Wikipedia talk:Content assessment/Archive 4

Proposal - adding C-class between GA-Class and Start-Class
While interest on this page is still high, let's try to decide on some substantial changes to the assessment scale. There were two major issues discussed in the section above: The positioning of A-Class, and whether there is enough granularity in the scale between GA-Class and Start-Class. From my reading of the discussion above, while we're still arguing about A-Class, there is enough of a consensus to pursue the second course of action. So, here is my proposal of the revised scale:

Discussion start convenience header
The main difference is in the text for B-Class and the proposed C-Class. We can argue about colors, precise wording and examples later (because I didn't pay that much attention to those things), but in general, B-Class would correspond to what is being proposed as B+-Class (I actually used part of the WikiProject Mathematics assessment scale for guidance), and C-Class would be equivalent to the less positive aspects of the current B-Class grade. In brief, B-Class articles would be somewhat close to what is currently considered a GA, but they would need some content and polishing to get there, while C-Class would be for articles that while complete, are not even close to either GA or FA status. The revised scale also clarifies that articles that violate copyright, neutrality or original research policies are generally not acceptable for stable releases (which was a complaint of the previous scale), and uses a whole letter to indicate that separation. So, what do you guys think? Tito xd (?!? - cool stuff) 23:05, 14 May 2008 (UTC)
 * Looks good to me. John Carter (talk) 00:22, 15 May 2008 (UTC)
 * Do we really need another assessment class? What problem is this addressing? I think there's a high risk of making assessment too fine-grained to where contributors start using too much of their time re-assessing (especially if this forces projects to re-comb their articles) when that time could be better spent actually improving the article. Remember, the purpose of the assessment is not so much to have a handle on the individual article, which usually moves to the next level up a while before a re-assessment reflects it, but instead to look at the larger "forest-from-the-trees" view at the project level as an aggreggate of article statistics. Creating another assessment level is not going to substantially help the projects' idea of where they stand, but it will create a great deal of new work for them in order to determine which of their articles should now be populating the new class. Girolamo Savonarola (talk) 00:30, 15 May 2008 (UTC)
 * (See also my comments at the bottom of the above thread.) Girolamo Savonarola (talk) 00:34, 15 May 2008 (UTC)
 * The problem this proposal addresses is that there are two types of B-Class articles. There's the articles that are very close to GA, but that have something missing; and then there's the articles that while complete, just flat-out suck. The previous B-Class rating allowed articles with copyright problems to be rated as B's, as well as articles with neutrality or NOR issues. In a way, WP:1.0 would get in massive problems with copyvios, so we want to clarify that such is not acceptable anymore, while being realistic that many B-Class articles will form part of our static releases. I too was not 100% convinced about the necessity of this class, but after thinking about it, I see how it is beneficial. Tito xd (?!? - cool stuff) 00:41, 15 May 2008 (UTC)
 * Regarding Girolamo's comments about reassessment, it might be useful if more projects added a "last assessed in" parameter to their banners, so that they can periodically review articles for improvement, etc. I would add it myself to several banners if I had a clue how to set it up. John Carter (talk) 00:45, 15 May 2008 (UTC)
 * Um, I thought that was the point of the assessment tables themselves? I mean, Version 1.0 Editorial Team/Tropical cyclone articles by quality/1 contains the assessed old-id for all the articles in that page. Besides, unless we have the equivalent of Gimmebot going around and filling out those parameters, editors won't fill out that parameter. Tito xd (?!? - cool stuff) 00:48, 15 May 2008 (UTC)
 * You might be right, but depending on how many pages like that an individual project has to go through, it can be a daunting task, particularly as I've never found a way to get them to arrange in sequential order on those pages. John Carter (talk) 01:03, 15 May 2008 (UTC)
 * I'm not really sure what you mean by ordering the articles sequentially, as the pages go in an Quality -> Importance sort automatically. That said, there are current discussions for a major overhaul of User:WP1.0 Bot, so that could be a feature that could be implemented eventually. Tito xd (?!? - cool stuff) 01:08, 15 May 2008 (UTC)
 * I, for one, would love to include a 'C' ranking. Because it's simply adding a level to an existing system, nobody would be required to adopt it. Additionally, it's very true that there is a vast difference between "start" and "GA", which the single label of "B" simply isn't always sufficient to cover, especially in very large projects with thousands of articles of "start" quality . You have my full support. – ClockworkSoul 14:34, 16 May 2008 (UTC)
 * I don't have a strong feeling either way, though if I had to choose I would probably prefer to include an extra C-Class. Girolamo clearly lays out the principal objection - I'd like to see User:Holon's take on that.  However, I think we now have enough expertise in the community that we could probably manage C-Class.  The other main objection to C is that many projects would need to re-assess all of their B-Class articles, though a few like WP:MILHIST might not need to. Interesting - this proposal is beginning to look suspiciously like the very early days...! On balance I think it would be good to add another level at this critical level in the scheme, and it would resolve differences in how different projects handle B at present.  Once people have made their arguments, let's put it to a vote, along with the GA/A vote also - we can use voting patterns as part of the way to judge consensus.  Walkerma (talk) 17:25, 16 May 2008 (UTC)
 * Looks reasonable to me... there are a lot of technical issues vis its implementation, but they're all fairly trivial. Happy‑melon 21:11, 16 May 2008 (UTC)
 * Fantastic, I'm 100% in favor of this proposal.  Arman  ( Talk ) 04:00, 17 May 2008 (UTC)
 * Yeah, this is only the drafting stages of the proposal. I intend to submit it to a more general ratification vote when we all agree that the wording and stuff is reasonable. Tito xd (?!? - cool stuff) 08:02, 17 May 2008 (UTC)


 * We will get a better idea of where extra classes/grades are most justified or needed based on the exercise you've undertaken and further exercises if they're done. I'll worth through that with you once you've had aligned your Chem assessments with the scheme by doing the further comparisons. Still amazed by the near perfect consistency in your ratings of the 27 Chem articles! Holon (talk) 08:59, 17 May 2008 (UTC)


 * I support the creation of a C class. Most projects have a very large number of "start" class articles which vary greatly in quality. The gap in standard between "stub" and B class needs some gradation as well as within B. Excellent idea. Sarah777 (talk) 09:37, 17 May 2008 (UTC)

I'm not so sure I like this idea. For one, people are going to focus even more on the exact assessment of the article than the actual quality. The asssessment scale is meant to give a rough estimate of how much work the article needs, not an exact science. In my opinion, there's not going to be that much of a difference either way. It's about the same things as B class. I just think we're better off keeping it simple rather than trying to make a larger, more exact scale which people are going to obsess over. Juliancolton Tropical Cyclone  12:05, 17 May 2008 (UTC)
 * Keeping it simple is definitely a key requirement, hence the negative reaction to the proposal here. However, as noted above, the assessment scale is used for other purposes, and its original task of selecting articles for static releases is still relevant.  The boundary for material which is suitable for inclusion in a static release is currently located somewhere in the middle of the "B" rating, so this proposal is mainly to allow us to more accuratley divide wikipedia's 50,000 B-Class articles into those that are suitable for static release and those that aren't.  Happy‑melon 13:05, 17 May 2008 (UTC)
 * I support this proposal. I can think of many articles currently classed as "Start" that seem way beyond that description, yet have to be rated that way because they are clearly not polished enough to be rated "B" class. It would also allow us to have some confidence in "B" class articles as the minimum for V.1.0. The open question for me is whether to merge "GA" and "A". I can see some advantages to doing that. Sunray (talk) 17:35, 19 May 2008 (UTC)
 * I definitely support this proposal. It would help prevent copyright violations from being in WP1.0 and, like Sunray said above, allow for more flexibilty between "Start" and "B"-class articles. -- Imperator3733 (talk) 19:56, 19 May 2008 (UTC)
 * I'm in full support, because currently "B-Class" is such a wildcard. I've seen articles that are deserving of B-Class, and I've seen articles that are not, but not definitely not Start-Class either which is why I am completely in favour of a new rating to distinguish the good and bad "B" articles. It current seems to be used as a middle ground between Start and GA, and that really isn't a good thing. -- .: Alex  :.  09:52, 1 June 2008 (UTC)

In the table above, classes B and C have the exact same example. That doesn't help understand the difference between B and C. :-) --Itub (talk) 12:01, 12 June 2008 (UTC)

Conclusions
I'm posting comments now, as I need to head to bed soon, and I may not have internet access for the next day or two. The voting will close fairly soon, but - barring any last-minute surges, it looks pretty clear that:
 * 1) There is definite interest in a new C-Class, though a similar number of people are opposed.
 * 2) There is less obvious support for putting A below GA, and some are strongly opposed to that.
 * 3) There is moderate support for removing GA from the scheme altogether, but some strong opposition to that idea as well.
 * 4) Other options do not seem to have gathered "traction".

Regarding #2, I don't think there is a strong consensus for change, so I would judge that we should leave things the same. Regarding #3, I actually made made such a proposal in December 2006; although many supported the idea then, those opposed were much stronger in their views, and there certainly wasn't a strong consensus for change. The arguments made now seem similar to then, and despite my personal support for #3 I don't think the consensus for change is there now. If the bot is amended to indicate GAs in the table automatically (like we do for Version 0.5 content) then perhaps we could initiate another discussion on that topic.

Several of the "no change" voting people indicated clarifying the current system - we are in fact doing that very thing. Please help out with this!

That leaves #1. I think we do need a second round to get a wider set of opinions before we decide either way. My own thought is to gather further opinions by spamming the WikiProjects. After all, that is where the assessments are coordinated. But I'd like to get people's thoughts on what to do next. I'm proposing: Please give suggestions below. Walkerma (talk) 22:54, 25 May 2008 (UTC)
 * 1) Titoxd and Holon (if they're willing!) and any other volunteers should get a few good examples of articles indicating how things work (at least supposedly!) now, and how those same examples might be evaluated under the proposed S/S/C/B/A &(GA/FA) assessment system.
 * Just us two? The more the merrier... ;) Tito xd (?!? - cool stuff) 19:20, 27 May 2008 (UTC)
 * Definitely the more the merrier!
 * 1) Work out how best to gather consensus - voting? If so, using approval only, or yes/no?  I actually like how the things worked above - you can count approval votes, but also read support and oppose comments to gather people's views.  If we do go for a simple "To add C or not to add C" poll, that should be simpler to judge than the above discussion.
 * 2) Then ask WikiProject members to vote/comment/whatever on the add C/don't add C question.


 * Comments:

These examples may suck, as they're as usual about hurricanes, but since the articles are similar, we can compare and contrast more easily, I think. Any other examples? Tito xd (?!? - cool stuff) 10:02, 29 May 2008 (UTC)
 * Thank you for calling this poll, for presiding over it, and commenting on it so eloquently. A new vote on C or not-C is the obvious way forward. My own interest is of course in the GA separation issue. I was really surprised by the fact that very few editors commented on this issue. There was some support from editors here, but also objections from significant GA regulars. This poll, unfortunately, occurred just after a poll about the introduction of the GA green dot into the mainspace. Consequently, I believe, insecurity about GA fed into the opposition. I hope a clearer case for separation can be made in the future. Geometry guy 21:44, 26 May 2008 (UTC)
 * When we reach a consensus on this, could someone post the final decision at User talk:Pyrospirit/metadata? That way I can adjust the assessment script/gadget to reflect the new scale. I'll be rather busy with final exams over the next couple weeks, so I might not have time to check this page. Pyrospirit  ( talk  ·  contribs ) 00:13, 27 May 2008 (UTC)
 * Definitely, and Happy Melon as well - these scripts are invaluable for people doing assessments. Thanks a lot for your support with this work. Walkerma (talk) 23:11, 28 May 2008 (UTC)
 * Aye, thanks Walkerma, it's good to revisit these things now and then ;) Perhaps next time if we propose to separate both FA and GA from the scale... EyeSerene talk 09:26, 27 May 2008 (UTC)
 * I believe a final up/down !vote is the way to go as well. Tito xd (?!? - cool stuff) 19:20, 27 May 2008 (UTC)
 * Yes, I agree that we will probably need to revisit the GA separation issue at some point. That wasn't on the original agenda for this vote (I didn't want to be seen as a "bad loser" by pushing it a second time!), but on balance it is something we should probably debate again.  What I'd like to do is to see about getting the bot to display GA (and maybe FA) status in the tables automatically - which should be easy - and then perhaps it would be appropriate for us to debate this topic.  Thankfully we don't need to worry about that issue while rewriting the details of the criteria.  As for the current debate, it looks as if people are OK with a simple (but wider) yes/no vote on C-Class, but I'd like to see some good examples before we open the polls. Walkerma (talk) 23:07, 28 May 2008 (UTC)
 * I'll get working on those tonight. I'd really prefer we open them soon, though, because that way, we still have the inertia of the previous discussion with us, which will hopefully help us gain more consensus. I'll also spam Watchlist-notice with this vote when it starts, as nobody objected to me doing so when I asked about it. Tito xd (?!? - cool stuff) 05:54, 29 May 2008 (UTC)
 * Well, I didn't have that much time to be able to make a fully-fleshed out example comparison, but here's what I would consider examples of Start-Class, C-Class and B-Class, using articles I'm familiar with:
 * C=Terracotta Army; B= Burr-Hamilton duel; A=Freyja? Ling.Nut (talk) 11:43, 29 May 2008 (UTC)

Here's examples from what I'm used to:

These are good examples IMO, and I feel that USRD will be having a field day with C-class. Mitch 32contribs 12:00, 29 May 2008 (UTC)

Let me make a comment here about how we describe these classes and cite examples. Most of the descriptions I've seen of Start- and C-classes are rather negative – they focus on faults, deficiencies, and problems, such as (from above) "completely unreferenced...lacks a lot of content...issues with the prose...weak and short lead section...poorly illustrated." All of this is no doubt true. Yet let's think about the "preferred" progression of an ideal article from Stub- to A-class. It would be possible for a thoughtful editor to husband an article through Stub-, Start-, C-, and B-class statuses over the course of weeks, as it evolves in public through a series of carefully written versions. The Start version might thus consist of an introduction and a good outline. The C-class might fill in major details and supply key references, and might begin to draw on a pool of other editors with subject matter expertise. In the later versions, in-line references and topic depth would increase. This process might in fact be an ideal model for article development, rather than expecting them to burst on us "from the brow of Zeus" in complete A-class glory. My point here is that we have tended to describe the lower classes in a pejorative way, despite the fact that every good article will usually go through these stages. A good Start-class outline might even be more useful than a prosaic B-class treatise. In other words, I think there's such a thing as a good Stub, a good Start-class article, and a good C-class article, and I think we should be offering guidance on how to create these in a way that will their facilitate later expansion into B- and A-class articles. Trevor Hanson (talk) 18:59, 29 May 2008 (UTC)
 * Or to put it another way, a good series of class examples might show the same article in a progression of stages, showing how it improves, rather than a series of different articles each with problems needing correction. I haven't found a good one to cite, but perhaps somebody can suggest one. Trevor Hanson (talk) 19:02, 29 May 2008 (UTC)


 * Trevor, this is an excellent point! Start-Class articles are frequently perfectly adequate for most purposes, and many print encyclopedia articles are just stubs.  I think we could do this "article timeline" using article history.  If you could find a suitable set of article versions, we could include it in the poll description, and also afterwards in the assessment scheme (which we're rewriting this spring).
 * I will look. Here is one sequence, based on the evolution of Mobile Bay jubilee: as stub, as Start-class, as C- or B-class. However, I think we can find examples showing a more dramatic progression. Active editors reading this can probably think of articles that have evolved well and smoothly under their stewardship; I encourage you to spotlight your own examples. Trevor Hanson (talk) 20:21, 30 May 2008 (UTC)


 * Dr Cash (see "another option" below), I happen to agree with you; but at this point I’d like to get the C/notC vote done first. After that is sorted out, I’ll initiate a discussion on the GA/FA topic, with three options – no change, remove only GA, remove GA/FA.  But I’d like to find out how the bot can be set up first (I don’t think we’ll even need a separate template parameter, actually).
 * Regarding the examples, thanks for a lot for these! I'm a little concerned that the standards may be set a bit too high; I think they probably represent the standards in use on a few active projects, rather than on an average WikiProject.  I clicked “random article” a few times and this is what I found as current assessments, and I think these represent more typical examples:
 * ’’’Start’’’: Charles-François Bailly de Messein, 501st Legion, Moirae, Douglass High School (Georgia), Lewis Wade Jones, Spartacist League, Lake Pontchartrain Causeway, Nicolai Gedda, Ba F.C., Fruitcake.
 * ’’’B-Class’’’: Prevention of dementia, Vibraphone, Sonic the Hedgehog (video game), Alexander Woollcott, Elaine Belloc.
 * This selection is made up of the first ten Starts and the first five B-Class I found. Please don’t “tut tut” if you disagree with some assessments, as this will only prove my point – this represents a cross-section of assessments as done now  You’ll see that some Starts are very basic, whereas others fit well with the C-Class definition.  The new proposal does involve raising the standards for B considerably, as WP:MILHIST has already done for some time (see this and this), and this change will mean that some of those Bs will become Cs.  I’d like to invite people to do one of Holon’s assessments on this random set.  If you’re interested, please email me (my username AT potsdam DOT edu), and I will try (I’m on vacation) to send you an excel spreadsheet promptly to use for the analysis.  We can put this information together to get a random selection across the topics. Walkerma (talk) 00:21, 30 May 2008 (UTC)


 * Another option:

I still remain quite strongly opposed to any proposal to decouple the GA process alone from the wikiproject assessment system. GA is as much a part of the wikipedia-wide assessment system as FA (and PR, for that matter). Despite whatever negative feelings that Raul, Sandy, or any other FA regulars have towards GA, it isn't going away, and shouldn't be discounted as the 'odd man out' in the assessment scheme. That would do nothing more than to just demote the whole process and ultimately lead to its demise. Furthermore, wikiprojects are interested not only in which articles are A, B, start, stub class, but also which of their articles are both FA and GA (e.g. those neat little, bot-generated tables listing the number of articles by class and importance for each project). So I think we want to keep GA in this system somehow, since it's still involved.

However, I think a better solution would be to decouple both GA AND FA from the wikiproject assessment scheme together. Let's let the wikiprojects assess their articles on a simple grading scale on their own (e.g. A, B, C, start, stub). Yes, let's add C to this system, since we're removing GA & FA from the class= parameter. Since GA & FA are both independent projects from the wikiprojects, they should both be separated from the immediate assessment system. I think they should still be included somehow, though. Maybe have a status= parameter for each wikiproject tag, for either a 'GA' or an 'FA'. Ideally, articles should achieve the A-class rating across all relevant wikiprojects prior to being nominated for FA. However, FAs may not remain at A-class; if the article degrades after a year or so of being FA, a wikiproject could reassess it as B-class, and this might ultimately aid WP:FAR in reassessing the FA status. Similarly, I would expect all GA nominees to be at least B-class; but having this separate would give WP:GAR another metric to help gauge the article's current status. So, we would have A-class featured articles, B-class featured articles, A-class good articles, or B-class good articles. We'd also, of course, have regular A-class and regular B-class articles, without being FA or GA. Likewise, the bot that keeps track of the number of articles for each assessment type should still be able to easily keep track of all of the assessment grades, as well as GA & FA status together. Dr. Cash (talk) 22:47, 29 May 2008 (UTC)


 * Comments: In the examples given in both cases (cyclone and route), the start class and C-class articles seem too close in quality. The current description and example of Start class in the "official" grading scale seems much more lenient than the new examples. One more suggestion: We may consider locating two examples for each class one for the lower limit and the other for the upper limit.  Arman  ( Talk ) 02:30, 30 May 2008 (UTC)
 * Procedural comment – Please replace the plain links with permanent ones, or the examples are likely to soon slide out of their current classes due to their potential improvement. There is a reason why this is the case with the master table on the main page. Waltham, The Duke of 17:22, 30 May 2008 (UTC)

The next poll
(Most of my new comments are above.) Should we start the new poll (C or not C) on Sunday/Monday at midnight or so, or do we need longer? How should we contact WikiProjects? Is there anyone who can use AWB to spam the project talk pages, or should we just add a message bar on the top of the assessment template? Walkerma (talk) 00:21, 30 May 2008 (UTC)


 * Could you not attempt to get a site-wide message put up? -- .: Alex  :.  09:50, 1 June 2008 (UTC)


 * I suggest lodging a notice at the WikiProject Council; many people from various projects watch this page, which serves as a centre of co-ordination. Waltham, The Duke of 11:03, 1 June 2008 (UTC)
 * We intend to do that, and to cross-post to several pumps and noticeboards. Tito xd (?!? - cool stuff) 06:35, 2 June 2008 (UTC)


 * Walkerma has asked me to start the vote in case he isn't available, as he's on vacation and has sketchy Internet access. (I'm not comfortable closing it, so let's hope he comes back before that.) In either case, would we be okay starting the vote by Wednesday, June 4, at 0000 UTC? Tito xd (?!? - cool stuff) 06:35, 2 June 2008 (UTC)


 * I'll be back by June 14, so I can close the vote after that. I think we should have a full fortnight (2 weeks) for the vote, is that OK?  Tito, can you add a tag at the top of the assessment template to announce the poll? Thanks, Walkerma (2 June about 18:37 UTC).

Evolution of an article - an example
This example of the article "Atom" may help demonstrate the use of a new "C-class":
 * The article started as a stub [] on Oct 1, 2001.
 * By 8 October, 2001 it received some additions and approached the upper bound of the definition of stub: []
 * On 20 September 2002, the article was enriched with some more information and it moves into start class: []
 * The version as of 3 June 2004 [] is still in start class. There is a meaningful amount of information - but it needs further structuring improvement.
 * On 24 June 2004, the article receives another important addition - a useful image. It has reached the upper bound of start class, but still not good enough to get a C-class rating: [].
 * On 18 September 2004, the article looks like this: []. Some of the sections have expanded and now it barely meets the C-class requirements.
 * By 31 August 2005 the article has been expanded further and can now be comfortably called a C-class article: []. The article's major shortcoming is in referencing.
 * The version of 12 December 2005 [] has enough content and structure of a respectable article. In spite of its lack of in-line citation, the article is approaching the upper limit of C-Class. If it were properly referenced, we could have considered rating it B-class.
 * After addition of several new images and contents from a book (now cited), the article looks like this [] on 19 August 2006. Though some editors would still hesitate to rate it B-Class, due to its lack of in-line citation; content wise it deserves a B-Class.
 * By 23 March 2007 [] new contents and references have been added and the article would now safely pass a B-class assessment by any editor.
 * The October 17 2007 version of the article: [] is nominated for a Peer Review. Review closes on February 9, 2007 leaving the article like this: []. The review ensures completeness of the content, addresses many MoS issues and inline citation problems and upgrades the article to A-class.
 * Next day [] the article is nominated and listed as GA.
 * On February 12, 2008 the article was nominated for FA and this version of the article: [] (as of 17 February 2008) was promoted.

Thanks a lot! This was a lot of work, I'm sure, but it will be a very useful guide, and it shows that articles don't burst on us "from the brow of Zeus" (to use Trevor Hanson's phrase). We'll add this into the main write-up for the scheme. Thanks, Walkerma (talk) 20:45, 18 June 2008 (UTC)

Redirect class?
I have noticed that during some article mergers, the original article assessments have stayed the same (e.g. Start class). Does the Editorial Team make allowance for "NA" (see NA-Class—"NA This page is a Redirect and does not require a rating on the quality scale."), for use on the redirects' discussion pages? Should it? If not, what should be done to these ratings? G.A.S 05:51, 2 June 2008 (UTC)
 * Not officially, but it is in use. I have also seen a number of project use a "redirect-class" for exactly the purpose you're inquiring about. There are many such "non-class classes" floating around. – ClockworkSoul 06:27, 2 June 2008 (UTC)
 * One use of a redirect-class that USRD has been considering adopting is using it to track "redirects with possibilities", such as articles merged to a list which may be split someday or other merges that may need to be resplit in the future. Redirects that are merely alternate names don't really need to be tagged. —Scott5114↗ [EXACT CHANGE ONLY] 01:10, 4 June 2008 (UTC)

The better thing to do in the case of redirects is to just delete any wikiprojects from the redirected artice's talk page, since there's no article. Adding a 'redirect-class' just adds unnecessary clutter to the wikiproject's assessment system. Dr. Cash (talk) 02:38, 4 June 2008 (UTC)

Expert consensus?
[I have relocated this discussion to Wikipedia talk:No original research, which is a venue where others with this interest are more likely to contribute.] Spinality (talk) 02:28, 28 June 2008 (UTC)

Suggestion: Abolish A-Class
In my view, I think A-Class should be abolished, as between GA and FA, many people tend to go from GA and jump straight to nominating it for FA-Class. The gap for A-Class seems to vague. I believe only one or two projects actually use A-Class properly. Looking at the statistics scale, 913 articles out of the 1,646,082 articles assessed are A-Class, that's 0.05% of articles, a very minority. Because of this, and my own opinion, I think the A-Class is a tad useless and should be abolished. D.M.N. (talk) 12:08, 16 June 2008 (UTC)
 * I don't like the idea. Many WikiProjects use A-class as a way to recognize featured-worthy content without the star. Juliancolton Tropical Cyclone  21:40, 16 June 2008 (UTC)
 * This seems like a good idea to me. What exactly is A-class?  Between B and GA?  GA and FA?  I've seen it both ways.  How can something be between GA and FA (both designations requiring peer-review) despite lacking the requirement for peer-review or community consensus?  Why should an article be classified as one thing according to one project, but A in another?  Shouldn't an article's quality be consistent regardless of which group of interested Wikipedians reviews it?  A-class seems to unnecessarily complicate matters, and as I mentioned in my comments above about the C-class thing, I think we need to simplify and clarify the assessment scales whenever possible.  Getting rid of A-class would accomplish both. Drewcifer (talk) 07:45, 17 June 2008 (UTC)
 * At Milhist, we use A-Class as a project review immediately prior to FAC. It requires three supports and no substantial opposes for promotion. In my view, this is a much more testing and meaningful review than GA, which requires only one editor to promote. GA reviewers often have no background whatsoever in the subject matter and the reviews vary enormously in quality and throughness. -- R OGER D AVIES  talk 07:53, 17 June 2008 (UTC)
 * I also don't like having the A-class. It kind of makes it an exclusive class for projects that have a large member base. <b style="color:#0000FF;">OhanaUnited</b><b style="color:green;">Talk page</b> 15:51, 17 June 2008 (UTC)
 * Why do you need a large member base for A-Class? -- R OGER D AVIES  talk 04:33, 20 June 2008 (UTC)

Anyone support a "Ratification vote" like above with the proposed C-Class to get a good consensus? D.M.N. (talk) 15:56, 17 June 2008 (UTC)
 * Personally I'd like to see the dust settle on this vote first. Whatever the outcome of the poll, we will be revamping the classes, and rewriting the definitions to make them sharper.  Once that process is complete we can discuss this.  Personally, I'd like to see it modelled on the MILHIST system of WikiProject peer review - this would have been impossible a couple of years ago, but should be possible for many WikiProjects now.  I don't see that as exclusive - it provides an incentive for smaller groups to get active.  WP:CHEMS developed the original scheme when it had only about 5-10 active members. Walkerma (talk) 16:19, 17 June 2008 (UTC)
 * I agree: let's let the dust settle before taking on the A-class thing. But that doesn't mean we can't talk about it in the meantime.  Anyways, I've already described my issues with the A-class thing up above, but I want to mention that I don't think what the Military History project described (as their precursor to FAC) is much of a defense of the class.  The project can still do a project-wide review (and it's great if they do), but I don't see a point in having that result in another class-change merely in preparation/anticipation of yet another class-change. Drewcifer (talk) 03:04, 18 June 2008 (UTC)
 * I dislike dust. :-)
 * Still, I agree. Personally, I don't find it necessary to abolish A-class, but I do not mind discussion after some time passes. Waltham, The Duke of 07:29, 18 June 2008 (UTC)
 * Let's wait until the first poll closed first. We don't want to start too many fires at the same time. <b style="color:#0000FF;">OhanaUnited</b><b style="color:green;">Talk page</b> 13:48, 18 June 2008 (UTC)

Well, if a project really needs finer levels (let's say Project XYZ want to identify those B-class articles ready for GA), then they can simply tag those articles doubly into two assessment categories (like B-class XYZ articles and GA-ready XYZ articles). There's no need to have obsolete or additional classes. (But I find it interesting that the A-class scheme seemed to have predated the GA project according to Walkerma below. That's probably why we are in this FA/A/GA mess right now.) --seav (talk) 03:06, 20 June 2008 (UTC)

This is an extremely bad idea. If we delete the A class after adding the C class, we are right back where we were with six classes before the addition. The natural progression from stub-start-C-B-A-GA-FA is the way to go. GA and FA are community processes with formal reviews. C-B-A are (or can/should be) project processes, however they set up their assessment process. Stub is essentially automatic, and the start class can be applied by any established editor, except the creator. At least that's my view of the organic whole. Part of the problem is that A class, to some people, is between GA & FA classes, when it should be below the GA, as the assessment process is quite different. Keeping the A class within a project, and below GA should help focus content improvement using project expertize, as C-B-A rungs in the ladder. Isn't that what we were looking for, more rungs in the assessment ladder to help article assessment by eliminating big jumps between classes, especially between start & B? — Becksguy (talk) 03:38, 20 June 2008 (UTC)


 * Does much of this problem go away if we remember that an article need not be A-class before submittal to the GA-class review? The non-peer-reviewed process goes Stub/Start/C/B[/A], followed by submittal for peer review. The two peer-reviewed processes go  /GA and  /FA. In some WikiProjects, A-class is a convenient internal quality level – an FA-class "release candidate" after having achieved GA level. In other WikiProjectss, where such a formal internal review process does not exist, it could make more sense either a) to reach GA after first achieving A-class, or b) to omit A-class entirely, and complete final tweakage in the peer-reviewed step from GA to FA. Each sequence of levels can make sense within a particular set of editorial resources and procedures, and can use the same quality standards. What varies is whether an additional WP-internal review occurs between FA- and GA-class; and both approaches seem reasonable. Spinality (talk) 06:52, 20 June 2008 (UTC)
 * There's a few interesting things I can glean from some of the comments from above. First: there is still a confusion as to where to place A-Class.  At the risk of picking on the people who have been so kind as to provided the opinions: Becksguy says it goes A-GA-FA.  But if you look elsewhere (below for instance, A-Class is between GA and FA.  Second, and related to this confusion, is that different projects treat A-class differently.  Milhist uses it as pre-FAC.  Others might use it as pre-GAC.  Why should two inherently different (and internal) processes be assessed the same thing, and why do these processes even require a change in assessment-class anyways?  And lastly, a very good point from above: if A-class is to be understood as a WikiProject specific thing, then why reserve a separate class for it?  Why not let the WikiProject itself categorize the page internally? Drewcifer (talk) 07:06, 20 June 2008 (UTC)


 * If it's not abolished, it should at least be amended. I have no use for it when making assessments.  If an article is good enough for peer review, I send it to GA or FA; if it is not, I tag it a B-class.  --Arctic Gnome (talk • contribs) 03:00, 24 June 2008 (UTC)

I agree with D.M.N. I always didnt understand why have a A class and a FA class. Yes if we get rid of A class is will make a bigger gap between GA and FA, but come on, why do we need a A class. Gears  Of War  03:02, 24 June 2008 (UTC)


 * That reminds me of a discussion I had with a fellow editor regarding an article's assessment, which came down to one point: If you want an assessment for an article, the one the community decides is often most valued. The criteria between GA and FA class is so close to each other that there is very little room for A-class: by the time it can succeed in FAC; you nominate it there—not for A-class ("Good articles that may succeed in FAC should be considered A-Class articles"). It is only in the largest Wikiprojects with assessment departments where this may differ; as this is an opportunity to gain further input (i.e. yet another peer review).
 * With the criteria of B class going up as well, GA class seems to become only a formality itself (although valued); so this leaves me wondering about the use of A class; as often the Wikiproject would have provided input at B-class level: Why would it need to do so again at A-class review? G.A.S 05:27, 24 June 2008 (UTC)
 * With the criteria of B class going up as well, GA class seems to become only a formality itself (although valued); so this leaves me wondering about the use of A class; as often the Wikiproject would have provided input at B-class level: Why would it need to do so again at A-class review? G.A.S 05:27, 24 June 2008 (UTC)

Results of the poll
Ratification vote on C-Class

Quality assessment scale is Wikipedia-wide

Optional implementation... Reversed

Suggestion: rename to Start and Average

I just closed the poll as agreed, two weeks after it started. I will be going through all the comments (over 50 pages when printed out!) carefully over the next 24 hours, and I'll post a conclusion here. Walkerma (talk) 03:20, 18 June 2008 (UTC)

RESULT: C-Class will be added to the assessment scheme.

I (Walkerma) have spent much of today reading every word of the above. The simple tally of votes (152 to 106) corresponds roughly (IMHO) to the strength of support on each side. Here are the main arguments as I see them:


 * Arguments in favor
 * More refined definition of the Start/B area, which is a very large and important stage in article development. Some Bs are close to GA standard, others are very poor.
 * It's a lot of work to turn a Start into a "good" B – this would give editors something to aim for that is more within reach.
 * For the 1.0 project, we need to distinguish between Bs that are OK for publication, and those which aren't. We need to tighten up standards for B, to exclude those lacking sources or with other problem tags (NOR, POV, etc.).

More levels means more reassessment as grades change more often.
 * Arguments against
 * The purpose of assessment is simply to provide WikiProjects with a rough idea of where each article stands, and the current system does this perfectly well.
 * Keep things simple – we don't want people to obsess over the details of assessments instead of improving articles.
 * The system is unclear right now, and this will only make things even worse!
 * This will be a nightmare of work, trying to re-assess thousands of articles that have already been assessed as B or Start!
 * There are enough levels already. Dealing with so many orthogonal parameters (breadth, depth, refs, readibility, etc.) means that many assessments are already haphazard/arbitrary, this makes things worse.

(GA/FA/A discussions continue) Define the levels more sharply (being done anyway) If it passes, announce the new C-Class in the tables. If it passes, downgrade all Bs to Cs? Or perhaps just Bs with cleanup tags? Instead of an extra level, let's rename Start as C, raise the standards a bit for B and C, and make Stub-Class bigger.
 * Suggestions
 * Notes
 * I believe I (Walkerma) have been very conservative regarding changing the system – this is only our second ever major discussion on changes in over two years. I agree that we need to spend more time improving articles and less on discussions like this – but I also don't think the system should remain forever frozen with the 2005 standards.
 * The basic definition of B-Class was written in mid-2005, when GAs didn't exist, inline citations were rare and the vast majority of articles had no sources at all. Thankfully the situation has improved a great deal, and this change will allow us to tighten up the B-Class definition to represent where we are in 2008.
 * If you look in the archive, we had a lengthy discussion on the A/GA issue only last month and it was inconclusive. The aim was to consider the possibility of that issue at the same time as the new C (with one of the original options being "change A/GA and add C"), but only the "C alone" option garnered enough support to continue.  However, once the rewrite and the bot coding is done, we can perhaps revisit that issue.
 * The point has been made that "C-Class is optional". I expect that it would be very hard for a project to forbid its members from using the system – but of course, this is a choice that can be made, perhaps by limiting parameters in the template.

I believe there is enough support for the change to justify the addition of a new C-Class level to the assessment scale. This would be the first change to the scale in three years. Many of the opposing arguments can be somewhat mitigated while still adopting the new C-Class.
 * Conclusions

One of the main arguments against the new class is that the current system is already unclear. This whole debate originated from an initiative to sharpen up the definitions and examples – we plan to write a very simple set of definitions for newbies and non-assessors, and a more detailed set of definitions/range of examples to clarify the scale for those doing large numbers of assessments.

Another common thread has been to say, "Fix the current system (e.g. by raising the standard for B), don't add another level." But it may be very hard to impose the WP:MILHIST criteria for B-Class onto a small project – but having a new C-Class will (IMHO) allow us to harmonise the definitions of B-Class to this higher standard.

I don't think projects should immediately go on an assessment spree, unless they have a lot of outdated assessments and energy to spare. It should not become a nightmare of re-assessment; articles are evolving all the time anyway. Assessments should really be updated at least annually just to remain relevant – note that the bot-generated worklists include the assessment date (though the sort feature is rather broken). So I would say: just aim to keep assessments up to date, and after 12 months or so (if annual) all the articles will incorporate the new C-Class perfectly.

There was a fairly popular suggestion that all Bs should be demoted to Cs immediately, then reassessed upwards as appropriate. I would say that for a project using stricter standards than normal (e,g, MILHIST), this might not be the best way, but some projects may decide it is appropriate for them. I liked one revision of this, where only B-Class articles with cleanup tags are automatically demoted. We will see if someone can spare a bot for this.

Speaking personally, when this whole debate began (see archive) I did not have strong feelings one way or another (indeed my default position with such things tends to be "leave it alone!"), but the arguments presented above (and before) have gradually made me a mild supporter of the idea of a new C-Class. I think that a new sharper set of definitions and a higher-standard B-Class should alleviate at least some of the misgivings of those opposed to C-Class.

This change won't happen overnight. We will need to come up with really sharp definitions and great examples in order to make this work. We'll need to get templates and bots written. Please help where you can! We have seen a high level of debate and cordiality throughout, and on behalf of the 1.0 project, I'd like to thank everyone for their opinions. Walkerma (talk) 03:14, 19 June 2008 (UTC)

Comments
Just as a matter of point, I'm not convinced that a 60-40 vote shows any consensus. (Correct me if I'm wrong, but that wouldn't be threshold for any of our other processes which take voting into consideration.) Now, I'm not trying to throw a monkeywrench into the works or anything like that - whatever is decided is decided, but given the large quantity of votes cast, and the fact that a small percentage of defection (less than 10%) would have changed the majority, I don't think that a consensus itself was reached. However, if the judgment is that the supports made a more reasoned and thorough argument, so be it. As I said, I'm not against the decision, just the idea that somehow there was a clear consensus. I'm happy to move forward either way. Girolamo Savonarola (talk) 06:16, 19 June 2008 (UTC)


 * The only place where consensus means 70 percent supermajority is RfA: elsewhere consensus is determined by weight of argument in the light of the best interests of the encyclopedia, not by counting !votes. I have complete confidence in Walkerma's reading of the argument, and would like to thank him for his thoughtful closing of the discussion. I myself did not participate, and had no strong view on the merits of C-Class, but I will be happy to support it now that it has been introduced. Geometry guy 07:57, 19 June 2008 (UTC)
 * I concur with Geometry guy; I think Walkerma did a commendable job in analyzing and summarizing the positions, and in proposing an action that seems consistent and reasonable. The "oppose" comments were thoughtful, and raised important issues – issues that can and should be taken into consideration in the new implementation. They showed clearly that simply adding a new C-class and leaving the rest of the system unchanged would be a poor idea. It is thus important that we follow through with sharpening the assessment criteria and examples. They must provide clear guidance, both for document editing and for assessment, and must discourage practices that are against policy. These changes should make the entire scheme more useful, intuitive, and nonintrusive. Spinality (talk) 20:20, 19 June 2008 (UTC)
 * Thanks, Walkerma, for your analysis and hard work. I agree that only B-Class articles with cleanup tags should be automatically demoted to C-Class. As for the 60/40 vote, in a political election, that would be considered a landslide. It means that 50% more voters supported than opposed. ThreeOfCups (talk) 22:24, 19 June 2008 (UTC)

May I ask: Can we impement this now in talkpages by changes levels from Start to C or B to C, or should we wait? Just wondering, someone may jump the gun. D.M.N. (talk) 11:04, 19 June 2008 (UTC)
 * To get further comments, I have left a notes at WP:VPP and at WP:AN about the new C-Class. D.M.N. (talk) 11:10, 19 June 2008 (UTC)

Setting C right now will have varied and sometimes very interesting effects on banner displays and categorisation, depending on how rigorously the templates have been coded. Picking some examples from CAT:WPB, there are three likely scenarios. The most common code looks like this:

{{#if:{{{class|}}}|....... {{!}} {{{{{class}}}-Class}} {{!}} This article has been rated as {{{class}}}-Class on the quality scale.

.......

{{#switch:{{{class}}} |FA | fa     = |A | a       = |GA | ga     = |B | b       = |Start | start  = |Stub | stub    = |NA | na     = |#default = }}

Will display as "This article has been rated as C-Class on the quality scale", but will categorise into Category:Unassessed astrology articles. This is the most likely outcome. There may be the occasional template that has code like this:

{{#if:{{{class|}}}|....... {{!}} {{{{{class}}}-Class}} {{!}} This article has been rated as {{{class}}}-Class on the quality scale.

.......

}}

This is very poor code, as it will allow the page to be categorised into a (at this point probably nonexistent) Category:C-Class foo articles, or indeed Category:D-Class foo articles, Category:Cheesecake-Class foo articles, etc. Any templates which do this should probably be recoded. Then there are templates which already perform more rigorous checks on the class parameter: {{tlx|WPMILHIST}} will currently mark C-Class articles as unassessed; all banners using {{tlx|WPBannerMeta}} now correctly handle C, categorising into Category:C-Class foo articles. So in short, I'd recommend that we wait for the necessary (trivial) changes to be made to WP 1.0 bot, then start rolling out the changes to banner templates in a systematic way. What we really need in the "offers of help" section below is people who are comfortable with template code such that they can help go through CAT:WPB to make sure that all banners handle C-Class in a sensible way. We also need to create Category:C-Class articles and populate it with all the Category:C-Class foo articles subcategories: this could probably be automated, cloning categories from, say, Category:B-Class articles. There's quite a bit of infrastructure to develop before we can really get this show on the road... and of course we need to rewrite the assessment criteria, preferably in a more objective style. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 11:33, 19 June 2008 (UTC)
 * H-m, the WPBannerMeta almost supports the new C-Class. It filters the pages correctly into the C-Class category, but with "class=C" (or "class=c") the message in the Project banner says that the page is "not an article and does not require a rating on the quality scale".  (But thanks to all involved.)  --Craw-daddy | T | 15:04, 20 June 2008 (UTC)
 * Fixed <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 17:50, 26 June 2008 (UTC)
 * Are you suggesting a new Cheesecake-Class? Great idea.  I presume that such an article would have to be quite firm (not as firm as FA), but digestible; perhaps the reader's experience would be "pleasant, with a sweet, slightly fruity flavor"?  Everyone loves cheesecake - let's have another poll! :) Walkerma (talk) 02:26, 20 June 2008 (UTC)
 * Mmm... cheesecake. No assessment debates there. By the way, the WP:1.0 bot now recognizes C-Class. Tito xd {{sup|(?!? - cool stuff)}} 07:04, 20 June 2008 (UTC)


 * That cracked me up, Walkerma! Maybe something for next year. Ok, so if 1.0bot is ready, step 2 is to create and populate Category:C-Class articles. Otherwise we're going to be making a lot of category redlinks in the near future.  Something to ask each project to take care of for itself? Or to do manually? Or automate? <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 08:43, 20 June 2008 (UTC)

WikiProject Banners Assessment Levels
Why is it that nearly every WikiProject banner includes a field for its assessment level which needs to be individually done for every banner? 99% of Wikipedia articles (basically every article that isn't rated as A-class by one of the relevant WikiProjects) are rated the same across the board, so why do it individually? In other words, a Stub is a Stub is a Stub, regardless of what angle (ie WikiProject focus) you're viewing it from. So this begs the question: why clutter the talk pages with repeated information? And, follow-up question: why is this form of assessment-per-project favored despite the fact that many articles are "unassessed" by some of the relevant WikiProjects, even though they have been assessed by others? This method just creates a big mess and a huge amount of redundant work for (as far as I can tell) no rational purpose. Drewcifer (talk) 07:51, 18 June 2008 (UTC)
 * The problem is that an article might be excellent in one area, yet completely miss the most relevant details that would interest other projects. For example, Labor Day Hurricane of 1935 doesn't contain nearly any information within the purview of WikiProject National Register of Historic Places (which has some interest in the article), yet it is adequate in the tropical meteorology department, so WikiProject Tropical cyclones assesses it higher. Tito xd (?!? - cool stuff) 08:07, 18 June 2008 (UTC)
 * This is true, but not reflected in the global assessment scheme, which syas that an article's coverage of its topic should be assessed, not the coverage of the aspect of its topic that a particular WikiProject is interested in. Of course most Wikiprojects have a localized assessment scheme, and that allows for these variations.
 * Really, whether this individuality is good or bad depends on your point of view whether assessments should be used to organize a Wikiproject's todo list (in the example above, there won't be anything to do for the cyclones project, while there is work to do for others) or to evaluate the global "quality" of Wikipedia (in which case an article that is "partially good" actually isn't good). Kusma (talk) 08:31, 18 June 2008 (UTC)
 * This is done to provide Wikiproject level statistics and categories. (For instance WikiProject Military history/Assessment and Category:A-Class_military_history_articles). I presume the template syntax would be too complicated to add them all in a single template. There was a bot (User:Bot0612) that ran through articles and set the assessment levels to be the same, not sure why it isn't running anymore. G.A.S 11:40, 18 June 2008 (UTC)
 * I seem to remember a bot doing that as well, but if I remember correctly it was pretty short-lived. But you do bring up a possible solution: a single assessment template.  All of the WikiProject banners stay, minus the assessment fields, and a single "B-class" template tells us what the quality of the article is.  As for the idea of different projects rating an article differently, I completely understand that (and confess that I didn't think about that), but cases such as Labor Day Hurricane of 1935 seem like the exception not the rule.  Is it really worth keeping a broken system going for the occasional exception when it creates such bigger messes elsewhere?  Every time a new WikiProject is created, this adds hundreds to thousands of talk pages (depending on the project of course) that need to be re-assessed for the second, third, fourth time.  Can't our resources be better spent somewhere else? Drewcifer (talk) 17:37, 18 June 2008 (UTC)
 * As noted above there are and have been bots that do auto-assessments. I know BetaCommand ran his bot to do that for awhile; but he updated all the templates.  The one mentioned above only does the specific template it's looking for.  §hep   •   ¡Talk to me!  17:42, 18 June 2008 (UTC)
 * As far as I know, no bot is currently auto-assessing articles. I was actually a part of the BetaCommandBot's little experiment with the auto-assess, and it seemed to work fine but I guess that part of the bot is no longer operational.  But that fact that "a bot can do it for us" still doesn't justify a redundant system in my mind.  The pros and cons of the assessment system remain the same, regardless of whether a bot is supposedly doing this stuff for us.  Redundancy is still redundancy whether a human is doing it or not. Drewcifer (talk) 17:50, 18 June 2008 (UTC)

I don't particularly like the idea of all WikiProjects banners reading the same assessment (for a given article) all the time - what I'll call uniform assessment. Controlled auto-assessment (a bot goes through a specific set of articles on a specific date and assesses them for a specific project) is one thing; uniform assessment is a completely different thing. I agree that usually assessments do all match, but that may be as much due to laziness as anything else. The WP1.0 selection bot looks at each WikiProject separately, so if an article is ranked by five different projects it will consider the article five times; in other words, the individual WikiProject assessments DO matter for article selection. This is a Good Thing, IMHO. Some pros/cons I can see with auto-assessment and uniform assessment:
 * As Tito argues above, projects may see the same article differently. An article on a metal may cover the physics, geology and metallurgy well but be weak on the chemistry. Although the quality assessments are mostly independent of the WikiProject doing the assessment, there is clearly an element (no pun intended!) of perspective - the Chemistry folks may object to their "Start-Class" being raised to B because someone from metallurgy thinks it is B-Class.  As things stand, the chemists can note, "This needs a lot of work" but the metallurgists can say, "this covers the metallurgy pretty well".  We could even end up with two WikiProjects arguing over which assessment is right!
 * Related - different projects define the classes slightly differently. We at WP:CHEMS got a bit irritated when an outsider decided to downgrade several of our A-Class chemicals articles because (I think) they didn't have inline citations - these were articles that had passed a WikiProject-based peer review.  The person explained later they were using the WP:MILHIST assessment criteria.
 * If I'm from a small project, I may think that an article has reached B-Class, but with a uniform assessment banner I may be reluctant to change the assessment for eight WikiProjects. The result - many articles may languish below the level they deserve.
 * The strongest argument in favour seems to be that a new WikiProject (call it X) needs to perform many assessments on articles that have already been assessed, and which may already have 3 or more matching assessments. I think in such cases it would be fine (if WikiProject X agrees to it) to have a bot copy over assessments from other banners into the banner for X.  After that point, though, X will probably want to have the freedom to update its assessments; this would be much harder with uniform assessment.
 * An argument in favour of uniform assessment is that assessments would perhaps be kept more up to date; if an article improves from Start to B this will be picked up much faster if any one of several projects can revise the assessment for all.

Overall, this comes down to an argument between project independence versus efficiency. The best solution, IMHO, would be to have a bot that could (a) On demand, copy over any existing assessments from articles into WikiProject:X's template, and (b) Provide a list of all articles tagged by WikiProject:X that have differing assessments. This solution would allow efficient use of project members' time, but still allow full independence. Walkerma (talk) 20:34, 18 June 2008 (UTC)


 * I think it is the biography project that had an auto-assessor bot, don't know how it worked but it left a tag saying 'auto' so it would need reviewing if this was linked to a category it an ability to but in a 'no' so that the bot wouldn't asses any tagged that way it could be a useful combination, as most projects will want the same tags (one project coudl have worked on it improved it but the otehr project may not have go to it yet) but still have the ability to mak it manual is differin levels were needed. --Nate1481(t/c) 10:50, 19 June 2008 (UTC)


 * Taking as an example two projects I've been assessing for lately: Christianity and Saints. I've yet to encounter a Saints article that wasn't also a Christianity article, so they are commonly on the same page.  These have slightly different criteria for the different grades; Saints requires at least one reference for B class in addition to the 1.0 criteria, Christianity does not.  So in theory an article might meet the B criteria for Christianity but not meet the B criteria for Saints.  A different recent example was an article that a TV project member had assessed as B, I read it and thought 99% of the content was inappropriate or irrelevant and assessed Stub for Christianity and asked for a third opinion.  A different member of the TV project came, cut much of the content and did some other tuneup then assessed as Start.  With just one assessment parameter, this wouldn't have worked as well.  Other differences between project grading schemes exist now (MILHIST) and will in the future, and should be allowed to continue to exist in the future so that best practices can evolve.  GRBerry 13:11, 19 June 2008 (UTC)

The reason why I think that the project banners have these quality level parameters in their templates is because the article can then automatically be included into the relevant xx-class categories with just a simple addition of the template. De-coupling that parameter from the banners means that this auto-categorization can't be done efficiently. That said, I am in favor of a Wikipedia-wide uniform assessment scheme. If an article isn't well-rounded in scope, then how can it be a good Wikipedia article? As I said in a section above, the Assessment scheme is meant to grade articles for fitness into a Version 1.0 edition of the Wikipedia. The decision to involve WikiProjects is so that the assessment effort can be distributed. The Assessment via WikiProjects is therefore just a means to the Version 1.0 end and the fact that WikiProjects find use in it themselves is simply incidental. So, I'm very much in favor of having a clear Wikipedia-wide criteria for the assessment levels, possibly adopting the best practices that have evolved thus far now and then moving from then on with close coordination of the WikiProject Council. --seav (talk) 02:55, 20 June 2008 (UTC)


 * To sum things up a bit, the real question this issue seems to have brought up is the following: should assessments be applied to the article as a whole or on a per-emphasis basis? This question seems perhaps a bit more relevant than my initial proposal. Drewcifer (talk) 06:19, 20 June 2008 (UTC)

Added C-class to the grading scheme
I've added C-class to Template:Grading scheme. To give other people a chance to look it over first, I enclosed it in  tags so that it won't show up anywhere but the template page itself. Do the wording and example I used seem accurate and representative of consensus? Pyrospirit ( talk  ·  contribs ) 18:21, 19 June 2008 (UTC)
 * The next thing for us to do is completely rework all the assessment criteria, so I wouldn't be in too much of a rush. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 18:55, 19 June 2008 (UTC)
 * This was brought up before, and it looks like it's been changed, but I'd like to echo the concern that the colors for B- and C- class are too close together. This really isnt much of a problem, but the colors should be independently identifiable.  I would suggest darkening start and stub, making room for C-class to be a little more orange.  Again, as before, not the most pressing issue.- Running  On  Brains  19:34, 19 June 2008 (UTC)
 * Agreed. I personally think this color is slightly darker and better. Juliancolton <sup style="color:#666660;">Tropical <sup style="color:#666660;">Cyclone  19:39, 19 June 2008 (UTC)
 * The colors are not supposed to stay that way. I literally plucked it out of where the sun doesn't shine, so that's something else that we need to discuss as well. Tito xd (?!? - cool stuff) 19:45, 19 June 2008 (UTC)
 * Whatever color is chosen it would be best to define it on a hex code rather than a name. I'm all for "#ffcc66" (This color) which is half-way between the current colors for Start and B. I we go closer to the goldenrod we should make Start darker as well. Adam McCormick (talk) 03:13, 20 June 2008 (UTC)

Your template refers to a "B-" class in the B-class section. I don't recall any discussion to implement a "B-" class; when did this occur? Dr. Cash (talk) 20:49, 19 June 2008 (UTC)`
 * Does it? Where? Unless you're misinterpreting it where it goes "B-" then says "Class" on the next line, really, it needs extending a bit the column. D.M.N. (talk) 20:54, 19 June 2008 (UTC)
 * EDIT: I've made a change so it doesn't go onto a next line. D.M.N. (talk) 20:57, 19 June 2008 (UTC)

Colors
I seemingly have a tendency to get in color debates, but what the heck. Here's how these colors would look:

I personally think that the transition between B->C->Start is too light on the last one, and the goldenrod is a bit iffy on saturation values. I think we may have to modify the colors for Start and maybe B as well. Tito xd (?!? - cool stuff) 05:49, 20 June 2008 (UTC)
 * Just a comment here... The background color of the project boxes should also be kept in mind here. At the moment I write this comment, the color for C-Class articles is nearly indistinguishable from the background box color.  This, simply put, is bad.  I think there should be some clear contrast between the colors.  If this requires changing one or more colors of the classes, then so be it.  --Craw-daddy | T | 15:09, 20 June 2008 (UTC)
 * Perhaps the same color as goldenrod but with modified saturation levels would work better. Pyrospirit  ( talk  ·  contribs ) 16:58, 20 June 2008 (UTC)

This whole colors issues sways me more in the direction of replacing start w/C and making the stub class bigger. Picturing the table that way makes sense. But in anycase I prefer #ffcc66. Kevin Baastalk 17:46, 20 June 2008 (UTC)
 * Or maybe 'C' can take on the color of 'Start' (orange) and 'Start' can be made more red. I added a column for this option. Kevin Baastalk 17:52, 20 June 2008 (UTC)
 * How about we make the GA a deeper green, make B a lighter green and C as yellow? Plus we already have the right colour symbols to do that! [[Image:Symbol b class (alternate).svg|15px]] [[Image:Symbol c class.svg|15px]] -- .: Alex  :.  20:35, 20 June 2008 (UTC)
 * I've added an example. Not the exact shade I had in mind, but just enough to get an idea. -- .: Alex  :.  20:55, 20 June 2008 (UTC)
 * I think that takes away from GA, makes B look more important, maybe if we took the fourth column and just made Stub a darker red like #cc0000? Adam McCormick (talk) 21:01, 20 June 2008 (UTC)
 * It's a bit too deep. It looks as if someone's stabbed it half to death! :D -- .: Alex  :.  21:06, 20 June 2008 (UTC)
 * Ok, I'll moderate it, is #cc3333 preferred? Adam McCormick (talk) 21:19, 20 June 2008 (UTC)
 * How about #cc6666? And for the "Green B", #66cc66 for GA? Kevin Baastalk 22:44, 20 June 2008 (UTC)
 * Added another "green B" idea, from an swath used previously in GA-Class. Tito xd (?!? - cool stuff) 22:51, 20 June 2008 (UTC)
 * I prefer cc3333 to cc6666. I'd prefer 66cc66 as B and 66ff66 as GA. The sixth option could work. Maybe we should eliminate a few before we add any more? Adam McCormick (talk) 22:55, 20 June 2008 (UTC)
 * How about changing the "Redder Stub" from deep red to purple? See "Purple Stub" example. ThreeOfCups (talk) 23:23, 20 June 2008 (UTC)
 * OK, with deep red and purple I had to make the compromize (I know I started singing too, just wrong) Adam McCormick (talk) 23:39, 20 June 2008 (UTC)
 * Isn't that purple too close to Image-Class? Seems to be right between Image-Class and List-Class I like it. §hep   •   ¡Talk to me!  23:42, 20 June 2008 (UTC)

← I find that brown is dark enough a colour to make a good contrast with both Start-class's red and the background, but without being too dark for one to read the lettering. In addition, it is a rather good way to continue the colour scale, it is different from all other colours (including the non-article designations), and, in my mind, its relative drabness fits the designation of "the very lowest we can provide". Funnily enough, this tincture is the cross between the goldenrod and deep-purple proposals; the RGB values are the exact averages of these two colours. I had already decided on brown as my preferred option, but I was quite surprised by the result of this experimentation. Waltham, The Duke of 06:09, 21 June 2008 (UTC)
 * I've also crossed this brown with the "Redder stub" idea and created a more clay-like tincture, #c7503b. For those who don't like brown and prefer a redder shade.
 * Also, in order to save space, I propose removing the "Red start" and "Green B" options, which don't seem to have any support. Waltham, The Duke of 06:28, 21 June 2008 (UTC)
 * I'm still partial to the somewhat lighter red for stub class - i added the column "Redder Stub 2". Kevin Baastalk 15:10, 21 June 2008 (UTC)
 * We need to make sure that Stub-class offers sufficient contrast with C-class, so I think we should avoid warm browns or clay colors. Maybe a lighter version of Deep Purple? Or Redder Stub 2 with more blue and less yellow? (Okay, I admit I'm partial to purple, but it also seems to be the one color from the spectrum that's missing.) A light shade of plum- or burgundy-brown might work. ThreeOfCups (talk) 18:46, 21 June 2008 (UTC)
 * Why don't we aim to change as little as possible and just add a new color for C-Class? §hep   •   ¡Talk to me!  19:07, 21 June 2008 (UTC)
 * I don't usually have strong views on colours, but I'd suggest that it would be better to change the B-Class colour more than the Start-Class colour, as this would indicate the fact that B-Class standards are being raised by the introduction of C, whereas Start-Class is less affected. I don't like the idea of the new Start colour looking like the old Stub, as that sends the wrong message.  Also, bear in mind that the colour-scheme was designed to follow the order of the spectrum - unfortunately, until FA switches to purple, ThreeOfCups won't be too happy!  (I love purples as well, BTW)  However, I'll go along with whatever people decide here.  Thanks for your work! Walkerma (talk) 19:24, 21 June 2008 (UTC)
 * Oh yeah, that was something I forgot to add. I suggested a green B because it kinda represents the new standard of B-Class articles as higher quality articles well on the way to GA. -- .: Alex  :.  20:04, 21 June 2008 (UTC)
 * I think the GA-green is more or less fixed, but a greenish-yellow B-Class would seem like a good idea to me. Walkerma (talk) 20:10, 21 June 2008 (UTC)
 * I've added a new "yellow-green B" idea based on your comments. The new color is derived by measuring the halfpoint of the current B-Class and GA-Class colors in HSV space. The current B-Class color is then used for C-Class. Tito xd (?!? - cool stuff) 06:38, 22 June 2008 (UTC)

← If we keep all the other colors the same, but use a new color for the new C-Class, perhaps we could use gold #FFD700. That color falls between the C-Class shades used in columns 1 and 2. I added a column at the end to show how it would look. I like the idea of changing the B-Class color to signify the change in definition, but I'm not sure that a greenish yellow, like #ADFF2F, would provide enough contrast between B and GA. ThreeOfCups (talk) 03:03, 22 June 2008 (UTC)


 * ThreeOfCups, if you will see my proposal, you will notice that there is sufficient contrast between C-class and Stub-class, not only due to the latter being much darker, but mainly because it is has much blue, making it colder than other proposed tinctures. Kevin Baas's idea's principal difference from mine is that it has even more blue, and, frankly, I find it a bit "dusty". I have made a cross between Mr Baas's proposal and mine, resulting in #c76955; it might just be the ticket. Although I like purple as a colour, I find it unfit for this purpose, both because it is too dark and blue for the letters to be read and because purplish colours are already used for three non-article classes. Brown is used nowhere, and the new cross is also distinct from all other tinctures in use. (I'd make a new column for it, but one can simply picture the cross by looking at the two adjacent columns; besides, there are too many options right now. Could those which stand no chance be removed?)
 * Stepshep, the target is to have a colour scale that makes sense; changing only one colour is not necessarily simpler.
 * Walkerma, I find your view of the semantics involved very interesting, but I am not sure as to whether other editors will share your concerns. Are colours really that much associated in people's minds with classes? Especially the lower classes? Will it be so hard for them to get accustomed to the change? After all, Wikipedia is all about constant change.
 * As a final note, I think that colours that stand out too much from the rest of the scale, like the two golden ideas, should be avoided. We do not want flashy colours that will give undue emphasis to a certain class; a continuous spectrum of colours should be preferred, as Walkerma said. (A greenish-yellowish shade might be found for B, but in my little experience with colours these "cabbage" shades are usually rather ugly; on the other hand, the existing "#ffcc66 B" proposal is too light—close to white—and makes minimal contrast with A. As I see it, changing B-class is the option least likely to succeed. Well, after changing only C.) Waltham, The Duke of 04:20, 22 June 2008 (UTC)


 * As the table stands at this date, on my home CRT monitor the columns headed "#ffcc66", "Green B", "ffccff B", "Purple Stub", "Brown" all work. "Current" and "Gold C" the B and C colors are indistinguishable, "Goldenrod" the C and Start colors are indistinguishable, "Red start", "Redder stub", and "Redder stub 2" the start and stub colors are indistinguishable, "Deep purple" the stub color and text are blending too much. GRBerry 04:59, 22 June 2008 (UTC)


 * Looks like your monitor's color settings (temperature? contrast? brightness?) need some adjusting. Anyways, of the current selection, my top 5 choices are: Deep purple, redder stub 2, redder stub, Yellow-green B, & Green B. Kevin Baastalk 15:51, 22 June 2008 (UTC)
 * Yellow-green B works for me. It's a pleasant shade, the tone is consistent with the rest of the scale, and it provides sufficient contrast with both C and GA. (BTW, I'm sticking with my assertion that Brown and Redder Stub 2 don't provide enough contrast between Stub and C, at least on my screen.) ThreeOfCups (talk) 16:39, 22 June 2008 (UTC)
 * I agree with ThreeOfCup; Yellow green works for me (esp. given that B class is of higher quality) and that said contrasts seem insufficient. G.A.S 17:59, 22 June 2008 (UTC)
 * I'd support the Yellow-green B, that looks nice. §hep   •   ¡Talk to me!  18:05, 22 June 2008 (UTC)
 * Ditto. Yellow-green B works quite well without impinging upon other scales.- Running  On  Brains  18:25, 22 June 2008 (UTC)
 * I like both yellow-green B and brown stub. Either one would be fine with me, and I think both look significantly better than any of the other options currently suggested. Pyrospirit  ( talk  ·  contribs ) 20:11, 22 June 2008 (UTC)
 * I like the yellow-green as well. -- .: Alex  :.  20:46, 22 June 2008 (UTC)
 * Urm... Like which? You could be commenting on any of the messages above. :-)
 * Sorry, some of the text didn't appear when I typed it in for some reason. -- .: Alex  :.  10:21, 23 June 2008 (UTC)
 * I like the yellow-green shade quite a lot, and I find it fitting that it should be close (colour-wise) to GA in the same manner that A is to FA. I am only unsure that its contrast with the neighbouring colours is any better than that of the original (currently "current") C-class shade. (Still, I suppose, the lesser of two evils...) I still believe that my brown provides more contrast, in addition to its other benefits, but I find this last solution more beautiful. Waltham, The Duke of 01:29, 23 June 2008 (UTC)


 * So, does everyone like the green-yellow B? Tito xd (?!? - cool stuff) 06:16, 23 June 2008 (UTC)
 * It seems that way: Added by yourself, supported by Kevin Baas, ThreeOfCups, myself, Shep, RunningOnBrains, Pyrospirit, .:Alex:., Waltham. There are no objections. The other choices do not seem to have such strong support. G.A.S 10:53, 23 June 2008 (UTC)
 * Should we go ahead and make the change then? -- .: Alex  :.  11:28, 23 June 2008 (UTC)
 * I think so. Looks good. :) D.M.N. (talk) 11:57, 23 June 2008 (UTC)
 * I've requested the change to be done here. Done! -- .: Alex  :.  12:05, 23 June 2008 (UTC)

← Now that the color has been changed, can someone please adjust the corresponding images like accordingly ? The best would be to change directly this image, so that all links are automatically updated. Oh, and we will also need this kind of image for C-class. SyG (talk) 06:18, 24 June 2008 (UTC)
 * Those are already linked above ([[Image:Symbol b class (alternate).svg|15px]] [[Image:Symbol c class.svg|15px]]), although those could have their colors adjusted slightly. Tito xd (?!? - cool stuff) 06:23, 24 June 2008 (UTC)
 * Thanks for pointing those to me. Obviously the color is wrong, but at least that is a start. SyG (talk) 18:23, 24 June 2008 (UTC)
 * I apologise if I'm jumping the gun here (I see people saying the colour is wrong, but I'm not sure if they've already been changed) - I think the B-classicon above (the green one) is too close (exact, in fact) to the GA one. If this is something already being done then sorry for butting in. Nice work everyone with the change. Frickeg (talk) 07:31, 25 June 2008 (UTC)
 * The changes have been requested, and we are awaiting the new image colour. It will overwrite the current yellow B-Class icon, so all existing icons of this colour can be automatically updated. The green B-Class icon was just the closest example for the time, but will not be used. -- .: Alex  :.  15:38, 25 June 2008 (UTC)

Implementation of the new C-Class
OK, let's coordinate most of the implementation on this page. Perhaps people can sign up next to the tasks to be done, and strikeout the job when it's been done (completely)? Please add other necessary tasks to the list, I'm sure I've missed a few, but if we need to discuss anything in depth we probably need separate sections. Walkerma (talk) 02:57, 20 June 2008 (UTC)

To do

 * Define really precisely how the change will affect the definition of B and Start-Class. Should we adopt MILHIST's definition of B-Class, for example?  We'll need to debate the details.
 * Rewrite the definitions for B, C and Start. See the section below for details, since this is part of a bigger rewrite we're doing.
 * Develop a strong set of examples of B, C and Start, showing both the "centre of the range" as well as the border between ranges. I've been in touch with Holon for his help on this, and although he's on holiday he can help a bit. Sign up if you can spend some time doing comparative assessments of examples.
 * Inform CBM/Oleg of the change, so they can update the WP1.0 Bot. (Request here)
 * Done.
 * Update the assessment scheme template - see above.
 * Make sure that the automated creation tool described on Version 1.0 Editorial Team/Using the bot creates C-Class with the rest of the scales.
 * Document in step-by-step detail what steps WikiProjects have to take to use the new class.
 * Choose a color scheme (esp for B/C/Start) that works well - also see above. Perhaps the goldenrod color suggested by Julian above?
 * ✅ see color discussion above, templates have been updated to reflect the color change
 * Generate a global C-Class category, as well as a new C-Class category for each of the WikiProjects. The second part sounds like a job for a bot - but how do we take account of those projects who may wish to opt out of the scheme?  (I understand that MILHIST plans to debate the issue, and others may do also.) See the comments section just above for details.
 * BRFA is up. I'll have it check the deletion logs and not recreate anything that's been deleted; so the 'opt out' is to delete the category as soon as it's created. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 08:58, 20 June 2008 (UTC)
 * ✅ <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 17:10, 2 July 2008 (UTC)
 * Inform Pyrospirit and get the help-with-assessment script (like Outriggr's script) updated.
 * Pyrospirit done.
 * Inform ClockworkSoul of the change (now done), so that Igor can be updated.
 * Done.
 * Create a bot that automatically assess old B-class articles as new C-class, for all the articles within a given WikiProject. The choice to launch this bot would be let to the decision of each WikiProject. SyG (talk) 09:00, 20 June 2008 (UTC)

Rewrite process
The main rewrite is occurring at Version 1.0 Editorial Team/Assessment rewrite (talk), but brief comments may be left here.

For an overview of what was planned, see here, and the complete background discussion is here. The idea is that we can draft the wording here, and afterwards upload it into the main assessment page.

The basic idea is to have a summary page that gives a very simple overview of the scheme, then a more detailed page giving a very thorough definition for each level as well as a good selection of examples (middle of each level as well as borderline examples, also examples at each level from several different article types - e.g., biography, place, science etc. We can also include the timeline of development, found here.  Walkerma (talk) 03:07, 20 June 2008 (UTC)
 * Added a couple of things to do above. Tito xd (?!? - cool stuff) 05:08, 20 June 2008 (UTC)
 * We have WikiProject Council/Assessment FAQ as well. That page can have a lot of more detailed information, and can be used as a "subarticle" (for lack of a better term) of Grading scheme and this page. Tito xd (?!? - cool stuff) 05:10, 20 June 2008 (UTC)
 * I think it should be full steam ahead with the assessment rewrite. Where are we going to coordinate that? Wikipedia talk:Version 1.0 Editorial Team/Assessment rewrite?? <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 09:01, 20 June 2008 (UTC)
 * Yes, I think so, and we can do the rest of it (except the FAQ) here. I should be able to spend lots of time on it this coming weekend. Walkerma (talk) 13:08, 20 June 2008 (UTC)
 * I have completed a first draft of the "summary" here. I'd like to get some new examples inserted once we have had several people agree on these (I have been emailing an assessment macro to interested parties, please contact me if you can help.)  This is just a rough draft, please feel free to edit and polish it, but don't make it (much) longer! Walkerma (talk) 03:51, 25 June 2008 (UTC)


 * Request for comments:
 * 1) Should we reserve "Stub-Class" for stubs in the "classical" sense of the term?  The definition implies that a stub can be a long collection of (mostly) nonsense - but I've never seen such a thing myself (wouldn't you delete much of the nonsense?).  If it's longer but simply bad, isn't that Start-Class, especially now we have C-Class as well?  My choice (since the intro of C) would be to reserve Stub-Class just for very short articles, but we should agree on that.
 * 2) For B-Class, should we require the five criteria from WP:MILHIST as found here (also see FAQ)?  Or should we just recommend it?  Or should we require only certain of those criteria be met?  I think we finally have a chance to tighten up the meaning of B-Class, and personally I'd like to see the WP:MILHIST system adopted community-wide.  But is this realistic?
 * 3) For A-Class, I'd like to take advantage of this shake-up to tighten things up there too, with (a) inline references and (b) WikiProjects perform some level of peer review.  Initially I'd thought (b) couldn't be done at many smaller projects, but upon reflection I think it could be done more informally as follows.  Person A sees that a lot of new content and citations has been added by a newbie to History of Foo, and reckons that it is now A-Class.  Person A then asks Person B (maybe just on their user talk page) to see if he/she agrees it is A-Class.  Alternatively, if A wrote the new content, he/she asks two people (B and C) to assess the article.  This is weaker than a formal system as used at WP:MILHIST, but it would set a minimum standard that would be much more reliable than "Looks pretty good to me."

Comments please! Walkerma (talk) 04:29, 23 June 2008 (UTC)

Stub Class
Should we reserve "Stub-Class" for stubs in the "classical" sense of the term? The definition implies that a stub can be a long collection of (mostly) nonsense - but I've never seen such a thing myself (wouldn't you delete much of the nonsense?). If it's longer but simply bad, isn't that Start-Class, especially now we have C-Class as well? My choice (since the intro of C) would be to reserve Stub-Class just for very short articles, but we should agree on that.
 * Yes. G.A.S 06:04, 23 June 2008 (UTC)
 * We're talking 800,000 articles here. While WikiProject Stub sorting would like it, I'm not sure whether this change would be really disruptive to WikiProject scales. Tito xd (?!? - cool stuff) 06:20, 23 June 2008 (UTC)
 * I've recently formed a nice definition of stubs, which might actually be unoriginal, yet I've seen nowhere so far: "In terms of length and amount of information, a stub should be equal or inferior to the lead section of a featured article for the same subject." In other words, a definition and some further basic information, not organised, mostly or completely unreferenced; the definition relies on the importance of the subject. Now that I think of it, removing the "equal to" part might bring the definition closer to reality; stubs are not that organised or spherical in coverage and often have problems with NPOV and proof of notability. Waltham, The Duke of 07:10, 23 June 2008 (UTC)
 * Yes I agree with the proposal and Waltham's definition. I've always thought of stubs as... well the stub of an article that needs major expansion, and that is exactly what this designation was designed for. Small articles are Start-Class and C-Class represents the previous lower end of the B-Class scale. -- .: Alex  :.  11:41, 23 June 2008 (UTC)
 * The classical sense of "Stub" is an article that is so poor that an editor with no specialized knowledge of the subject could easily improve it, either by web searching or visiting a library. An article can be very long yet still be a stub; length has nothing to do with it.  GRBerry 13:59, 23 June 2008 (UTC)
 * Yes I would argue that a stub has to do with the "amount of meaningful content" not the length. But generally the two end up being synonymous except in deletion candidates. Adam McCormick (talk) 22:52, 23 June 2008 (UTC)
 * Yes. If an article is clearly beyond stub length, it should not be a stub. Now that we have C-class as an option, Start-class can be used for longer articles with poor content. Pyrospirit  ( talk  ·  contribs ) 03:09, 24 June 2008 (UTC)

OK, thanks! We can try and incorporate His Grace the Duke's suggestion into the description as at least a guideline. Walkerma (talk) 01:43, 25 June 2008 (UTC)
 * Excellent! - Waltham's description is short and snappy while being very clear and precise. Further, it encourages editors to improve the article however they can, not just by "hoop-jumping" to tick of specific points. I hope similarly simple definitions can be found for the other categories! Martin  (Smith609 – Talk)  08:51, 26 June 2008 (UTC)

The general view seems to be that we can incorporate Waltham's description as a guideline. Although a few people believe that a stub can be long, this seems to be a minority view, and so I don't think the concept of a "long stub" should be explicitly used in the definition of Stub-Class. (My own view (Walkerma) - we almost never see a long article that only has a sentence or two of real content, because the "nonsense" portion would normally be given short shrift and removed. So I think it is mostly moot anyway.) Walkerma (talk) 15:29, 2 July 2008 (UTC)
 * Conclusion


 * Comments on Stub-Class
 * For the purpose of "stub class", I would exclude infoboxes, short lists, the mere fact that there may be sections from consideration. Refer to the following stubs for my reasoning: Yukimasa Obi, Brutish Mine, and Polpo (manga). G.A.S 05:03, 3 July 2008 (UTC)
 * Yes, good point. We can include one or two of your examples in the definition page, to drive the point home. Walkerma (talk) 14:09, 3 July 2008 (UTC)

B-Class
For B-Class, should we require the five criteria from WP:MILHIST as found here (also see FAQ)? Or should we just recommend it? Or should we require only certain of those criteria be met? I think we finally have a chance to tighten up the meaning of B-Class, and personally I'd like to see the WP:MILHIST system adopted community-wide. But is this realistic?
 * Yes. Also note that WP:MILHISTS grading template requires a "yes" for each of the criteria, otherwise it is not rated as "B-Class". G.A.S 06:04, 23 June 2008 (UTC)
 * We should be careful with having the GA and B class criteria too close to one another (Although it seems that Milhist's criteria leaves a safe margin). G.A.S 06:02, 24 June 2008 (UTC)
 * Yes. The Milhist system is an easy way to identify article strengths and weaknesses. -- R OGER D AVIES  talk 06:13, 23 June 2008 (UTC)
 * My only concern is that small WikiProjects may not be able to implement this, but otherwise, yeah. Tito xd (?!? - cool stuff) 06:18, 23 June 2008 (UTC)
 * This has come up a couple of times now and I don't understand it. Can you explain to me please why they can't? -- R OGER D AVIES  talk 06:20, 23 June 2008 (UTC)
 * Smaller WikiProjects have smaller amounts of users able to do assessments. While theoretically there is nothing stopping these projects from doing checklist assessments, there needs to be a way to do this with a minimal amount of bureaucracy possible—these projects may not have an assessment department (or even an assessment page) written, and operate mostly informally. Although thinking about it, maybe providing a robust checklist may help these projects, not hinder them. Hmm... Tito xd (?!? - cool stuff) 06:25, 23 June 2008 (UTC)
 * Thanks! From experience with various Tag & Assess things, checklists are very easy for editors, especially if there's a FAQ.-- R OGER D AVIES  talk 06:28, 23 June 2008 (UTC)
 * Is there an available FAQ? Tito xd (?!? - cool stuff) 06:31, 23 June 2008 (UTC)
 * Yep, -- R OGER D AVIES  talk 11:10, 23 June 2008 (UTC)
 * The project I am most active in (WP:Anime) does not have an "assessment department". As such, most assessments are informal, usually by the applicable editors, or a passer-by. This is why I reckon that a checklist would by most helpful. G.A.S 06:51, 23 June 2008 (UTC)
 * Yes – With a good checklist, any Wikipedian experienced and knowledgeable enough in certain areas (referencing, mainly; the rest is easier for most people) can make assessments, unrestricted by specific projects. I think this process is meant to be the projects' counterpart of GA: quick and informal. The more formal processes should be reserved for the next level, namely A-class and Featured Articles. Waltham, The Duke of 07:10, 23 June 2008 (UTC)
 * Yes, please (note that the Germany project also uses a variant of this). The checklist system tells an editor whose article has been assessed in what areas the article needs to be improved. This is much more useful than a mere "you article needs to be improved" that a simple numerical assessment without clear criteria tells the editor. Kusma (talk) 08:57, 23 June 2008 (UTC)
 * Recommend The key standards should be the same, the fine details could vary. I see B2 and B1 as the key standards.  B5 could be better reworded as "where appropriate it contains..."; not all articles will merit any of the examples.  (E.g. Filioque where it would be inappropriate to have an infobox or diagram, and any images on this topic will be merely decorative - so if they weren't freely licensed we couldn't use any images.)  B3 will normally just happen, it doesn't really need to be in the checklist.  B4 in my eyes is subsidiary to B2; a major grammatical error will distort the content of the article. This is particularly an area where projects should be free to have additional standards.  The Saints wikiproject might reasonably insist on having coverage of the beatification process for modern Catholic saints or on mentioning the standard life of the saint for older saints.  GRBerry 14:49, 23 June 2008 (UTC)
 * This makes sense in principle, but the actual list needs thought. In particular, "all major points have appropriate inline citations" is way too strong a citation requirement for B-Class. Even GAs only require inlines for "direct quotations, statistics, published opinion, counter-intuitive or controversial statements that are challenged or likely to be challenged, and contentious material relating to living persons." I propose that the B-Class requirement should be something like "It is suitably referenced and has reliable sources". Geometry guy 15:10, 23 June 2008 (UTC)
 * PS. Can we please use bullets, not numbers, for a discussion? There's been enough !voting in recent months! :-)
 * Changed. I just started with numbers as normal indent would have been hideous to follow. G.A.S 16:10, 23 June 2008 (UTC)
 * Support inclusion of five points, although I think it would be better to make a few wording changes to allow a little flexibility.
 * "the article should meet the following criteria..."
 * B1 should be changed to something like "Contains reliable references for all major points, as well as inline citations where appropriate." There has been talk in past FAC's about how inline citations are over-required and can actually detract from the quality of an article (in some rare situations), so I don't think they should be required.
 * I echo the call GRBerry made for a change in B5 from "contains appropriate supporting materials" to "contains supporting materials, where applicable (or appropriate or some other qualifier)"
 * With these changes, it would be a great idea.- Running On  Brains  16:21, 23 June 2008 (UTC)
 * Support/Yes Very reasonable criteria if the FAQ was heavily generalized and a generic "checklist" template created Adam McCormick (talk) 21:26, 23 June 2008 (UTC)
 * Recommend use of the MILHIST criteria for B-class. B-class should remain a less formal designation than A-class or GA, so I do not think we should require use of the MILHIST criteria. However, strongly recommending their use in judging whether an article is B-class would strengthen the definition of B-class and is quite a good idea. Pyrospirit  ( talk  ·  contribs ) 03:16, 24 June 2008 (UTC)
 * Agreed. G.A.S 06:02, 24 June 2008 (UTC)
 * For those interested: WikiProject Tropical cyclones has just adopted a modified scale, based on the MILHIST checklist, the comments here, and our own needs (see B6 in the link above). The banner for the WikiProject has also been changed to automatically assess as C-Class those articles that are marked as B-Class, but that are not marked as passing the criteria above. (The relevant template magic is at Hurricane/Temp, Hurricane/B check and Hurricane/B list.) What do you guys think about the wording there? Tito xd (?!? - cool stuff) 06:09, 24 June 2008 (UTC)
 * (edit conflict) It seems to make sense. I believe we should agree on each of the criteria individually, as I believe we will only go around in circles if we try to decide on them as a "package". I like the way in which the shortcomings (as opposed to GA class) is clearly described. G.A.S 06:15, 24 June 2008 (UTC)
 * This additional criterion B6 may need to be rephrased, as it is based on a false premise: "Wikipedia is meant to be a general-purpose encyclopedia". This is a significant distortion of Pillar One (!) which states "Wikipedia is an encyclopedia incorporating elements of general encyclopedias, specialized encyclopedias, and almanacs". If Pillar One is not definitive enough, perhaps the well known Jimbo Wales quote ("Imagine a world in which every single person on the planet is given free access to the sum of all human knowledge. That's what we're doing") brings this issue into focus.
 * Should articles be as widely accessible as possible? Yes. Must they all be entirely accessible to the general public? No. Geometry guy 21:01, 24 June 2008 (UTC)
 * Chalk that one up to poor wording. The point intended is for articles to not be mired with excessive technicisms where avoidable, and for articles to be written with a viewership greater than the meteorological community in mind. In short, for them to not be tagged with technical. Any suggestions for alternate wordings? Tito xd (?!? - cool stuff) 21:11, 24 June 2008 (UTC)
 * Something like "The article presents its content in an appropriately accessible way", with a corresponding explanation. The gold standard here is "The article presents its content in as widely accessible way as possible", but we shouldn't be asking for that at the B level. Geometry guy 22:24, 24 June 2008 (UTC)

It seems clear that there is a consensus: Some version of the MILHIST criteria should be adopted. We can tweak the language as needed over the next couple of days to allow a little more flexibility. I think G-Guy's final wording (right above) for criterion B6 sounds very good. Walkerma (talk) 15:11, 2 July 2008 (UTC)
 * B-Class Conclusion


 * B-Class Comments

A-Class
For A-Class, I'd like to take advantage of this shake-up to tighten things up there too, with (a) inline references and (b) WikiProjects perform some level of peer review. Initially I'd thought (b) couldn't be done at many smaller projects, but upon reflection I think it could be done more informally as follows. Person A sees that a lot of new content and citations has been added by a newbie to History of Foo, and reckons that it is now A-Class. Person A then asks Person B (maybe just on their user talk page) to see if he/she agrees it is A-Class. Alternatively, if A wrote the new content, he/she asks two people (B and C) to assess the article. This is weaker than a formal system as used at WP:MILHIST, but it would set a minimum standard that would be much more reliable than "Looks pretty good to me."
 * Moved below. Support; although I would reckon that person A should not ask person B "is it A class?", but "should we consider listing the article at WP:FAC?". G.A.S 05:36, 25 June 2008 (UTC)
 * Comment: I believe that we should first figure out the whole A class vs. GA class issue: "Is the A class grading higher than GA class?" G.A.S 06:04, 23 June 2008 (UTC)
 * Refer to this recent example re the current confusion regarding the status of A and GA class:
 * Black hole was listed as an A class article.
 * This was assessment was given after the FAR (diff), and later reconfirmed on 2007-01-10 (diff).
 * On 2007-04-27 the article was listed as a candidate for "Wikipedia 1.0", confirming its status as A class. (diff).
 * On 2006-12-27 the article was submitted as a good article nominee, but failed. (Mainly due to problems with citations - it is unclear to me why it was submitted to GA nomination, as A class is technically higher than GA class (WP:ASSESS—"Good articles that may succeed in FAC should be considered A-Class articles, but having completed the Good article designation process is not a requirement for A-Class.").)
 * What should the assessment for the article be then? There was no formal review to classify it as A class, and it failed GA class.
 * (The article was subsequently re-assessed as B class on 2008-06-09 by myself (although it seems that this would go down to C class soon if the B class criteria is passed).)
 * G.A.S 05:39, 24 June 2008 (UTC)
 * Comment How about something like "the proposal to promote to A-Class should be made on the article's talk page and supported by two (or three?) uninvolved editors, with no significant opposes"? This is less formal than Milhist but more formal than editor's talk pages. -- R OGER D AVIES  talk 06:11, 23 June 2008 (UTC)
 * Wait until we sort out that whole GA v. A mess. Tito xd (?!? - cool stuff) 06:19, 23 June 2008 (UTC)
 * I think the mess has been sorted out; the proposal to move A-class down has not garnered enough support, and I can see why: a good, MILHIST-style A-class review is better than a GA review. Applying even a less stringent version thereof project-wide will settle this issue for good. Personally, I see two pairs: B-class–GA (one person) and A-class–FA (committee). The difference is that the community-wide process in each pair has more legitimacy and a higher status and level of difficulty and professionalism. I like viewing the grading scale as a two-tiered system, with three levels for developing articles and four levels of increasingly formal evaluation honing an already useful article into perfection. Waltham, The Duke of 07:10, 23 June 2008 (UTC)
 * Comment – GA standards on referencing are strict enough; I'd say that, from GA-class upwards, improvement concerns almost exclusively presentation and completeness. Therefore, there should be no sourcing gaps for an article to be A-class. As far as the process is concerned, I think I prefer Mr Davies's version; more formal than one-person reviewing, but certainly more flexible and informal than FAC. Walkerma's proposal is too weak, in my opinion, especially considering that such a system is easier to be rendered meaningless by editors working on an article calling on a friend to have it safely tagged as A-class. I also suggest—and this might, at least partly, solve this problem—having a checklist for this process as well, to help the reviewers; it could be the community-wide processes' counterpart of "criteria". In this case, of course, the checklist would not be part of the banner, but would be given in, or linked from, the reviewing thread. Waltham, The Duke of 07:10, 23 June 2008 (UTC)


 * Comment - I know at WP:SCH/A we require unanimous support of our dedicated assessors to rate an article and A. Also I am of the opinion that A is just fine where it is, it's a great way for a project to say "this GA article is close to FA standards" and to mark it as such. Adam McCormick (talk) 21:30, 23 June 2008 (UTC)


 * Comment: I would have brought this up earlier had I been active in this debate, and I'm not sure this is the place for it, but I would support merging GA-class into A-class, with the community consensus aspect of both remaining in place. As others have pointed out, both classes contain near-"complete" articles, and it is not obvious at all (especially to newcomers) which is considered better.  I see no reason why we feel the need to abandon our letter scheme for another bureaucratic classification, especially when Start/C/B/A/FA would be so much simpler. &mdash;Pie4all88 (talk) 22:44, 23 June 2008 (UTC)

The suggestion below - to split different types of assessment - did not give a clear consensus, and so we can not proceed rapidly at this point. We can perhaps revisit the issue once the technical aspects have been clarified, in a week or two. We should not hold up the rollout of C-Class while this debate rages on. I have split that section from the A-Class main discussion, as it is a much wider issue than just A.
 * A-Class basic conclusion

There does seem to be a consensus for beefing up the WikiProject-based peer review element. I would like to propose that we adopt Roger Davies description, "the proposal to promote to A-Class should be made on the article's talk page and supported by two (or three?) uninvolved editors, with no significant opposes." We could also recommend (not require) that larger projects set up a formal WikiProject-based peer review scheme like WP:MILHIST. It seems that there is NOT a consensus to require inline references. Walkerma (talk) 15:23, 2 July 2008 (UTC)


 * Comments on A-Class

You probably want to say something like "For WikiProjects without an A-Class peer review scheme, the proposal to promote to A-Class ..." right? Geometry guy 18:33, 2 July 2008 (UTC)

Suggestion—Split community and Wikiproject assessments

 * Suggestion - In my opinion, FA and GA should be separate from the assessment scale, becoming instead project-wide designations of quality for articles. The assessment scale would then be Stub/Start/C/B/A, turning A-class into the highest rating WikiProjects could give an article; almost all featured articles would be A-class, and A-class articles would usually be close to FA if not already there. I think this would make the assessment scale much simpler and more intuitive. Projects would give each article one of five possible ratings, and FA and GA would be like "stamps of approval" for a certain level of quality. Doing so would also make the definition of A-class much clearer. (I think this might have been proposed earlier; if so, we should renew discussion on it.) Pyrospirit  ( talk  ·  contribs ) 03:27, 24 June 2008 (UTC)
 * Comment: If the wikiproject does not consider an article "A-class"; it should not" be featured. The only use I see for A class, would be if only an A class article (where the Wikiprojects has an assessment department) or'' a GA class article (where the Wikiproject does not have an assessment department) may be presented as an FAC. G.A.S 05:39, 24 June 2008 (UTC)
 * I like this idea alot Pyro. Seems very logical to me, and would clear up alot of confusion.  Also, for the first time, the C>B>A progression will make sense.  This idea also works because since GA and FA are peer-reviewed qualifications, they should be separate from the normal "do-it-yourself" (for lack of a better term) assessment scale. Drewcifer (talk) 05:45, 24 June 2008 (UTC)
 * I disagree: would Wikiprojects be willing to give up the ability to have an article rated FA class? GA class? Should this be a dual rating (which would duplicate the article in statistics)? Currently Wikiprojects pride themselves on the FA class ratings and GA class ratings bestowed upon articles within their scope. Why would a wikiproject want to "overwrite" a community rating (eg. GA to A class)? In the same breath: why would the wikiproject not list an article for FA when it is at that "A class"? The A class rating = "ready for promotion"; but so is "FAC". <Confused/> G.A.S 06:09, 24 June 2008 (UTC)
 * I think we need to devote some time to discussing this separation proposal fairly soon. This fuzziness at the top end of the assessment scale keeps cropping up every time we look at A-class, and until the A/GA issue is settled it will be difficult to form a proper idea of which criteria we should be basing A-class on, and where it then comes in the hierarchy. EyeSerene talk 09:35, 24 June 2008 (UTC)
 * I second the proposal — in a separate discussion. G.A.S 09:53, 24 June 2008 (UTC)
 * G.A.S, I think you might be misunderstanding what I suggested. WikiProjects would still be able to have FA's and GA's under their scope, and they would not "overwrite" community ratings at all. All that I suggest is that FA and GA be made separate from the assessment scale. The assessment scale is designed for use by WP1.0 and individual WikiProjects to assess articles; FA and GA, on the other hand, are community ratings given by a project-wide process. A WikiProject cannot assess an article as FA or GA without that process. This means that as things currently stand, most articles that have passed FA or GA effectively have no WikiProject assessment, since their WikiProjects just duplicate the community rating. With the change I suggest, WikiProjects would rate even FA's and GA's on the stub/start/C/B/A scale, allowing their assessments to continue to be of use. So, if a good article's WikiProject assessment rating goes from B-class to A-class, then it's a good candidate for featured status; likewise, if an A-class featured article drops to B-class, it's a sign that it probably needs to go to FAR. As for the issue of duplication in the statistics, the numbers for FA and GA would be moved elsewhere as they would no longer be steps in the assessment scale. Think of it this way: articles that pass FAC (or GAN) would gain "featured (or good) status", but they would still keep their WikiProject assessment ratings. Does this make more sense? Pyrospirit  ( talk  ·  contribs ) 15:33, 24 June 2008 (UTC)
 * That clarifies it quite nicely. I am not sure whether the Wikiprojects would see it this way (For said explanation above). For the most part, these ratings are so integrated into the assessment scheme that I fail to see use in removing them from it. Consider also: A Wikiproject has 10 "A class" articles; of which 9 are featured. Without cross referencing two categories, it would be impossible to see which one has not been featured yet. We are back at the very reason why C class was introduced. G.A.S 15:46, 24 June 2008 (UTC)
 * C-Class was introduced because the gap between B-Class and start-class was too wide. It was added to deal with an issue on the lower end of the scale. Separating FA and GA from WikiProject assessments would not affect C-Class articles, and the existence of C-Class does not resolve the confusion over A-Class and GA. The use in removing FA and GA from the assessment scale is that it would resolve the issue of confusing definitions at the higher end of the scale. I agree that it could occasionally require cross-referencing categories (though the categories in question would usually be fairly small), but making the assessment scale clearer and simpler will be more beneficial in the long run. Pyrospirit  ( talk  ·  contribs ) 21:53, 24 June 2008 (UTC)
 * What should be done where Wikiprojects does not use A class at the moment, because they do not have an assessment department for instance? I would rather that they keep to the current schema. This may work, but someone has to play devil's advocate. G.A.S 05:23, 25 June 2008 (UTC)
 * I believe that it should be possible to amend the bot so that WikiProjects can see in their worklists which of their A-Class articles are FAs or GAs. But would that keep people happy?  Would we need to amend the statistics table as well (which currently provides a nice summary of no. of GAs and FAs for many projects)? Walkerma (talk) 14:50, 26 June 2008 (UTC)
 * Lets continue the discussion re table layout below. These threads will start to unravel otherwise:) I believe that it would be valuable to cross reference the "status" of articles with their "assessments"—see version 5 below. G.A.S 15:21, 26 June 2008 (UTC)


 * Support Oppose to inline citations. As long as there is a proper review from the WikiProject, if the WikiProject judge the article does not need an inline citation on a specific paragraph because the offline citations are good enough, who are we to boss them ? Sorry to say that to avid internet users, but the documents available online are just miles away from the quality and abundance of the documents available offline. SyG (talk) 06:24, 24 June 2008 (UTC)
 * I think you've got inline citations and online citations slightly mixed-up SyG... tricky letter. We just mean that the article has to show proper referencing. The sources don't have to be from the internet. Best, --Jh12 (talk) 06:53, 24 June 2008 (UTC)
 * Oh yes, my bad! Thanks for correcting me. Then obviously I support this proposal. SyG (talk) 18:21, 24 June 2008 (UTC)

So, to clarify visually (something that might help conceptualize the proposal), would the proposed new system be something like this?

Or this? Drewcifer (talk) 18:54, 24 June 2008 (UTC)


 * The second one matches the proposal more closely, I think. Note that a GA that is a strong candidate for FA might be A-Class, and an FA that needs its status reviewed might be B-Class. Whether an article is A-Class or B-Class would be up to the WikiProject. Thanks for the illustration, by the way. Pyrospirit  ( talk  ·  contribs ) 20:09, 24 June 2008 (UTC)
 * I like the first one better. B-class has a well-established use as meaning "almost good enough for a GA review".  A-class, on the other hand, is poorly defined in all but the few wikiprojects that have build an assessment scale around it.  --Arctic Gnome (talk • contribs) 21:10, 24 June 2008 (UTC)
 * Either one would probably be fine. It might turn out under this proposal that most GA's would be A-Class, or it might be evenly split between A and B, or maybe mostly B. The important thing is to separate FA and GA from the WikiProject assessment scale. Pyrospirit  ( talk  ·  contribs ) 21:19, 24 June 2008 (UTC)
 * I agree. I have long argued for separating GA from the WikiProject scale, and have been convinced that FA should be treated similarly. Here is a visualization which combines the two previous ones. Geometry guy 22:05, 24 June 2008 (UTC)


 * Wow, this is looking pretty good so far, guys! I have to admit, there was a time that I thought that the A/GA situation would be more or less intractable. Excellent ideas all around. Girolamo Savonarola (talk) 05:59, 25 June 2008 (UTC)

Support Splitting the assessment scale from the Peer Review scale is a stroke of genius (literally thinking out of the box). Having two interleaved conceptual scales with different processes was driving me crazy. That an A class article could be listed as better than a GA class when it hasn't gone through a community review was my sticking point. Granted, an A article may actually be better than a GA reviewed article, depending on the people/projects/check lists/criteria involved. This way the stub-start-C-B-A is an assessment process within a WikiProject (and fixes the gap between start and B), and the GA-FA are community peer review processes with the Wiki seal of approval. BTW, I added captions to the three wikitables to differentiate them. I like version 1 better. — Becksguy (talk) 08:35, 25 June 2008 (UTC)


 * Yes, I really like these tables as well, and in fact I was influenced by them when drafting the order for the new "short" summary table. I personally like Version 2, because I see it as meaning, "B is aspiring to be GA, and A is aspiring to be FA"; also, you don't have anything appearing twice.  However, I don't have strong feelings on this.  I think the idea of headers is great, it really clarifies WHY there are two columns. I think "Peer-review" is the wrong phrase for Col 2 (we're proposing an internal WikiProject peer review for A-Class), and I'd avoid "External review" because that may be interpreted as involving people external to WP (this is one option we're considering for the future).  I think "Community review" (Becksguy's own term) may be more appropriate.  Or is there something better still? Walkerma (talk) 14:12, 25 June 2008 (UTC)
 * Support proposal and agreed with Walkerma. I too prefer v2 for the same reasons and I don't see why A-Class needs to appear twice (I found this confusing when I first saw it). I kinda agree with changing the name, although I don't think it's a huge issue. Wikipedia's internal name for these processes is Editorial Validation, so how about Editorial review? -- .: Alex  :.  15:29, 25 June 2008 (UTC)
 * There is no significance in A-Class appearing twice. It simply reveals a limited skill with wikitable mark-up. I've demonstrated using rowspan=2 to eliminate the duplication of GA-Class in V3; something similar could probably be done with A-Class. Geometry guy 18:17, 25 June 2008 (UTC)
 * Neutral Support V2, propose V4 . But if this is the case, I fail to see the reason for having them in parallel. (B aspiring to become GA → B<GA and A aspiring to become FA → A<FA thus B→GA→A→FA. We could use a second column to specify who provides the assessment, but keep the ratings in series. G.A.S 16:59, 25 June 2008 (UTC)
 * This is a weakness of V2: a GA could be B-Class or A-Class. But I'm not sure why we are focusing on the way the tables look rather than the conceptual distinction between WikiProject ratings and GA/FA. V4 below reintroduces some of the current confusion: the missing equation is GA<A, which is what caused the whole problem to begin with. Geometry guy 18:17, 25 June 2008 (UTC)


 * Accepted. G.A.S 20:08, 25 June 2008 (UTC)

I think the problem we're having is re-conceptualizing the process. At the crux of the proposal (at least how I interpret it) is that an article is both FA AND A simultaneously. So in other words, FA is a subset of all A articles, and these articles have gone through an additional editorial-review to gain the additional status. In yet other words, FA does not supersede A. And the exact same things would hold true for GA's relation to A-class or B-class, depending on which proposal version your looking at. So therefore I think it's a little iffy to simplify things by saying FA>A, since it implies a linear movement from A → FA. This is why I'm not too keen on the single-column one above (v4), since it implies a linear hierarchy of assessment. Drewcifer (talk) 18:38, 25 June 2008 (UTC)


 * What Drewcifer said. The problem is using a linear scale in the graphic, since the process isn't linear. Although I really like the "By" column, as that clarifies wonderfully. — Becksguy (talk) 19:25, 25 June 2008 (UTC)


 * Maybe we should go for a sortable table instead;) I appreciate your liking the By column:) G.A.S 20:08, 25 June 2008 (UTC)


 * Yeah, the linear scale really doesn't properly convey that FA and GA would be separated from the scale. Version 3 seems to represent it best (though an FA might drop to B-Class shortly before going to WP:FAR). Anyway, we shouldn't get too hung up on the appearance of the representation; it's the proposal itself that matters. Pyrospirit  ( talk  ·  contribs ) 20:03, 25 June 2008 (UTC)

I strongly support this separation proposal. It makes a huge amount of sense, and think version 3 is closest to how I had envisioned things working, although (no fault of Geometry guy's - no easy way to show this in a table!) it still implies an equivalence between the WikiProject assessments and the community ones. However, as Pyrospirit says, it's the principle that matters ;) EyeSerene talk 10:00, 26 June 2008 (UTC)


 * Strong support versions 3 and 4, as amended by Drewcifer. (A B-class FA should probably be nom-ed for FARC. Sometimes they'll improve a lot without being downgraded.)--HereToHelp (talk to me) 12:04, 26 June 2008 (UTC)


 * Support Version 3, version 4 is better than current but the de-coupling of the peer review & the project review is a good idea as it stops a class from being almost obsolete as it is now. --Nate1481(t/c) 12:44, 26 June 2008 (UTC)


 * Slight Oppose to Version 4. Most notably on the GA-Section. The problem with community-review of GA-Class Articles is that it only requires a single editor, and usually deals significantly with content and verifiability.  As of late, we at the GA-Wikiproject have experienced a number of difficulties with the background knowledge of subjects by the reviewer.  as an example, during the review of Sources of Islamic law, the first reviewer's opening line in the subpage was "I am by no means an expert in any sort of law, let alone Islamic law, but I think that will give me a great "outsider" perspective", which puts the review of to....a very bad start.  I'm of the mindset that wikiprojects should coordinate the GA-Review process within their category on the WP:GAN subpage.  We've had an in-depth and lengthy discussion concerning the specializing of reviewers Here.  I think if specific wikiprojects were able to coordinate their GA-Reviews better, we might be able to alleviate this problem.  Respectfully, Cam (Chat) 01:23, 27 June 2008 (UTC)

I thought about this: Would it really be necessary to keep FA and GA as assessments (for this purpose)? Adjust A-class's criteria and description to that of FA/FL, and A class becomes valuable. B-class's criteria is also being adjusted upwards, and as such will also become valuable as an assessment. (I.e. do NOT call FA and FL "assessments" but rather "status") Or is this what everyone has been saying all along? G.A.S 14:40, 26 June 2008 (UTC)
 * More or less, yeah, that's the idea. Now how soon can we implement this?--HereToHelp (talk to me) 15:12, 26 June 2008 (UTC)
 * Yeah, FA and GA being a status rather than an assessment is pretty much what this has all been about. We should implement this as soon as we have a consensus on the matter, whenever that would be. There seems to be a lot of support and not much opposition to this proposal, so maybe we should just let discussion go on for a few more days to give people time to voice their opinion. After that—assuming people continue to support it—we could probably safely implement the proposal. Pyrospirit  ( talk  ·  contribs ) 15:45, 26 June 2008 (UTC)
 * I believe I was the only one with doubts regarding this. Although in essence, these would be presented in separate tables in future:). Maybe we should design the assessment statistics template for the Wikiprojects in the meanwhile? G.A.S 15:53, 26 June 2008 (UTC)


 * Is it wise to just implement it, or should we go through one of those ratification vote things? There does seem to be a good amount of support, but I don't think we should be hasty.  This isn't as massive of a change as it seems, just a slight rethinking of a process which is poorly defined, but I don't feel like we're getting the whole community's consensus here.  Besides, making a formal proposal to the community will help us further strengthen the idea until it is ready. Drewcifer (talk) 15:55, 26 June 2008 (UTC)


 * These alternatives certainly clarify the fact that it's not a linear sequence Where exactly would you see these tables being used? In the stats table for each project?  If so, can you show what (say) Version 5 would look in situ?  If there is a clear consensus on the 1.0 talk pages (which there is here), we have usually just gone ahead with minor changes like this.  Likewise with changes of colors.   Incidentally, the French Wikipedia is also considering removing FA/GA from their scheme, and adding an extra class - see their draft grid here. Walkerma (talk) 16:11, 26 June 2008 (UTC)
 * I don't think another ratification vote is a good idea. As Walkerma said, there is a clear consensus here, and if there continues to be a consensus, we should just implement it. A ratification vote also has its own problems associated with it. For example, in the ratification vote over C-Class, many reasons in support were extremely brief or absent and of little use, and many reasons given in opposition seemed to miss or ignore points that had already been discussed (such as stating that there was no need when that need had already been demonstrated). This is why discussion is preferable to polling: because in a vote or poll, people simply state which arguments previously brought up that they support, while in a discussion, people regularly introduce new ideas. I think that discussing this for a few more days will allow for sufficient community input to determine consensus. Perhaps we should add this discussion to cent as well, as long as it's clear that this is a discussion, not a poll or vote. Pyrospirit  ( talk  ·  contribs ) 16:51, 26 June 2008 (UTC)


 * (ec) I would not consider splitting off the GA and FA process would be a "minor" change. I for one think this needs much broader input. I have no specific objections but this affects how every WikiProject runs assessments and will cause at least as much havoc as did the addition of a C class. This should at least be posted to cent if not a site-wide notice. Adam McCormick (talk) 16:59, 26 June 2008 (UTC)


 * Sorry, I was unclear. A change in layout of how FA/GA are presented in tables can be done tomorrow if needed.  The things shown above simply represent the current reality more clearly than existing "linear" representations.  But removing FA-Class/GA-Class from the system altogether would definitely need a much wider discussion or poll - there would be some clear opposition - but I think that at present we're only talking about formatting changes, aren't we? Walkerma (talk) 17:46, 26 June 2008 (UTC)

Coming in to this discussion from a break, I'm most concerned about technical implementation. There seems to be support for redefining FAs as a subset of A-Class articles, and GAs as a subset of A/B-Class articles. How is that going to be implemented in practice, by the 1.0 bot, by the category tree from Category:Articles by quality, and by the tables that are produced for the wikiprojects. Making this change, more so than C-Class, will require modifying hundreds of templates, thousands of categories, and/or millions of pages. What is the tangible benefit that this redefinition provides? What problem is it solving, and why is that problem significant? And is there a way of solving that problem with a minimum of disruption? <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 18:53, 26 June 2008 (UTC)

(←) I believe that at least Pyrospirit, Drewcifer, EyeSerene, Becksguy and myself are talking about conceptual change as well as formatting changes. We are putting forward the case that FA and GA are a different kind of designation than Stub-Start-C-B-A, so that FA's and GA's would also have WikiProject ratings on the Stub-Start-C-B-A scale (almost invariably B-Class or A-Class for GA's, and A-Class for FA's). It is heartening to see the French taking forward a similar idea.

This does require a wider discussion, but I too would prefer a community discussion to a poll. Several recent polls have convinced me of the wisdom of the traditional wiki lore that polls generate division much more easily than they help to build consensus. This is particularly true of polls where support and oppose !votes are in separate sections and numbered. At the very least, all points of view should be discussed together to encourage editors to read as well as write.

There will be opposition, and it may be better to proceed first with formatting changes to better reflect the current reality, and return to conceptual change a little later. On the other hand the community/WikiProject assessment split idea, which has been kicking around for over a year now, does seem to have finally captured the Zeitgeist and maybe it is time to roll with it. Geometry guy 18:54, 26 June 2008 (UTC)


 * It may be best to introduce formatting changes first, as part of the many changes coming in with the new C-Class. If we can get the bot to start recording the information, then better still.  (That would be helpful even now, for spotting if an article is listed as B-Class but is a GA, for example.) Then we can debate the more serious change of removing GA/FA from the system, and I think by then (if the bot can do the trick) then there should be a strong case to be made.  When I proposed removing GA (only) from the system a while back, opinion was split down the middle, but the main objection was "we want to know if one of our articles is a GA".  If we remove that objection before putting it to the community, there should be little objection IMHO.
 * Happy-melon: re the technical issues, I'm technically very poor, but the 1.0 bot is able to report (in the project worklists) whether or not an article is in the V0.5 release. It will do the same for V0.7 when it comes out (hopefully this fall).  It does this by checking for the talk page category representing V0.5, I believe.  I think it should equally well be able to check for GA or FA.  GA is even easier, since GAs are (like V0.5) recorded by itself.  But I'll ask Oleg and CBM about the practicalities for the bot.  Any migration to such a new system should easily be doable by bot, I think, and any reassessments shouldn't be too onerous because there aren't so many articles in these high levels.  Walkerma (talk) 19:52, 26 June 2008 (UTC)
 * I agree with this approach, particularly in view of the technical concerns raised by Happy-melon: this needs to be done in stages to minimize disruption. In reply to Happy, I would say the problem is a serious one: there is conceptual confusion between WikiProject ratings and community processes, which causes headaches and extra work for editors when either type of assessment changes. (Example: if I delist a GA, do I downgrade WikiProject ratings from A-Class to B-Class or not? Also, for an amusing anacronism, see Category:Good articles by quality.)
 * For the moment multiple templates are populating GA-Class and FA-Class categories, and we can't change that overnight. However, we can advocate some changes during the introduction of C-Class, and begin to move towards identifying GAs and FAs using Category:Wikipedia good articles and Category:Wikipedia featured articles instead of GA-Class and FA-Class categories. Input from Carl(CBM) and Oleg (Alexandrov) will be invaluable here, but I agree with Walkerma that the relatively small number of articles involved should mean this is not too onerous. Geometry guy 20:26, 26 June 2008 (UTC)
 * I originally opposed the C-class on grounds that it would create too much complexity for a rough-and-ready system, but with the removal of GA and FA I think it's now a welcome addition. We're getting somewhere. If you guys could figure out the technicals that would be great, because it's over my head.--HereToHelp (talk to me) 20:58, 26 June 2008 (UTC)
 * Re your question: "Example: if I delist a GA, do I downgrade WikiProject ratings from A-Class to B-Class or not?" I would add that a delisting from GA / FA is only an indicator that the WP assessment should be adjusted: Refer to the "Black Hole" example above: It was delisted from FA, and assessed as A class. However, given the quality of the article, B class – even C class – would have been more realistic. G.A.S 08:25, 28 June 2008 (UTC)

Arbitrary section break: Require GA for A?
There are a number of technical issues with anything of this sort that need to be resolved. Regardless of where FAs fall into the scale, it's obviously necessary to keep track of them in places like the assessment log. Would FA status be treated similarly to importance, as a third scale along which articles move? If FA-Class categories are removed, how would a project browse their featured articles? And so forth.

More generally, I happen to think that it's perfectly legitimate, and even desirable, for WikiProjects to orient themselves around the presence of FAs as the highest mark of quality with which an article can be labeled. Excising FAs from the scale would have a number of problematic consequences: In short, I don't think this change would be beneficial—at least not to those projects for whom dealing with FAs and A-Class articles is a regular and substantial undertaking rather than a rare joyous event—or that attempting to transform the assessment scale from a WikiProject-employed one to a WikiProject-exclusive one would be beneficial to the overall role of WikiProjects in the community. Kirill (prof) 23:15, 26 June 2008 (UTC)
 * 1) WikiProjects would be encouraged to reject the Wikipedia-wide process in favor of their own assessment of an article's quality.  At present, it's virtually impossible to have a featured article assessed at a lower grade, because FAs are (automatically) graded at FA-Class.  With a separate system, it would be legitimate to have, say, a C-Class FA; aside from fomenting disputes between pro-WikiProject and pro-FA elements, this would make 1.0-related selection more difficult—does the resulting article score as a C-Class, or as a featured article?
 * 2) While projects with no formal process governing A-Class status would not be unduly troubled by the possibility of an A-Class FA, those projects (such as MILHIST) with a formal process would suddenly find themselves dealing with all manner of issues: to deal with a deficient A-Class FA, do we take it to FAR to remove the FA rating, or to A-Class review to remove the A-Class one?  Or even to both, sequentially or simultaneously?  And what happens if the processes produce different results?  How good is the article in reality?
 * 3) If features in the project template make it necessary to know whether an article is featured, but an explicit value for <tt>class=</tt> is not permitted, it would require the project to add a separate, redundant field merely to provide that status to the template code.
 * 4) At present, the choice of which "featured" items are considered to be FA-Class is entirely up to the project.  Featured articles would be tagged as "featured" under this scheme, and presumably featured lists would be as well; but what about featured portals, featured pictures, and all the other types of featured content that might have a project tag on them?
 * I have to come to the same conclusion as Kirill, not just for the reasons he delineated above, but also because of the following:
 * The current scale indicates that the highest quality assessment in Wikipedia is Featured article status. Almost by definition, any article that meets FA criteria will satisfy any WikiProject's A-Class criteria. If it doesn't, then the appropriate way to deal with that article is to ship it to WP:FAR. Hence, there is no need to decouple FA from the top of the assessment scale. Decoupling FA from the scale would actually dilute the meaning of the bronze star.
 * There are large technical challenges to having an ambiguous GA-Class classification. Unless something changed while I wasn't looking, the 1.0 bot stores each article's class as a numeric value; that way, the bot can compare that 50 > 40, or that GA-Class is "better" than B-Class. But what do you do when GA-Class does not mean anything in relation to A-Class and B-Class? A change of this type would require a complete rewriting of the 1.0 Framework, from our category trees to the way the bots operate. The logistics of such a change are terrifying.
 * Removing FA-Class would cause a lot of truly unneeded work for WikiProjects. Having to reassess a lot of FAs to discover that there was nothing wrong with those articles to begin with is a waste of assessor time and resources. Worse, it causes a potential "if it's an FA, it must be an A" mindset, defeating the whole purpose of having those articles reassessed.
 * Most importantly, both FA-Class and GA-Class are what maintains continuity between WikiProject assessments and Wikipedia-wide assessments. Without GA-Class, for example, the boundary between what would constitute a B-Class article and an A-Class article would gradually creep, with some WikiProjects making it more strict than others. The current situation pegs the scale and provides useful reference points: A-Class articles should be better than GAs. B-Class articles are lacking a little bit to be GAs, but meet checklists. C-Class articles are lacking a little bit more to be GAs, and fail checklists, and so on and so forth.
 * Overall, I don't even see the problem with the current scale (I can't stress this enough—the fact that WikiProject assessments can come ahead of community assessments is a testament of the purpose of both FA and GA, since the former is trying to recognize our best content, and the latter is trying to recognize content that is "good enough"), and I see lots of drawbacks to the separation of grades proposed here. The only change I would approve of is version 4, which is adding a "Who can grade" column to grading scheme. Everything else would get a strong oppose from me. Tito xd (?!? - cool stuff) 06:08, 27 June 2008 (UTC)


 * The problem with the current scale is that the way it's set up creates confusion between A-Class and GA-Class. If an article is listed as a Good Article, but it is also meets the A-Class criteria for one of its WikiProjects, which should it be listed under? The A-class criteria are often stricter than the GA criteria, which would imply that it should be listed as A-class; however, GA is a community assessment, which should in theory take precedence over a single project's assessment. While GA is still inserted in between A-class and B-class, there seems to be no way of resolving this problem. There is also another issue stemming from GA's awkward position: the scale is non-intuitive and confusing to people not familiar with Wikipedia. I would be fine with leaving FA at the top of the scale since that does not seem to cause any problems, but we really should do something about A and GA. Separating GA (and possibly FA) from the scale seems to be the simplest way to go about doing so. Pyrospirit  ( talk  ·  contribs ) 17:41, 27 June 2008 (UTC)


 * I agree. I've been around Wikipedia for a while, but not too long, and the scale always confused me to no end, which is why I let other editors classify articles. It was only today, as I read this page, that I finally understood how the scale all works. Splitting FA and GA off would make it much, much easier for newer editors (well, all editors) to understand. Alinnisawest (talk) 17:46, 27 June 2008 (UTC)


 * I completely disagree that this is the best or easiest way to solve this "confusion." IMHO the confusion is all in the statement that A is better than GA but doesn't require GA. The best and easiest way to correct this is to explicitly make GA rating a prerequisite for an A Class rating. All we really have to do is remove the parenthetical from the definition of A and maybe add an image like the GA next to the A. That would be the "easiest" way of fixing the confusion, and much less confusing than instituting multiple class scales. Adam McCormick (talk) 17:55, 27 June 2008 (UTC)
 * "An image like GA" How about: Image:Symbol a class.svg, Image:Symbol b class.svg, and Image:Symbol c class.svg? I'm not too fond of the typeface but it's a start.--HereToHelp (talk to me) 18:06, 27 June 2008 (UTC)
 * I was hoping something to visually imply that A requires GA maybe a cross between and  or between  and  Adam McCormick (talk) 18:20, 27 June 2008 (UTC)
 * I'm inclined to agree with Adam. If an A-Class article should be better than GA, then there is no reason why it would not pass GA.  If it cannot meet the GA standards, then it can't possibly exceed them.  The current situation, with GA not being a prerequisite for A-Class, makes no sense. Some statistics on the percentage of A-Class articles that are also GAs would be very helpful. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 20:35, 27 June 2008 (UTC)
 * The reason behind the lack of requirement is that several WikiProjects have had legitimate complaints with perceived idiosyncrasies with the GA process. However, those were at the time when GA was still trying to establish itself as a process, so I don't know if those are still issues or not. For some data: even though there is no de jure requirement to do so, all of the articles in Category:A-Class Tropical cyclone articles are GAs, so it becomes a de facto requirement. The only exception to the rule are lists, which cannot be GAs, and those are sent to WP:FLC very quickly. Tito xd (?!? - cool stuff) 20:52, 27 June 2008 (UTC)
 * I originally liked the proposal because it would eliminate the issues of the "optional" GA. (That, and it would make my template design job easier but oh well.) However, this would send more traffic to GA (if they can handle it, they'd like that), improve the quality of assessment, effectively forcing some manner of review for A-class, but most importantly, fir within the current system. I like the mandatory GA-for-A idea.--HereToHelp (talk to me) 22:16, 27 June 2008 (UTC)

To repond to Krill and Titoxd, you guys bring up some interesting points, but I think you're misunderstanding the proposal in a couple key ways. I'll try and explain. Hopefully I've been able to clear up some confusion, or if the comments were made with a clear understanding of the proposal, then I hope I've been able to argue for the other side adequately and fairly. Please let me know what you think. Drewcifer (talk) 04:41, 28 June 2008 (UTC)
 * "it would be legitimate to have, say, a C-Class FA" this is absolutely not what we're proposing. Basically, FA would be a subset of A articles.  So all FAs are A-class, but not all A-class articles are FAs.  It is a similar (though not completely identical) case for GA: GA is a mutual subset of B and A: not every B is a GA, not every A is a GA (since it could be FA), but every GA is either A or B.  So, to play devil's advocate, it's again impossible to have a C-class GA.
 * "But what do you do when GA-Class does not mean anything in relation to A-Class and B-Class?" This is not a new issue.  Right now, if an article is GA-status, but gains A-class status, does it cease to be a Good Article?  In terms of other Wikiprojects, the obvious answer is no.  In terms of the Wikiproject that classifies it as A, I think that question is extremely vague.  The GA templates stay on the page.  The article stays listed at WP:GA.  Which makes it, for all intents and purposes, a Good Article until promoted to FA or demoted from GA-status.  For a random example of what I mean, take a look at Talk:Malaria.  An A-class article which still exists within the GA-world.  With the current assessment scale, the dual-personality of the article is just strange and a little unintuitive.  If GA-status was something separate to A-class, then it would make much more sense.
 * "Removing FA-Class would cause a lot of truly unneeded work for WikiProjects" Like Titoxd pointed out, "by definition, any article that meets FA criteria will satisfy any WikiProject's A-Class criteria". That said, reassessing an article from FA>A is a simple matter, since every FA, by definition satisfies A-criteria.
 * As for the the technical stuff which I admit to not understanding: that would probably be a headache to figure out, but, and stop me if I'm wrong, the technical stuff should work around the best possible system, not the other way around. If the community likes this proposal (like alot of proposals), there will have to be some tweaks to some bots and some templates, but IMHO that's not a reason to avoid improving the system (again, if the community deems the proposal a improvement in the first place).
 * "At present, the choice of which "featured" items are considered to be FA-Class is entirely up to the project. [...] but what about featured portals, featured pictures, and all the other types of featured content that might have a project tag on them?" This is actually an unexpected benefit to the proposal, one that I didn't realize until just now.  Since we separate FA and FL class from the assessment scale, the FA/FL designation will mimic almost identically the FPortal/FTopic/FPicture designation, since FA/FL becomes something separate from Stub/Start/C/B/A, something which does not apply to portals, topics, or pictures.  So in a sense, this would clear up alot of confusion about what to do with those other featured content designations: nothing.  Instead of making FPortal/FTopic/FPicture processes mimic FA/FL (which is impossible due to the lack of assessment otherwise), we do the opposite.  Simple solution to an age-old problem.
 * "Without GA-Class, for example, the boundary between what would constitute a B-Class article and an A-Class article would gradually creep" I believe that is simply a problem of being clearer about assessment definitions, something which is desperately needed anyways.
 * I understand your presented proposal. I have no doubt that, given time, it could work. However, I don't like decoupling the two scales and still saying that "GA is a subset of A and B" or "FA must be A" because substantively they mean the same thing as putting them in a line. You are still saying that an article must be a B to reach GA, you are still saying that an article gets to GA before A and still saying that A connects GA and FA, the only difference, from a set theory perspective, is that you are allowing an article to still be labelled B after it reaches GA and A whether it's GA or FA, It doesn't change or lessen the confusion about which classes are prerequisites for which, and IMHO it would drastically increase problems with making the scales get along. As it stands, while it still says that GA isn't required for A it really doesn't work that way in practice, so why not just formalize the linear nature? Adam McCormick (talk) 05:30, 28 June 2008 (UTC)
 * The solution of making GA a requirement for A-Class is attractive in many ways. This means (in theory) that it gets the intense focus of one uninvolved editor representing the community before it returns to the project for peer focus by several uninvolved editors. This broadening of involvement at each step is an excellent way to ensure that the article really does have wide consensus for its status. Under this system, by the time it arrives at FAC, it should have received the support (and input) of at least four uninvolved editors. This will help relieve the reviewer-shortage problems at the higher levels of assessment as fewer articles will be prematurely nominated. For symmetry, A-Class should probably be a pre-requisite for FAC, though this suggestion will probably prove more controversial. -- R OGER D AVIES  talk 05:50, 28 June 2008 (UTC)


 * As I see it, the main reason in favour of removing FA-Class/GA-Class is that it is confusing to have GA (and to a lesser extent, FA) community-based assessments mixed in with WikiProject based Stub/Start/C/B/A. It is my perception that many people doing assessments currently misunderstand this aspect, and it leads to the continual debate about things like "Which is higher, GA or A?" (To me that is a meaningless question, because they are different animals.)  The same problem is frustrating my counterpart on the French 1.0 project.  At the same time, some veteran reviewers (who are very clear on such distinctions) from active projects make a strong case for keeping these levels in the system; as I understand it, the central argument is that the community-wide assessment schemes provide a reference point (or call it a "reality check") against which to judge the WikiProject assessments, and this blend of methods is healthy.  Personally, I think in an ideal world (where everyone understood the subtleties perfectly) the hybrid system would be better, but in reality many inexperienced reviewers are confused.  I'd like to discuss the technical aspects with our bot people but that may be tricky for me as I'm going out of town on Sunday (also CBM is away right now); however, I think technology may be able to help.


 * I don't like the idea of requiring GA for A, because that would lock us into a hybrid system, though if we decide to keep GA/FA in the scheme perhaps it would be better to go "all the way" and do this. The net effect of this would be that all A-Class articles would need to meet WP:WIAGA, and that would IMHO make A-Class almost redundant.  As I see it, we currently have two approaches to FA:
 * Someone inexperienced with wiki protocols expands an article within their field of expertise, adding loads of great content with references. The article is now a comprehensive description of the topic, well laid out and thoroughly sourced.  The WikiProject members look at it and say, "Wow, this is amazing!"  However, it would fail GA on a couple of technicalities/MOS issues.  This sort of article is currently A, because an experienced Wikipedian could clean up the style issues and send it straight to FAC.
 * An article is gradually improved, and an experienced Wikipedian focuses on bringing it up to GA standard. Then another subject expert gets involved, and (perhaps after wider peer review) together they bring it to FAC.
 * Really, it boils down to this: Do we want to track these two approaches to FA, (current system) or only one (proposal)? If we require GA for A then in the first example it would be assessed as B until cleaned up for FAC, and it might then get lost among the other Bs (though it could still go B -> FAC).  Track #1 is probably less common these days, but I think often knowledgeable subject-experts like to be able to focus on making the article complete, without worrying about MOS issues, and WikiProjects often appreciate those articles for what they are - potential FAs.  Do others see things similarly, or not?  Or is all this really quite unimportant anyway, in the wider scheme of things, and maybe we can adapt to a merged or closer A/GA? Walkerma (talk) 06:22, 28 June 2008 (UTC)
 * I agree with Walkerma, and indeed am firmly against requiring GA for A. There does seem to be some wishful thinking about the current system and a certain amount of fear of change here ("Fear leads to anger. Anger leads to hate. Hate leads to suffering" :-) if I remember rightly). Walkerma correctly identifies to distinct ways in which content improvement takes place:
 * WikiProject driven activity using the Stub-Start-C-B-A scale to track progress, perhaps with reference to GA or (project) peer review along the way;
 * An individual or small group taking an article of interest through GAN, peer review, and FAC, often without reference to WikiProjects.
 * Do we want to track the second in WikiProject templates? We certainly don't need to, as there is already a mechanism which does this: ArticleHistory. It does it much better as well, by recording the history, and permanently linking to the reviews. This provides the community processes with a quality assurance role and some accountability. SSCBA is an entirely different beast: it is primarily about tracking, not assuring, allowing WikiProjects to address gaps in coverage and analyse progress.
 * A hybrid linear scale works well for those handful of WikiProjects, like WP:MILHIST and WP:TROP, which have thoroughly integrated tracking and assurance, and interact relatively seamlessly with GAN and FAC. This success generates a myth that there is just one encyclopedia wide scale of article quality, which everyone agrees on. In reality, SSCBA ratings are WikiProject dependent, because they are for WikiProject use: that's why every WikiProject has its own template! If you don't believe me, see Talk:John von Neumann.
 * Asking GA to synchronise WikiProject ratings across the encyclopedia is asking it to do something it never did well and simply cannot do. GA primarily provides a style and compliance check against quite basic community wide principles. It cannot, for instance, ensure that Film GAs and Literature GAs have a similar standard of coverage of style, genre, themes and contextualisation; it is even worse and providing consistency of content across technical science articles. By its nature (typically non-expert reviewers) GA focuses on form over substance.
 * Distinguishing GA and FA from the WikiProject scales does not prevent WikiProjects from using GA as a reality check before A-Class, or using FA as the ultimate seal of approval for its A-Class articles; it gives them the choice. One of the spurious arguments made above is that this rationalization would "dilute the meaning of the bronze star": FA is perhaps the single main driving force for the entire encyclopedia, and it gets its meaning from the rigorous nature of FAC, not from being equivalent to FA-Class! On the other hand requiring GA for A-Class would make A-Class increasingly redundant, and tie the hands of many WikiProjects, particularly those involved with highly technical content, where GA is at its most inadequate. Geometry guy 08:35, 28 June 2008 (UTC)
 * Hear, hear! Couldn't have said it better myself. Girolamo Savonarola (talk) 08:42, 28 June 2008 (UTC)
 * So, just to be clear, SSCBA provides a gauge of how well an article follows the guidelines of a specific WikiProject, while FA/GA (especially FA) look at overall quality without a specialist's eye. You've meade it quite clear that you think GA should not be a requirement for A, but are you also supporting removing them entirely from SSCBA? (Me? My opinion changes with the arguments. This is probably the least of concerns, but an FA or GA talk page has the FA/GA template and then a bunch o projects who will (usually) assess it as such=redundant.) Regardless of the decision made, we need to summarize it clearly and reflect it in the new assessment scheme.--HereToHelp (talk to me) 11:19, 28 June 2008 (UTC)
 * That's not a bad first approximation, although I think you meant to say "especially GA", since FAC/FAR often (though not always) has independent specialist input. However, these things vary from WikiProject to WikiProject, and, at GA, from article to article and reviewier to reviewer.
 * I support clarifying the distinction between GA/FA and SSCBA and I agree with you about the redundancy in stating 3-4 times that an article is a GA. However, as Kirill and Titoxd point out, WikiProjects (including this one) want/need to be able to keep track of their GAs and FAs. At the moment, they can do this with FAs, but not so easily with GAs, because some GAs are A-Class. It is simpler, more robust and more accurate for the bot to obtain this information from Category:Wikipedia good articles and Category:Wikipedia featured articles than from a plethora of GA-Class and FA-Class categories. Geometry guy 14:34, 28 June 2008 (UTC)
 * It seems that you are simultaneously arguing that GA doesn't force an article to have any substance and that it is so close to FA that it makes A-Class redundant. Am I missing something here? How would requiring a low standard (as you seem to describe GA) before reaching A Class make it any less stringent a rating? I don't believe that GA is such a low rating that it is "Inadequate" as a standard. Sure I would never argue that GA is enough but it should be necessary to move up to an A (ie A articles should always satisfy GA criteria). And none of the dehybridized versions above contradict that this would still be the case. Adam McCormick (talk) 18:24, 28 June 2008 (UTC)
 * You are missing almost every nuance, and misusing what I said :-) The main point (read what Walkerma says too) is that GA is qualitatively different from A-Class. It isn't inadequate or a low standard; indeed it is often extremely rigorous in the quality aspects it concentrates on. But it is weak at evaluating technical content, or matters requiring expert input. The current system and most of the proposals do not require GA for A-Class, as A-Class isn't "moving up" from GA; it has a completely different focus. Placing them in order shows a failure to understand how many WikiProjects and content contributors proceed. They focus on depth of content, not on number of inline citations, quality of prose, or Manual of Style issues. They want to create good content first. Hence some WikiProjects have many A-Class articles which are not GAs. WP:GAN can take a month or more, only to yield issues that content contributors find irrelevant; if this were a requirement for A-Class, many would not bother with A-Class. Geometry guy 20:36, 28 June 2008 (UTC)

Another arbitrary section break

 * Let me interject a point I don't recall seeing made. Much of the above discussion seems to be about generic articles in the abstract: an article traverses assessments this way or that way depending on how it evolves. But topics vary widely in their potential depth, and I suspect that the alternative routes around A-class would apply only to more complex articles. Consider a traditional print encyclopaedia article on a deep topic, e.g. the positron. I can conceive of an excellent, well-written, publication quality article of 1000 words that is ideal for a lay or first year undergraduate reader. I can also conceive of a broader article of 3000 words. Which is "better"? A lay editor might consider the first article a FA, and certainly a GA. Physicists might award B-class to the first, and A-class to the second, because the first has perfect scores on all its orthogonal assessment criteria, except subject matter depth. The idea that there is just one ideal scope goal for an FA article strikes me as untenable. Different reviewers will have legitimately disparate opinions on appropriate coverage, and in particular subject matter depth is somewhat orthogonal to editorial/publication quality. Spinality (talk) 19:07, 28 June 2008 (UTC)
 * Geometry guy: Yes, especially GA (oops). Adam: If As should meet GA requirements, should we require GA for A? If we do, Geometry, than WikiProjects know that all of their As are also GAs. Spinality: Good point, but what do you suggest we do about it?--HereToHelp (talk to me) 20:09, 28 June 2008 (UTC)
 * I replied to Adam above: the answer to your question is that WikiProjects want to have A-Class articles which are not GAs, because GA is a process outside their control which addresses different issues and has a different focus.
 * Spinality makes the same point I make: there is not one universal linear scale of quality. Different interest groups measure quality in different ways. Each WikiProject A-Class rating is one way to define an article which is "complete". Not every one of our 2 million+ articles can be FAs any time soon, but many many more could be A-Class, because WikiProject ratings scale with the number of WikiProjects and WikiProject editors; there is no bottleneck. And what a great encyclopedia that would make. Geometry guy 20:46, 28 June 2008 (UTC)
 * ...and probably the two most important quality scales are: a) subject matter coverage and expert assessment of sources, and b) MOS/Wikification/NPOV/citation completeness (the latter two from an editorial rather than a subject matter perspective). This is reflected in generations of editorial practice – it's why all books have both authors and editors. And to answer the question "what do you suggest we do about it" my answer, which I thought was implied but I will make explicit, is to regard GA/FA as independent Wiki-wide editorial statuses, that consider but are not part of the Start/C/B/A levels; the latter are intrinsically more focused on content and technical completeness within a WikiProject. No doubt progress along both scales will normally follow hand-in-glove; but depending on the WP and the article, subject matter or Wikification may lead progress toward deathless prose of glistening, well-sourced accuracy. Spinality (talk) 21:21, 28 June 2008 (UTC)
 * Sorry, I'm nor very good at inferences. Anyway, that would be opposed to the GA-required-for-A position. I think the "independence" idea will work, but the distinction between the authors and editors, which you explained, will have to be emphasized. (This distinction is leas true at FAC, where formatting and accuracy matter.) Moreover, it means that we won't GA/FA talk templates duplicating the assessment, or causing confusion if they don't (i.e. an A-class GA). This is good.--HereToHelp (talk to me) 21:55, 28 June 2008 (UTC)
 * Actually I don't think that "independence" excludes "GA-required-for-A." I do believe it's helpful to have two frameworks for measuring these different qualities of excellence, rather than conflating them. But we (or rather, individual WikiProjects) can still choose to establish strong or weak dependencies among steps in otherwise independent scales. Moreover, I agree that there is not one clear line between editors and authors; most contributors here play both roles. But this does vary. The assessment scheme(s) need to reflect the wide range of topics and articles as they exist. In some cases, an article's MOS conformance and coverage of technical issues will evolve in step; in other cases, particularly in highly technical subjects, the subject matter expert will often not be involved with Wikian arcana. (And of course these are precisely the kind of experts we hope will contribute.) Spinality (talk) 22:50, 28 June 2008 (UTC)
 * Naturally. Some WikiProjects could even decide to make GA a requirement for A. However, if we are to separate the scales, we cannot enforce that requirement from the top down. But when it happens to happen, that's fine.--HereToHelp (talk to me) 23:00, 28 June 2008 (UTC)
 * Yes, yes. The only way to enforce such requirements would be to create Wiki-wide standards and eliminate WikiProject discretion on such matters. Clearly different groups are currently doing a good job using their own measures. We can and should provide guidance and describe intent through clear examples, e.g. illustrate where it would be helpful to require GA for A, and where it might be undesirable. The existence of a prerequisite does not imply a shared, linear scale. Spinality (talk) 23:17, 28 June 2008 (UTC)
 * Whoa, I go out and a debate breaks out. But back to the topic: One thing I do not understand is why some WikiProjects say that they don't want to classify their A-Class articles to be GAs. The only criterion in WP:WIAGA I can see as potentially controversial is criterion 1 (style); the rest cover issues that WikiProjects should be looking at as they write articles. Certainly A-Class articles should be stable, neutral, and at least broad in their coverage? Tito xd (?!? - cool stuff) 23:26, 28 June 2008 (UTC)
 * Just to hazard a guess: maybe it implies that the A-class article meets Ga standards but does not exceed them (which it should, being A-class). I agree, though, that GA status should only be removed by the GA delisting procedure or when they "graduate" to FA.--HereToHelp (talk to me) 01:35, 29 June 2008 (UTC)
 * ← Just finished reading through the latest discussion here, and I see all sorts of excellent points being made that I never thought of when I suggested this. I definitely agree that GA and FA should not require any particular WikiProject rating, and that A-Class should not require GA—though it wouldn't be a problem for individual WikiProjects to choose to require GA for A-Class. (I think some people misunderstood what I said earlier and thought that I was suggesting some of these requirements; I was not.) Wikipedia works by a bottom-up organizational approach, which is the only practical way for a radically transparent and open wiki of this size to operate. We should use the same approach here, and not force requirements such as those on WikiProjects. Also, it seems clear that a fairly strong majority of editors here support separating GA/FA from the WikiProject assessment scale (start-stub-C-B-A). I recommend that we start a new section in Wikipedia talk:Version 1.0 Editorial Team/Assessment rewrite that clearly and concisely summarizes the proposal and the major points made during this discussion; this can provide a starting point for discussion by the wider community. Perhaps a collaboratively edited post would work best to start it off. Pyrospirit  ( talk  ·  contribs ) 04:31, 29 June 2008 (UTC)

Another "assessment"
There are only one place that I know about where the GA assessment is used outside of the Talk and Wikipedia (Wikiprojects) namespace: Featured topics. The criteria are as follow:
 * 1) It is a set of similar, interrelated articles that cover a specific topic
 * 2) There are at least three distinct articles.
 * 3) The articles have a clear similarity with each other under a well-defined topical scope.
 * 4) All articles in the topic are linked together, preferably using a template, and share a common category or super-category.
 * 5) There is no obvious gap (missing or stub article) in the topic. A topic must not cherry pick only the best articles to become featured together.
 * 6) The topic has an introductory and summary lead article or list.
 * 7) Each article is of high quality, including references.
 * 8) A number of articles are featured class (featured articles or featured lists), according to the following rules:
 * 9) At least twenty percent (rounded up) of the articles are featured class.
 * 10) At least two articles are of featured class.
 * 11) All other articles are good articles.
 * 12) Items that cannot achieve a high rating (as stated in 3.1 and 3.2) due to their limited subject matter have passed an individual audit for quality which included a completed peer review with all important problems fixed. Such items cannot be used to fulfill criterion 1.a.

The reason that I mention these, are that the articles that cannot achieve a high rating, are marked with. Should we maybe co-ordinate with those editors, as technically, this also a community assessment, albeit one with a very limited use? It seems to me that this is used for A, B, list class articles. — G.A.S 05:17, 26 June 2008 (UTC)


 * Differentiation of assesment from status seems sensible as it will avoid confusion, this is essentially v3 with simpler names --Nate1481(t/c) 15:07, 26 June 2008 (UTC)

A proposal about the proposal I'd like to propose
Since the above discussion seems to be getting massive and I feel that good points are getting lost along the way or not being discussed thoroughly, I'd like to propose something a little more formal. Something that I saw at the recent GA-icon discussion struck me as a really smart way to approach these kinds of things: after the straw-poll portion, I suggested that a list of the pros, cons, and meta-issues be simply listed for easy access for the closing admin. It acted as a really nice way to summarize a ton of good points that were related but impossibly dispersed throughout a large discussion. That said, I think we could further improve upon that: make a list of pros, cons, and meta-issues as their own little sub-sections, then discuss each one there, rather than in a meandering linear discussion like the one above. Any thoughts on using this method to help organize the above assessment discussion? Drewcifer (talk) 20:19, 28 June 2008 (UTC)
 * PS, here's a link to the GA icon thing I mentioned. Drewcifer (talk) 20:22, 28 June 2008 (UTC)
 * Sounds good, but good luck!--HereToHelp (talk to me) 20:35, 28 June 2008 (UTC)
 * It would be better to start a new section, at least. If it has some structure based on the issues raised so far, so much the better. Discussing all this with lots of level 5 "arbitrary section breaks" is a bit painful :-) Geometry guy 20:54, 28 June 2008 (UTC)
 * As I mentioned in my last post to this page, I recommend we start a new section at Wikipedia talk:Version 1.0 Editorial Team/Assessment rewrite. We can then summarize the proposal and the major points brought up during this discussion in that section (possibly with a collaboratively-edited post or through edits to its corresponding project page). This would both aid the discussion, and it will make things run more smoothly after the proposal is announced through cent or MediaWiki:Watchlist-details. Pyrospirit  ( talk  ·  contribs ) 04:41, 29 June 2008 (UTC)
 * FYI: I'm traveling tomorrow, and I'll be at a conference until Wednesday night. Also, I'm doing lots on Monday (at the conference), so I'll be focused on that until Monday evening.  I should have good internet access the whole time, but if I disappear, that's why! Please come to a clear consensus on everything while I'm away! Walkerma (talk) 04:46, 29 June 2008 (UTC)

Disambig-Class
Would anyone be opposed to moving Disambig-Class and the corresponding categories to Disambiguation-Class? We would keep the current text of Disambig, and keep redirects for ease-of-use. The name of "Disambig" just seems illogical. If it's going to be shortened, it should be the shortest, most easy-to-use (Dab). If it's going to be long, it might as well be the whole word. A bot could then also adjust the templates so they use the new title, which would result in the templates reading "This article has been rated as Disambiguation-Class on the assessment scale", instead of "This article has been rated as Disambig-Class on the assessment scale".  JohnnyMrNinja  08:07, 23 June 2008 (UTC)


 * Disambiguation-class is just too long; I should prefer using Dab-class instead. Where whole sentences are written, however, the entire word could appear—there are no limitations there. Waltham, The Duke of 08:28, 23 June 2008 (UTC)


 * It's the way that project templates like WikiProject Video games use the class templates. While showing the Disambig-Class template automatically when the text "class=Disambig" is entered, the description text they use is based off the exact text entered under "class=". So if you put "class=dab", the template will read "This article has been rated as dab-Class on the assessment scale".  JohnnyMrNinja  05:47, 24 June 2008 (UTC)


 * WPSchools, the template we use at WP:SCH/A uses Dab as it is. "Disambiguation" is just so much longer than any of the other class names it messes with the layout of the templates. Adam McCormick (talk) 14:35, 24 June 2008 (UTC)


 * Then would there be consensus to move it to Dab-Class (which is currently a redirect)? If it is going to be shortened, it should probably use Dab, as it is a very common abbreviation for Disambiguation across Wikipedia. The only place I have seen Disambig is on this template.  JohnnyMrNinja  03:58, 25 June 2008 (UTC)


 * I could agree with that but it's really the color of the bike shed at that point isn't it? Adam McCormick (talk) 04:30, 25 June 2008 (UTC)


 * Please take a look at the above section and tell me if there is anything ironic to be found in your metaphor. I am just trying to be logical and aesthetic, while doing some tidying around the classes at WP:VG.  JohnnyMrNinja  07:43, 25 June 2008 (UTC)


 * This is a threat: the next person to use this "bike shed" expression will end up looking for their head. :-D We have actually managed to agree on a matter purely concerning colours; we can certainly reach an agreement on a matter which, as Mr McCormick himself has said, is also practical. The classes should have short names; moving to "Dab-class" is the way to go. (Literally.) Waltham, The Duke of 07:54, 25 June 2008 (UTC)


 * While not opposed, I do not see a reason to rename the class. Dab-Class redirects to Disambig-Class, and most project templates currently accommodate  .  However, if the decision is made to rename the class, the corresponding category,  and its sub-categories (for all projects) should also be renamed.  --Scott Alter 08:05, 25 June 2008 (UTC)
 * Sorry, what I mean is that it's a template and that it doesn't matter which name redirects to the other as they're all transcluded anyway. Now as far as text length/content is concerned, that is a worthwhile discussion. Adam McCormick (talk) 16:01, 25 June 2008 (UTC)
 * I got myself all caught up thinking about colors and I distracted myself. When I switched horses midstream earlier, I didn't stop to think how silly my new course was. My desire was to see the words Disambiguation-Class, but as it seems that would be too long for templates, would there be opposition to changing the text of the template to Dab instead of Disambig? My main concern is that Disambig is an awkward shortening, and looks ugly on a template. And I take the bus, so these bike shed concerns are unfounded.  JohnnyMrNinja  08:21, 25 June 2008 (UTC)
 * I find "disambig" to be much more self-explanatory than "dab", and would express a general preference for the longest thing which could possibly fit visually. Even after editing wikipedia a long time, I still often need to think twice about "dab", and "disambig" is in wide use as well (at least as the name of disambig. Kingdon (talk) 14:16, 25 June 2008 (UTC)
 * As long as the banners, etc... account for both words I'm in favor of leaving the text as is if only because moving categories is a pain. Adam McCormick (talk) 16:01, 25 June 2008 (UTC)

New C class live
I see that the new C class has gone live for every project that links directly to Version 1.0 Editorial Team/Assessment. I think this is premature, as the definition of C posted is identical to that of B. I propose that it be reverted until the new definition of B class is agreed and posted. --Hroðulf (or Hrothulf) (Talk) 11:09, 23 June 2008 (UTC)
 * I've reverted the inclusion. C-Class isn't really ready for usage yet. Yes projects can start designating articles as C-Class at their discretion, but a suitible description of C-Class standards is still being drafted. What's currently on that section of the template is just placeholder text (the example is an FA article!) -- .: Alex  :.  11:47, 23 June 2008 (UTC)

FA color: gold?
Hi guys. I know that we just messed with the color scheme for C-class, but I would like to know why FAs (and FLs) are categorized in blue. We already have a very nice gold (#cba135) used in places like imbox for featured-related templates. I'm over at tmbox (talk pages) and we're contemplating having color coded assessment classes. But it seems rather silly to have a standard featured template in gold and then an assessment template in blue. I would like to know if there's any support for using the gold instead. (If there isn't, it was worth a try.) vs.


 * Hello again. I shall make no comment, given that I am clearly biased in favour of the present scale; I shall refer you, however, to the debate above for an example of how the scale is seen here. Waltham, The Duke of 23:49, 25 June 2008 (UTC)
 * Alright, never mind. Color-coded tmbox classes don't seem to be happening, anyway.--HereToHelp (talk to me) 00:28, 26 June 2008 (UTC)
 * Well then maybe:

? Gears  Of War  00:34, 26 June 2008 (UTC) (Note i left out A because it cans stay the same. Gears  Of War  00:36, 26 June 2008 (UTC)
 * FA/FT/FL:Platnum
 * GA:Gold
 * B:Silver
 * C:Bronze
 * Start:Copper
 * Stub:Who knows?

←How about: ...and so on as we have now. Or entirely as we have now, if it's too set in.--HereToHelp (talk to me) 01:02, 26 June 2008 (UTC)
 * FA/FL/Fwhatever: Gold
 * A: silver
 * GA: green


 * (edit conflict; part replying to Gears of War) But it won't work if you leave A-class as is; what does light blue have to do with precious metals? (Unless it's some alien metal, mwahaha.) Ah, well, it won't work anyway, unless you make them a little shiny and glossy, departing from the current one-uniform-colour scheme.
 * It's fun, though. :-) It reminds me of the service awards.
 * (new part, replying to HereToHelp) Now that I think of it, no precious-metal scheme would work, even theoretically, simply because FA is too associated with a bronze star. Waltham, The Duke of 01:07, 26 June 2008 (UTC)
 * I agree with The Duke of Waltham. I always liked the blue myself: It reminds me of a "royal blue"; that said, metals just aren't shiny on a monitor. G.A.S 04:57, 26 June 2008 (UTC)


 * Urmm... "Shiny-looking"?
 * "Appearing to be shiny"?
 * "Conveying to the viewer a vivid impression of profound shininess"? :-D Waltham, The Duke of 05:05, 26 June 2008 (UTC)

No use in metallic colours, since monitors do not display them properly. G.A.S 05:21, 26 June 2008 (UTC)
 * Shiny: Adjective 1. reflecting light.


 * Okay, so we leave everything else alone and just change FA to bronze/gold. Because a split between ABC and FA/GA is imminent, having metallic (as in metal-like, not shiny) colors in both schemes doesn't make much sense. With FAs out of the assessment scale, the talk assessment templates can say A-class--with sky blue--and a separate template, like what we have now, can say featured--with gold/bronze.--HereToHelp (talk to me) 12:13, 26 June 2008 (UTC)


 * Sky blue just won't be the same without the star :(. I would say though that I am biased towards the current scheme, and FA has had this color for a long time though. It seems to me that the blue text does not work so well on the gold background and neither does the bronze star (it seem to disappear), where the blue background provides a very nice contrast (While the text seems quite visible). Remember also to test for Accessibility. Is the sky blue such a big problem within the new style templates? G.A.S 13:33, 26 June 2008 (UTC)

Edit point
←Well rather than having a small swatch on an assessment template, if featured and A-class (sky blue referred to A-class, not featured, sorry) are separated, featured will be denoted more like


 * I see: Copied here for convenience:


 * If this is to be done, I would rather that A class inherit the old color, like this (although I would still prefer FA to be blue) —If what we are going to do, happens, FA effectively becomes N/A as a assessment. G.A.S 14:31, 26 June 2008 (UTC)


 * A, B, and C keep their current colors. FA and GA status will now be indicated by a gold or green (respectively) colored template. We won't see this--


 * --because we're using the above template. Most/all FAs will also be A-class, so on their assessment templates they will have the sky blue. (Ok, now that I think about it FA's blue is closer to sky blue. Oh well.) This is different from the color bar because the star doesn't go on the color bar. (It might be possible to color-code the assessment based on the class, but it's probably more trouble than it's worth. it's too early now, anyway.)--HereToHelp (talk to me) 15:20, 26 June 2008 (UTC)


 * Exactly: But consider the above discussion: FA will actually not be an assessment anymore. FA articles will thus have two templates as follow: One for FA status and one assessing them as A-class (As opposed to the current situation, one for FA status and one assessing them as FA class). All I rather is that the (less than pretty) color for A class be replaced with the one of FA class (which will use gold). G.A.S 16:02, 26 June 2008 (UTC)

←Ah! You want (#66ffff) to become (#6699ff ) right? Fine by me. Be forewarned that there will be a certain amount of social inertia (people who don't like change) to fight. I don't particularly see any benefit to this (although the "new" color is further from the "notice" color on various templates) but hey, it's a free wiki. You'll have to wait until FA is removed from the assessment scale though.--HereToHelp (talk to me) 21:11, 26 June 2008 (UTC)


 * I've been following all this "secession" discussion somewhat bemused, and I have two words to say: Slow down. Waltham, The Duke of 09:45, 27 June 2008 (UTC)

As someone who is colour blind, I would much rather that we use a very wide range of colours, than simply those colour that might represent "metallic colours". The current FA colour is much better than the proposed "gold" colour... Bluap (talk) 13:47, 27 June 2008 (UTC)


 * My point was that we're going to have GA and FA templates on talk pages, and assessment. It would keep everything consistent, because the gold on the separate template isn't changing. If you want a wide range of colors, for A and FA, two blues are better than blue and gold?--HereToHelp (talk to me) 22:26, 27 June 2008 (UTC)

Huh?
Why is B Class suddenly showing up as the same color as GA? Naerii 00:49, 27 June 2008 (UTC)
 * It's not the same color, it's a yellower version. See discussion above. Adam McCormick (talk) 01:02, 27 June 2008 (UTC)
 * I think the problem is...where is the C-class? nvm --haha169 (talk) 18:33, 27 June 2008 (UTC)
 * I have to agree though, that on my computer the colors are very hard to distinguish. Can the GA green be made less yellow? --EncycloPetey (talk) 18:33, 28 June 2008 (UTC)
 * Actually, when I posted my comment, it was temporarily removed from the list for more revisions. It was just re-added back on. Mozilla Firefox 3 views the colors perfectly, easy to distinguish. --haha169 (talk) 04:56, 29 June 2008 (UTC)
 * On IE7/XP, LCD screen, the two colours are distinguishable, although they could be more so. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 11:34, 30 June 2008 (UTC)

B-Class colour
I have proposed a minor revision to the B-Class colour to help make it more distinctive. See the template's talk page for details and discussion. Tom pw (talk) (review) 10:17, 29 June 2008 (UTC)
 * On a related note, would it be sensible to create CSS classes (, , etc) for the various colours and load them into MediaWiki:Common.css? That way, any updates like this could be done easily in one place, rather than having to change the colour in a number of different places. It would also make the colours fully skinnable. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 10:28, 29 June 2008 (UTC)
 * Happy-melon, sounds like a good idea to me. This definitely seems like the sort of thing that should be a CSS class. Perhaps we should also create "dark" versions of these classes that are similar but slightly darker, made for use as a text color rather than a background color; the current colors work fine for backgrounds but are too light to be readable text colors. Pyrospirit  ( talk  ·  contribs ) 02:47, 30 June 2008 (UTC)
 * Ok, I've now added CSS classes for FA, A, GA, B, C, Start, Stub, List and Dab to common.css - 30 day recaching lag starts now, so we can implement those classes everywhere as of 31st July. I created two classes for each grade: , for instance, sets a background colour and   sets a text colour.  So the former will be useful for , etc, while the latter can probably be worked into the metadata script.  <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 11:43, 30 June 2008 (UTC)
 * Thanks. I'll start working on adding those into the metadata script. Pyrospirit  ( talk  ·  contribs ) 18:17, 30 June 2008 (UTC)
 * Don't take it live yet, or it'll blank out for all those with cached versions of common.css (which is everyone who hasn't purged their cache or had it auto-recache in the past 8 hours). Good to lay the groundwork though. I took the current text colours straight from the metadata script, so there shouldn't be any display changes when you do implement it. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 18:23, 30 June 2008 (UTC)
 * I already updated the live version to use the new classes; however, this shouldn't be a problem since I also placed a prominent notice at the top of its documentation page mentioning that you may need to bypass your cache for the colors to display. The page is linked from the gadget's description in Special:Preferences, so it is the first place people will look if the colors stop displaying for them. Pyrospirit  ( talk  ·  contribs ) 05:21, 1 July 2008 (UTC)
 * It's not working: all the headings show up as black (but with the "an XX class article from Wikipedia" underneath). I've cleared my cache several times, and the classes appear correctly in User:Happy-melon/sandbox3, so it must be the script and/or its interface with IE7.  Any ideas? <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 12:10, 1 July 2008 (UTC)
 * Same here. D.M.N. (talk) 13:46, 1 July 2008 (UTC)
 * I've reverted to a stable version for now. There's no rush to get this out, so take your time to get it working correctly. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 15:26, 1 July 2008 (UTC)
 * That's odd. I tested it pretty thoroughly, and it worked just fine for me. I'm using the test version in my userspace, which is using the classes, without noticing any problems. I'll try testing it in Mammon IE7 to see if that's causing the problem. Pyrospirit  ( talk  ·  contribs ) 16:59, 1 July 2008 (UTC)
 * ← (outdent) It was just a problem with how the  attribute is accessed in IE. I made a few changes and tested them out in Firefox 3 and IE7, and it seems to work just fine in both. But you're right, the differences between versions appear pretty minimal from the outside, so we can probably wait a little longer before synchronizing it with the gadget version.  Pyrospirit  ( talk  ·  contribs ) 18:13, 1 July 2008 (UTC)

How should we announce the "official release" of C-Class?
Cross posted here.

I think we should make an announcement (I'm proposing Friday, is that OK?) to all WikiProjects that C-Class is now official. Is everything in place? We will have to leave the GA/A issue for now, though I think we can tighten up on peer review a little bit on A.   I think we can make some changes to B (and a little bit to Stub). I'm away at a conference at the moment, but I should be able to devote some serious time to it tomorrow evening.

Should we use AWB to post on all WikiProject talk pages? Or is there a better way? A bot? If we use AWB, should we break up the work? Walkerma (talk) 04:33, 1 July 2008 (UTC)


 * AWB running through talk pages of Category:WikiProjects participating in Wikipedia 1.0 assessments would be easiest. I'd also be happy to subvert and distrubute one of my own automated but manually-approved python scripts to help with the posting. Adam McCormick (talk) 04:48, 1 July 2008 (UTC)
 * I would be happiest if we can postpone the "offical release" until Friday so that I have time to make some updates to Igor first. – ClockworkSoul 04:53, 1 July 2008 (UTC)
 * You mean on Friday, or after Friday? Tito xd (?!? - cool stuff) 17:23, 1 July 2008 (UTC)
 * Split the difference... Friday evening? That'll give me the day to do my testing and uploads, and to hide the features that aren't ready for prime time. – ClockworkSoul 17:34, 1 July 2008 (UTC)
 * How will the codes for the statistics (such as this example) change to accommodate the new C-class? And by the way, we have other classes such as Disambig, Category, and Portal class but I can't get the codes to work in the example above. <b style="color:#0000FF;">OhanaUnited</b><b style="color:green;">Talk page</b> 17:59, 1 July 2008 (UTC)
 * This works. You should create an appropriate category (Category:C-Class_Solar_System_articles in the example) and then modify project template (See [|this diff]. For other templates it can be slighly different). You can add other classes in the same way. The bot will update statistics automatically. Ruslik (talk) 18:30, 1 July 2008 (UTC)

I'd like to suggest that we start at 1800h UTC on Friday, but try and get everything completely finished at least 12 hours before (0600h UTC), to allow time for any major blunders to be fixed. A couple of things we need to do:
 * Incorporate the results of the above Stub/B/A discussions from above into the long definitions table.
 * Write the wording of the announcement.

I'll be offline for most of the next 24 hours or so, but I've asked Titoxd to keep things moving along. Walkerma (talk) 15:41, 2 July 2008 (UTC)


 * You may be very short-staffed if you try to do this Friday. Remeber, that's a major US holiday involving fireworks. --EncycloPetey (talk) 16:01, 2 July 2008 (UTC)
 * I'll be available for a few hours today, so I'll try to do the most I can, but please, any and all help is appreciated. Tito xd (?!? - cool stuff) 18:17, 2 July 2008 (UTC)
 * Hehe who knows, perhaps the on-wiki fireworks will match the external ones! <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 20:02, 2 July 2008 (UTC)
 * I didn't know these colonials had got independence! I wouldn't have moved here if I'd known! (No wonder I ignored the data when I chose the day.)  If the worst comes to the worst, we'll have to finish things off on Saturday.  Walkerma (talk) 14:13, 3 July 2008 (UTC)

Category tree is done - I think that's the last of the infrastructure changes that need to be made. We just need to hammer out the actual criteria (decide whether we want to codify the MILHIST B-Class ratings, for instance) and hit the big red button to start the music! <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 20:02, 2 July 2008 (UTC)
 * The relevant B-Class changes have been done in the tables at Version 1.0 Editorial Team/Assessment rewrite. I also created Version 1.0 Editorial Team/B-Class criteria, which can be used for transcluding them elsewhere, as well as for easier arguing ironing out the final text. Tito xd (?!? - cool stuff) 22:09, 2 July 2008 (UTC)
 * With final text and a category to go off of, I can send out the notices with my bot; so you can spend your time doing other things if you wish. I should be available most of Friday and Friday evening.  §hep   •   ¡Talk to me!  06:38, 4 July 2008 (UTC)

Early archive
This talk page is currently very long due to the ratification vote. Is there consensus to archive the following to the 2008 archive, as these discussions have been completed (As the "vote" has been concluded). If needed we can keep the section here and link to the content in the archive. I also considered a collapsible archive header, but this does not solve the loading time of the page, nor does it work with subheadings. Regards, G.A.S 06:17, 23 June 2008 (UTC)

✅ Done G.A.S 04:52, 24 June 2008 (UTC)
 * Ratification vote on C-Class
 * Quality assessment scale is Wikipedia-wide
 * Optional implementation... Reversed
 * Suggestion: rename to Start and Average


 * Aye – Waltham, The Duke of 07:15, 23 June 2008 (UTC)
 * Sounds fine. I just wanted to wait a few days to allow the dust to settle from the vote. Thanks, Walkerma (talk) 14:25, 23 June 2008 (UTC)

More archiving
I have archived the following sections for the same reason (page bloat). <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 21:27, 4 July 2008 (UTC)

Main C-Class example
Which example do people like the best, to illustrate C-Class? Or is there something better? Additional examples
 * Tropical Storm Tammy
 * Pennsylvania Route 739
 * Terracotta Army
 * Atom (from 2005)
 * Exeter Cathedral

For our main example (the one in the main tables), we need to have something that is illustrative of the main problems in C-Class, yet showing the good features compared to Stub & Start. It should be on a topic that be allows it to be widely applicable for other subjects, i.e., a wide number of projects would find it a relevant example. The quality level should be right in the middle of the C-Class "zone", i.e., it should be such that at least 9 out of 10 reviewers would put it as C without any doubt.

Based on these, I'm inclined to shy away from roads and cyclones articles, because I think the style of these articles is fairly specific to those topics (though we SHOULD include such things in our supplementary examples list). I'd also like to have a C-Class that has one section cleanup tag if possible (I don't think any of the above have these!), to illustrate that aspect. But most important, we just need the best example! Which is the best? Walkerma (talk) 14:37, 3 July 2008 (UTC)


 * All of them seems good, but it is difficult to use Tropical Storm Tammy as an example, for the reason said.
 * The best idea would probably to go through a "representative Featured Article's" history to find the stage at which it was C class. — I will go have a look.
 * G.A.S 06:33, 4 July 2008 (UTC)


 * OK, I just saw your post come in, right after I added the Exeter Cathedral example (cited above) into the draft tables. I was a little uncomfortable with the four examples listed above, but I didn't get any response within 12 hours, so I went ahead and found this one which had been assessed by a project as C-Class.  I like this example better than the original four, because I think it's pretty typical/representative of an average C, but if others have something nicer that would be OK too.


 * If we want an example of a C-Class from something that became an FA, the one from above would be the 2005 version of Atom. However, I think that would be a non-ideal example of a MODERN C-Class, because it represents a more typical example of how things were in 2005 (when I would have rated it as B, by the way).  If there is a nice FA example with a fairly recent C version, I'd be happy to see that one replacing the Exeter Cathedral example.  (We'd need an FA that underwent a lot of improvement recently - probably a less well-known, recent FA.) Thanks! Walkerma (talk) 07:38, 4 July 2008 (UTC)


 * The new example is better than the original four. This means that the best place to look for applicable FA's is Featured article candidates/Featured log, specifically within the recent promotions. I will be able to look at them a little later. Can you identify a few of the recent ones that may provide good "generic" articles? I will then search through the histories. G.A.S 08:54, 4 July 2008 (UTC)

Some questions before the official rollout of C-Class
(By "this page" I mean the article space page corresponding to this talk page!)

This page is linked to by loads of banners, and the tables on it are used on loads of WikiProject pages, so we need to get this right. I'd like to get people's opinions on three interrelated questions. We now have a draft "sandbox" version available here. Please feel free to make minor tweaks to wording, but please discuss any major changes here before editing. I presume that each table will be done as a template, similar to Grading scheme. My questions are: Does it look OK as it is, or should we split the detailed table off onto a separate page? Would it be better done in a "summary style"?
 * Split the descriptions?
 * I think it's OK as one single page, even if more classes are added, but I'd like that view confirmed by others. Walkerma (talk) 07:17, 4 July 2008 (UTC)
 * I think we should merge the detailed table into the shorter ones. There is no need for the shorter ones to be so short, and no need for the detailed one to be so detailed. I believe the presence of two separate descriptions of the same thing originated in separate proposals, and it was never intended to use both. Geometry guy 08:27, 4 July 2008 (UTC)
 * Actually, IIRC, the detailed table was supposed to be the "official" one, and the short tables a summary of the official table. So I think that these should be two separate pages; e.g. grading scheme and grading scheme short. Tito xd (?!? - cool stuff) 08:36, 4 July 2008 (UTC)
 * Cough collapsible "more detailed criteria" sections cough :D <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 11:06, 4 July 2008 (UTC)
 * The issue here is that a casual reader of WP may just want a quick description of the grades (surveys tend to indicate that most people don't read lengthy descriptions anyway); however, we also need something with more depth as a reference for those who do a lot of assessments. Tito is right that the detailed one is the "official" version, BTW, that's why I asked.  G-Guy, this was how I tried to combine the two proposals (one to refine the scheme, another to simplify it), I proposed a summary style approach which people liked.  But would Happy Melon's collapsible version be better still?  I personally think it's a great idea, but I'd like to hear other's views.  HappyMelon, could you put something together in your sandbox for us to look at?  Walkerma (talk) 14:48, 4 July 2008 (UTC)
 * What do you think I've been doing all day? :D Just uploaded to the rewrite page. This discussion properly belongs there - shall we adjourn? <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 15:58, 4 July 2008 (UTC)
 * This new combined table looks great, thanks Happy-melon! Do others like it?  We can certainly do what you proposed, and thrash out the details here. Walkerma (talk) 16:24, 4 July 2008 (UTC)
 * I like Happy-melon's approach too, but I think it would be better to have separate tables for the three types of assessment (SSCBA, community, and other). This was a strong feature of the earlier summary tables. The "collapsible collapsed" detail is only needed for SSCBA: there is nothing official about the detail here for FA and GA; the official versions are WP:WIAGA and WP:WIAFA. Geometry guy 20:02, 4 July 2008 (UTC)

We have coverage now of List, Disambig, Needed and NA-Class in the "short version" table. Do we need to add these classes to the detailed listing as well?
 * Additional detailed descriptions for List, Disambig, Needed and NA?
 * I really don't mind either way. Walkerma (talk) 07:17, 4 July 2008 (UTC)
 * With no detailed listing, this is not an issue. These descriptions should remain short. Geometry guy 08:27, 4 July 2008 (UTC)
 * Don't really mind (although see the caveat below), as long as the difference between Needed-Class, Unassessed-Class, NA-Class and -Class is explained somehow. Tito xd (?!? - cool stuff) 08:36, 4 July 2008 (UTC)
 * Agree that the NA/Unassessed is the most important distinction. If we combine the tables with collapsible sections, as above, then this ceases to be an issue: if a class doesn't have too much detail, no need for a collapsible section. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 11:06, 4 July 2008 (UTC)


 * Additional nonstandard classes

Other classes are sometimes used, such as Category, Image, Portal, Template, Redirect. Should we add descriptions for these, as has been done here? If we do add these, should we have short descriptions ONLY for them, or should we have both short AND detailed descriptions? Thanks, Walkerma (talk) 07:17, 4 July 2008 (UTC)


 * I don't have strong views on this myself, though if I had to choose, I'd say that we should add short descriptions for any classes that are widely used, or at least list these classes. Walkerma (talk) 07:17, 4 July 2008 (UTC)
 * Short descriptions would be a good idea, as with List, Dab, etc. Geometry guy 08:27, 4 July 2008 (UTC)
 * WPBannerMeta mentions the "extended 1.0 assessment scale" as part of its template magic. In many ways, if we're going to describe frequently-used nonstandard grades (also including Future-Class and Current-Class), we could put these in an "extended" section of the page. Tito xd (?!? - cool stuff) 08:36, 4 July 2008 (UTC)
 * Hmn, I don't remember supporting "current class". Looks like I need to check that. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 11:06, 4 July 2008 (UTC)
 * Most of these nonstandard classes were not voted on, they simply evolved, wiki-like. One project decided they needed something, and other projects copied them, etc.  I've always taken the view that "If it's useful to your project, then use it", even in cases (like Image-Class) where we don't need such tags ourselves, and I think others in the 1.0 project have also felt the same way.  I think that for the sake of standardization, we should acknowledge these classes on this page, even if the bot doesn't read them.  The only controversial one was Needed-Class, because it resulted in a talk page with no corresponding article, and that goes against something in MOS (I supported it on grounds of WP:IAR.  Walkerma (talk) 14:30, 4 July 2008 (UTC)

Signup for posting to WikiProjects
Please people sign up to help in distributing our message to WikiProjects? We need to use We'll do this via WP:AWB, though people are welcome to use their own scripts if these are simpler. I'd like to see if we can get every project reached by 1800h UTC on Saturday July 5th if possible. Walkerma (talk) 17:27, 4 July 2008 (UTC)

I've been held up posting the draft message because I've been working on fixing the A-Class description, which (I didn't realize till this morning) doesn't match with our conclusions from above. I don't want to start the posts until the wording for A-Class is fixed - hopefully in a couple of hours, which will give people a chance to read and edit the draft message. Walkerma (talk) 18:02, 4 July 2008 (UTC)

OK, I'll copy over the "combined table" onto this page (article space), then we can start posting the message, I think! Walkerma (talk) 20:03, 4 July 2008 (UTC)


 * We should use a standard title for the messages. What will it be?  §hep   •   ¡Talk to me!  20:08, 4 July 2008 (UTC)


 * How about Changes to the WP:1.0 assessment scheme? If you like that, copy it into a comment in the draft below.  BTW, I hit a snag copying over the new criteria, so I'm holding off until Happy-Melon can figure it out for me (or can you help, Shep?).  Thanks a lot for your bot's help!  Will you start at the letter C?  Or would it be simpler just for you to do the whole lot?


 * We also need a list of the projects involved. I suggest we use this list.  Is that OK with your bot, Shep?  Walkerma (talk) 20:27, 4 July 2008 (UTC)


 * OK, it looks as if Happy-Melon is able to make the requisite changes to the scheme ready for copying over. Any minute now! Walkerma (talk) 20:33, 4 July 2008 (UTC)


 * (ec)Sorry, that coding is a bit over my head. I could either do it all or start at Z and move up. Whichever you like. I actually scraped a list of projects participating from the Index and Index 2. Hope that's okay. Title looks good aswell. As a side note, I'll be leaving in about a half hour 45 min to go watch fireworks. §hep  •  ¡Talk to me!  Today, 4:35 pm (UTC-4)


 * Yes, I need to leave soon (I'm in your time zone, and my American wife insists on dragging me to these things). Hopefully we can start soon.  I set a goal of getting this done within 24 hours - will you have any time later/tomorrow to do this?  I'd suggest you start with Z and work backwards - but can you tell us where you reach when you stop?  Thanks a LOT, Walkerma (talk) 20:43, 4 July 2008 (UTC)

Changes to are done - let's hit the road! If you want a list, ClockworkSoul has an XML database at http://igor.sourceforge.net/data/wikiprojects.xml to play with. Either way, roll the press!! <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 20:44, 4 July 2008 (UTC)


 * OK, Happy-Melon says we're ready to go! Assuming you don't get Z-A done now, I'll try doing some tonight when I get back from fireworks. Walkerma (talk) 20:46, 4 July 2008 (UTC)


 * Everything running smoothly here. §hep   •   ¡Talk to me!  20:49, 4 July 2008 (UTC)


 * There are 1,300 subcats of Category:C-Class articles, and yet only 200 in the category you're parsing. I'll try and generate a more comprehensive list. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 21:00, 4 July 2008 (UTC)

1,300 is close to the number I had. I didn't parse a category. §hep  •   ¡Talk to me!  21:02, 4 July 2008 (UTC)


 * Good good, looks like I'm playing around with the API for nothing, but I'm glad you got a complete list. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 21:18, 4 July 2008 (UTC)


 * Great! A couple of things, though.  Over at Grading_scheme there is a box in the documentation saying: "Please note that the C-Class section has been prevented from being transcluded on other pages as it is currently a placeholder and only shown here so other users may review the section addition of this class."  Is this still true?  If so, can we allow the transclusion?  If it's no longer true, can we remove this message box?  Secondly, weren't we going to add in the nonstandard classes into the table?  Walkerma (talk) 21:10, 4 July 2008 (UTC)


 * No, not any more, I'll remove it. Not sure about the nonstandard classes, but that's definitely not time-critical. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 21:18, 4 July 2008 (UTC)
 * .:Alex:. beat me to it. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 21:20, 4 July 2008 (UTC)


 * I didn't have the bot create any new pages, so if the project didn't have a discussion section, or is in an awkward location I may have missed; I put a log at User:ShepBot/Skipped and was hoping you guys could help me go through it. 1,131 projects were notified. Thanks!  §hep   •   ¡Talk to me!  03:59, 5 July 2008 (UTC)


 * Most of those look like task forces or subprojects that are really just part of bigger projects. I suspect there are only a handful that got missed, mainly because of nonstandard naming somewhere.  I'll take a closer look tomorrow.  Thank you! Walkerma (talk) 04:10, 5 July 2008 (UTC)

Draft of message
As you may have heard, we at the Wikipedia 1.0 Editorial Team recently made some changes to the assessment scale, including the addition of a new level. The new description is available at WP:ASSESS.
 * The new C-Class represents articles that are beyond the basic Start-Class, but which need additional references or cleanup to meet the standards for B-Class.
 * The criteria for B-Class have been tightened up with the addition of a rubric, and are now more in line with the stricter standards already used at some projects.
 * A-Class article reviews will now need more than one person, as described here.

Each WikiProject should already have a new C-Class category at Category:C-Class_articles. If your project elects not to use the new level, you can simply delete your WikiProject's C-Class category and clarify any amendments on your project's assessment/discussion pages. The bot is already finding and listing C-Class articles.

Please leave a message with us if you have any queries regarding the introduction of the revised scheme. This scheme should allow the team to start producing offline selections for your project and the wider community within the next year. Thanks for using the Wikipedia 1.0 scheme! For the 1.0 Editorial Team, ~

I can help

 * I'll start by doing letters A and B; I've never used AWB like this before, but if it goes well I'll be happy to do more. Walkerma (talk) 17:27, 4 July 2008 (UTC)
 * My bot (which uses AWB) can help as well. §hep   •   ¡Talk to me!  17:29, 4 July 2008 (UTC)
 * I've notified WP:VG, so please exclude that if possible from AWB/Bot -- .: Alex  :.  20:54, 4 July 2008 (UTC)
 * As a note; I set ShepBot to skip the page if it contains "the wider community within". A random string of text from the notice.  There should be no double posting if we do this.  §hep   •   ¡Talk to me!  20:57, 4 July 2008 (UTC)

Assessment category names
Categories for discussion/Log/2008 June 24 - This 12-day old CfD is for non-article assessment categories which I created using logical naming, following in the pattern of Category:WikiProject Video games articles. As they weren't articles, I thought Category:WikiProject Video games images made a lot more sense than Category:Image-Class video game articles (this format didn't even occur to me at the time). It was then suggested that they be renamed to fit with other projects. I conceded, as my intention the whole time was to make things as easy-to-use as possible. Now, even though I am the creator of the categories, and I was one of the few who got these classes added to our project template, I cannot get anyone to close this CfD because the old names are more logical. My post to WP:AN just got another vote to oppose. I do not care what they are named at this point, I would just like it closed so I can progress with updating our template. So I guess my points are this -
 * 1) Why is functional naming preferable (to WP 1.0) over logical/literal naming?
 * 2) Why are category naming standards not written if editors are expected to follow them?
 * 3) Will someone please close that freaking CfD. At this point, I do not care if all of the categories are renamed Judy Garland.

I realize that there are many categories that already exist, but look at Category:NA-Class articles and you can see that they aren't all the same at this point. I am talking about what ideally would they be named.  JohnnyMrNinja  13:32, 5 July 2008 (UTC)


 * These names have never been a standard part of the WP:1.0 system, and they were set up by individual projects to serve their own purposes. They're still not monitored by the WP:1.0 Bot.  If a standard naming system has evolved (I don't really know!) then maybe we should describe that here at 1.0; so far, we've just noted the fact that these categories exist. With something like this that just evolves, it's very hard for us to impose an alternative from outside, even though your new names make more sense.  And when you have standard systems you always have silly anomalies; in the same way you have Interstate highways in Hawaii, on WP you have Category:Good Articles by quality (which I in effect set up) - it's just simpler that way (GA can now track how many GAs there are using our bot). I'll ponder your comments before I respond at the CfD discussion.  Walkerma (talk) 17:28, 5 July 2008 (UTC)
 * Does anyone here have any interest in the category names? Because the CfD is still open aften 14 days, with little comment. I think people are too confused to know how to close it, since WP 1.0 was referenced.  JohnnyMrNinja  09:55, 7 July 2008 (UTC)
 * I ignored the fact that I'm hopelessly involved and closed it anyway. The categories were actually CSD candidates anyway. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 11:05, 7 July 2008 (UTC)

Question about A-class
Can an article be rated A-Class without being a GA? Sceptre (talk) 00:30, 6 July 2008 (UTC)
 * I'd guesss generally not. See A-Class criteria. And article should be almost FA, so an A article should be able to pass GAN.  Is that correct?  §hep   •   ¡Talk to me!  00:34, 6 July 2008 (UTC)


 * Quite wrong. It can be, and the current changes have made this more confusing than before. Geometry guy 00:36, 6 July 2008 (UTC)


 * Learn something new everyday. My apologies for the misinformation.  §hep   •   ¡Talk to me!  00:39, 6 July 2008 (UTC)


 * Yeah, this is obviously still a confusing matter for people. Something's gotta change. Drewcifer (talk) 00:55, 6 July 2008 (UTC)


 * A-class is not always a level higher than GA-class. It's WikiProject-specific. Gary <b style="color:#02b;">King</b> ( talk ) 07:29, 6 July 2008 (UTC)

We need to clarify this, as Geometry guy says. However, I thought there was some movement to change to require GA status. The A-Class criteria make it quite clear that an article shouldn't be able to get to A-Class without being GA material, so unless anyone can show examples of A-Class articles that have failed or wouldn't pass GAC, I'm inclined to say "no, an article can't be A-Class without GA", purely because if an article genuinely is A-Class material, GAC ought to be a breeze. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 10:09, 6 July 2008 (UTC)


 * GA and A-Class measure quite different things. They have different purposes, different processes and different editors involved. GAN most certainly is not a breeze, especially if you get a thorough reviewer such as Awadewit. On the other hand, GA is not good at assessing technical content. That is what WikiProjects do, and they need the freedom to do it, without having their hands tied by a long wait to jump through a hoop which they sometimes find irrelevant. These points have been made multiple times by Walkerma, myself, and others.
 * An easy way to clarify this would be to list the WikiProject, community, and other assessments in separate tables, as was done in the summary tables. Geometry guy 11:22, 6 July 2008 (UTC)
 * Do they? I'm not sure I agree with that. I don't deny that WikiProjects and GA reviewers are better at measuring different things, but I think that they are ultimately looking for the same attribute: that ephemeral level of "quality". I'm not sure what the problem is: indeed GA is not good at assessing technical content - that's why GA reviewers don't hand out A-Class!! A-Class articles need the level of technical detail that only WikiProjects can assess... but they also need all the things that GA reviewers are looking for.  Given that A-Class articles are supposed to be borderline FACs, GA really should be a breeze no matter who reviews the article: if the article is struggling at GAC, then that ought to raise serious questions about its A-Class standards anyway.  I agree and sympathise over the backlog at GAC, but surely that needs to be solved by more reviewers and participation, not by not reviewing articles that really ought to be easy passes. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 11:30, 6 July 2008 (UTC)
 * I disagree with the premise that A-Class articles are borderline FAC. The top end are.
 * But the bottom line is that WikiProject assessments should remain within the control of WikiProjects. Hence, for example, MILHIST can choose not to adopt C-Class, and Mathematics can continue to use Bplus if it wants to. The idea of a single and universal Wikipedia-wide quality scheme is an illusion. One more time, see Talk:John von Neumann. No hey banda. Geometry guy 11:46, 6 July 2008 (UTC)
 * G-Guy is absolutely right here. If there is confusion here (and there has been since day 1 of GA-Class), we should try to clarify the situation as best we can.  Anybody fancy drafting something clearer, that is not too wordy?  Or is there a better way? Walkerma (talk) 06:28, 7 July 2008 (UTC)


 * I don't know if this helps but I've always seen it as ping pong:
 * B-Class - assessment by a single editor representing the project
 * GA - assessment by a single editor representing the community
 * A-Class - assesssment by a team of editors representing the project
 * FA - assessment by a team of editors representing the community
 * -- R OGER D AVIES  talk 06:36, 7 July 2008 (UTC)


 * That explanation is beautiful. --EncycloPetey (talk) 06:37, 7 July 2008 (UTC)


 * I agree - that's a fantastic way of putting it! <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 11:06, 7 July 2008 (UTC)


 * Nice. I'm pleased to see "GA" and "FA" being used without "-Class" appended. Something like this comment, together with assessment tables which separate the WikiProject and community assessments would go a long way towards addressing the confusion. Geometry guy 21:35, 7 July 2008 (UTC)


 * Mr Davies's description represents the simple way in which I have perceived the scheme so far. Or at least the upper half—I see the lower half as a three-level course of increasing completion of the basic structure and critical mass of information, up to an acceptable level of quality (B-class). After that, it mostly comes to filling in the corners and improving referencing, presentation, and consistency. It's like a foetus, really: after weeks of dramatic changes, it reaches a point after which it just grows in size, mass, and complexity. I think I've been taking some things for granted, considering that most people seem to view it differently; only recently have I come to realise how things really stand.
 * Anyway, back to the subject. I've had this little theory about article development, the "Jumping Theory". This theory says that, provided that there is a certain amount of effort driving the development of an article, one class in the assessment scale can be by-passed at any stage during the development process. This applies throughout the scale: an article can go, after an organised attempt for its improvement, from a stub to C-class, from GA to FA, or from B-class to A-class. In the lower stages the progress can be somewhat faster, but in the upper stages it is slower, as more factors come into play and greater perfection is required, so jumping two classes needs a significant effort, and is harder to achieve the higher one goes, especially for complex subjects. This is just theory, of course (and deals with the scale as a single entity, although the ping-pong approach can join the two two-class columns of the upper half), but it could offer some food for thought to those who appreciate the abstract. It also gives a straight "yes" to the introductory question of this thread.
 * (Question: is there an equivalent of the trout for non-admins?) Waltham, The Duke of 00:06, 8 July 2008 (UTC)

Assessment of articles not covered by a WikiProject
Sorry if this question is already answered in the documentation, but how do I request the assessment of an article that does not seem to fall within the mandate of any particular WikiProject? – SJL 04:09, 6 July 2008 (UTC)
 * Most assessments are done within projects, but GA and FA are assigned to any article that meets the requirements. My suggestion is to try to expand the article in an attempt to attain GA status. --EncycloPetey (talk) 06:34, 6 July 2008 (UTC)
 * Thanks for your response. I'm not sure what more could be added to the article, to be honest. It's about a professional organization, and I've already written about everything that it does. – SJL 07:25, 6 July 2008 (UTC)


 * Please read WP:GA? for the Good Article criteria. It does not yet meet it; for instance, references are unformatted; please take a look at WP:CITE/ES for that :) Gary <b style="color:#02b;">King</b> ( talk ) 07:28, 6 July 2008 (UTC)


 * Well, I may have just fixed the problem; I added two WikiProject banners which seemed to be appropriate. Pretty much every article fits under some Wikiproject. Drewcifer (talk) 07:35, 6 July 2008 (UTC)
 * Quite. If in doubt have a look in CAT:WPB (needs cleanup, but everything's in there). You can pretty much always find a wikiproject to go with any particular article. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 10:11, 6 July 2008 (UTC)

If you do get stuck, you can request such assessments from Wikipedia talk:Version 1.0 Editorial Team/Work via Wikiprojects, see Version_1.0_Editorial_Team/Work_via_Wikiprojects for more details. User:John Carter has done sterling work in this area in the past, but I can have a go myself (as can many others who watch this page) if you need it. Cheers, Walkerma (talk) 06:23, 7 July 2008 (UTC)
 * You might also look at International Association for Plant Taxonomy, an organization article that did attain GA status. --EncycloPetey (talk) 06:39, 7 July 2008 (UTC)

Thanks for your responses everyone, they are very helpful. To be honest, I was put off of Wikipedia a few months ago after an unpleasant experience (see this proposal and the discussion that lead to its failure), and I've only recently decided to take another crack at it (when I can find the time). Friendly people like yourselves make that seem like a good decision. – SJL 03:17, 8 July 2008 (UTC)

Typical article imagemap
Why does the "typical article" example used on this page completely omit GA? Wouldn't a "typical" example include GA? --EncycloPetey (talk) 05:32, 7 July 2008 (UTC)
 * Wow, that was fast! I was just going to ask for an opinion!  The article went B to A/GA to FA very quickly, and you could only see it with a mouse rollover.  I've amended it now to show GA as well. Walkerma (talk) 05:50, 7 July 2008 (UTC)
 * Thanks. As a general impression question, then, does the "typical" article go relatively rapidly from GA to FA, with little time spent as a GA?  --EncycloPetey (talk) 06:08, 7 July 2008 (UTC)
 * Well, of course there is no truly "typical" article at all - but I was impressed by this particular example (which I admit, I didn't choose), because IMHO it is far more typical than most. It also shows every step along the way, including a formal peer review.  I think many articles completely bypass A or GA on their way to FA, and some get "worked on" and go from Start to GA or A very quickly. Walkerma (talk) 06:17, 7 July 2008 (UTC)

WikiProject responses
This section is for WikiProjects to post their queries regarding the changes, in response to our postings on their discussion pages. Please do not change the header once the postings begin, as this section is directly linked from the message. Walkerma (talk) 18:44, 4 July 2008 (UTC)


 * This is a minor question about referencing. Is the occasional fact tag, such as on Utah State Route 13, enough to drop it down to C? The only missing source is a map from about 1940 that will show US-191 on SR-13. (There are several unreliable sources that report this, so it should be correct.) If the answer is yes, I can see arguments starting by someone tagging a statement and dropping the assessment to C. --NE2 05:53, 5 July 2008 (UTC)
 * I would say no, as referencing is something that comes to 'perfection' at the B/GA divide. I would say that the article is borderline with respect to B2 (breadth and depth of content), but hey, it's a road, what can you say? :D. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 09:08, 5 July 2008 (UTC)
 * Plus, reassessing the article just because someone added a cn is pretty cheeky.--Father Goose (talk) 09:12, 5 July 2008 (UTC)
 * This is a good point that we should clarify in the wording. I agree, it should stay at B-Class.  I'll ask around to get others' views.  Are there any ideas on how to word this?  Thanks for raising this. Walkerma (talk) 17:32, 5 July 2008 (UTC)
 * Thanks for your work here; I've really felt the need for something between Start and B-class in working on WikiProject Textile Arts. - PKM (talk) 17:31, 5 July 2008 (UTC)
 * The Military History WikiProject has officially decided not to adopt C-Class per this discussion. -MBK004 18:23, 5 July 2008 (UTC)
 * The VG project is still undecided. See this discusion for more info. Gears  of War Go 'Skins! 18:58, 5 July 2008 (UTC)
 * Actually, if you read that discussion, it was a no-consensus to opt-out. The C-Class has already been implemented, we just need to update our assessment scale.  JohnnyMrNinja  08:03, 6 July 2008 (UTC)
 * WikiProject Ohio has accepted C-Class and has already assessed over 50 articles as such. §hep   •   ¡Talk to me!  19:14, 5 July 2008 (UTC)
 * WikiProject Furry has implemented class C. I'm personally in support of this change - there's a need for a slot for articles which are beyond "Start", but which aren't developed to the standards necessary to B-class - which is, after all, meant to be just a short step away from being a Good article. We have several long articles that have interesting information but which would be a long way from being considered "good". GreenReaper (talk)
 * WikiProject Seinfeld is still undecided. Right now, there's hardly any activities on the article. I just need a message about the new concept and with Joelster already retired, I'm already at a loss. Gprince007 is someone to talk to because he knows more than me about editing. Anyway since someone brought it up, I'll wait for the result in at least until the olympic starts. So I could use a bit more info about it on my talk page. It will be helpful in the near future. Johnnyauau2000 (talk) 04:15, 6 July 2008 (UTC) It's been decided that me and Gprince007 disagree with the notion of the new system. We're trying to bring our project further up the level and that trend doesn't help at this time. I recommending double-checking the difference between the two systems. Johnnyauau2000 (talk) 12:05, 7 July 2008 (UTC)
 * WikiProject India Assessment Dept. will not be implementing any changes this year. It will however consider changes for next year - Ncmvocalist (talk) 20:11, 6 July 2008 (UTC)
 * Update:WP:VG has decided to except the new class. Gears of War Go 'Skins! 22:40, 6 July 2008 (UTC)
 * WP:Energy is accepting and implementing new assessment scheme. Please help to update the Template:WikiProject Energy—right now it doesn't include C-class feature.Beagel (talk) 17:57, 14 July 2008 (UTC)
 * The WikiProject Pinball has accepted the new class. Unfortunately we jumped the gun, and I tried to implement its inclusion I think before the bot was ready.  There are a few articles rated C-Class (such as Indiana Jones: The Pinball Adventure), but the bot doesn't seem to be picking them up so far.  If someone who knows better what they're doing can jump in here and fix or tell me what I did wrong I'd greatly appreciate it.  Thanks, Fractalchez (talk) 21:55, 14 July 2008 (UTC)
 * The banner had to be adjusted to include material for the new class. I've adjusted the banner and it should work now. John Carter (talk) 22:00, 14 July 2008 (UTC)
 * Wow, thank you for your fast help! Fractalchez (talk) 22:05, 14 July 2008 (UTC)


 * The 24 Wikiproject has decided to implement C Class, per this discussion. Steve Crossin   (contact)  03:05, 16 July 2008 (UTC)
 * WikiProject Oregon has accepted C-Class going forward: future assessments will consider C, but existing articles won't be reassessed only for this purpose.  (discussion)  —EncMstr (talk) 15:55, 17 July 2008 (UTC)

I'm against it
as a member of "Wiki project --US supreme court cases," I oppose this new system. It is mere "instruction creep" and unnecessary bureaucracy. In our Wiki project we only have a handful of articles anyway, why add extra assessment categories? JeanLatore (talk) 16:48, 5 July 2008 (UTC)


 * If it's not useful to you, ignore it.--Father Goose (talk) 06:25, 6 July 2008 (UTC)
 * Umm... Which thread are you responding to? Is this about category names or C-Class?  JohnnyMrNinja  06:33, 6 July 2008 (UTC)
 * I've moved this section to where it makes more sense, and where it appears to have been initially intended.  JohnnyMrNinja  04:53, 7 July 2008 (UTC)

WikiProject Spaceflight (and WikiProject Space Exploration) - request for help
On behalf of WikiProject Spaceflight (into which WikiProject Space Exploration has been merged) I would like to request help modifying Version 1.0 Editorial Team/Space exploration articles by quality statistics which is transcluded (template-like) onto project pages. Is there some way the table there can be modified so that it shows C-class articles? Thanks! (sdsds - talk) 20:49, 7 July 2008 (UTC)
 * It should show C-Class right now. Tito xd (?!? - cool stuff) 05:31, 8 July 2008 (UTC)
 * Cool -- thanks! (sdsds - talk) 05:34, 14 July 2008 (UTC)

Question re: list class
There's a good chance this is already tucked away in the archives somewhere but I couldn't find it with a quick glance, so I'll go ahead and ask.

Why is it that there exist, currently, 7 different classifications for articles (start, stub, C, B, A, GA, FL) yet there are only 2 (List, FL) for lists? I understand that there are significantly less lists (and probably less interest in them as opposed to standard articles) but a little more granularity in the assessment scale for lists might be helpful. Recently there was a big push to create the C class to help distinguish between stub class articles and B class articles (a distinction that I was not entirely convinced was necessary). Right now, there is a single class (list) which encompasses every list from the barest stub all the way on up to an list that falls just short of featured status.

This seems a little like a double standard to me. As I understood it one of the primary rationale for the creation of a C class was to assist editors in determining "how much" work an article created, with the claim that there was some ambiguity between a "good stub" and "minimal B". But again, as I mention, right now there is no distinguishing between a barebones list and "almost FL" status. Has there ever been any consideration between providing a bit more substance to the way lists are rated? <b style="color:#0000FF;">Sher</b><b style="color:#6060BF;">eth</b> 20:53, 7 July 2008 (UTC)


 * As far as I know you hit the nail on the head: not enough people work extensively with lists to make granularity necessary. It's more the case that List-Class was created as a 'dumping ground' to avoid the problem of having to mark Lists within a project scope as something else (B, Start, A, whatever). <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 21:09, 7 July 2008 (UTC)


 * This is all water under the bridge, but I would have used the same quality system (Stub, Start, C, B, Good, A, Featured) for lists as for articles. A category like "FL" mixes two pieces of information: "What kind?" and "How well?" But as lists are a much rarer breed, it is probably not worth introducing a change at this point, while people are still absorbing the impact of the new C-class. Marc Shepherd (talk) 22:01, 7 July 2008 (UTC)


 * That may be all academic. I've not worked on many lists, but when I did a couple of them recently, I looked around at existing lists for format ideas.  My impression was that lists fall mainly into three groups: Stubs, long works in progress, and featured lists.  It seems as if each particular list either has some gung-ho contributor creating a long and well-thought list, or has a team of occasional contributors resulting in "C"-quality work.
 * I'm also under the impression that there are relatively few lists compared to the number of articles, and that most projects pay little attention to them, so a complex rating system may just be more work. On the project I work with primarily (WP:PLANTS), lists receive very little attention, and aren't even counted in the assessment statistics. --EncycloPetey (talk) 22:39, 7 July 2008 (UTC)


 * Although it is true that fewer people work on lists than what would be necessary to justify and support a classification system, I believe the lack of such a system is more relevant to the scope of lists than to anything else. You see, an article can be as descriptive as one can make it: it can be a B-class article and still be very useful, even though it could be improved to GA or A-class. With lists, however, there is a very specific type, quantity, and quality of information included, and completeness is paramount, or the list is simply not very useful. Sure, an additional column could be added to a table to add one piece of information to each entry, but the margins for improvement are generally much stricter than in articles. EncycloPetey has made a good distinction: stubs (the standard example of "just a definition"), works in progress (information present, but not yet complete), and FL (complete, with a couple of extra improvements in presentation and breadth of information). It's hard to make any finer distinctions... Perhaps a pre-FL class which indicates completeness of data but lacking presentation? One would argue that these things grow simultaneously, however. Waltham, The Duke of 00:22, 8 July 2008 (UTC)


 * For the most part, "list class" would refer to articles that can never be more than a mere list or indexes (as opposed to a categories), for instance List of symphonies by name; refer to Categories, lists, and navigational templates in this regard; I would rather that lists which are seen as articles, e.g. List of Black Lagoon episodes follow the normal schema (Stub>Start>C>B...>FL) G.A.S 04:59, 8 July 2008 (UTC)


 * 1) It does not make sense to have a standard FL class and not a standard List class.
 * 2) It does not make sense to have a quality-class based on "type of content", it should be one "style of content".

List class should only be used for articles that cannot be reasonably rated on the regular scale. Separating out lists by the fact that they list things would be like having a Biography class. To be more specific, Characters of Final Fantasy VII is a list of characters, but should not be rated as List class because it is formatted as prose and can be easily graded on the normal scale. List of Square Enix games is a collection of wikilinks, and impossible to rate on the standard scale, without being completely arbitrary. List of Square Enix games could probably get a better rating than List of basic chemical engineering topics, but why? What criteria are there? Formatting / visual appeal? That is (to a degree) a matter of personal taste. Completeness? Editors could scour the internet for hours and possibly never find all of the items that should be in a particular list. And nobody should ever have to do that. I propose this for list class, and FL by proxy (rough wording, hope it is clear) -
 * A list class article is an article that is comprised almost entirely of links to other articles. Effectively, it is like a nicely formatted (and possibly annotated) category.
 * Prose is limited to introductory statements (explaining what should appear on the list or in that section).
 * If a list ever progresses past this, to having expanded prose descriptions of entries, it would the be assessable by quality, and should be graded on the standard article scale.

Thoughts?  JohnnyMrNinja  17:12, 8 July 2008 (UTC)


 * Which is exactly what I said, just in more detail. FL is a curious case, though, as many a FL is much more than just a simple list. G.A.S 17:22, 8 July 2008 (UTC)
 * Yes, sorry, we've been discussing this same thing at WT:VG. FL doesn't seem to have much more criteria right now than a featured article that, BTW, is a list.  JohnnyMrNinja  17:26, 8 July 2008 (UTC)

Should there be more examples?
I think that there should be two examples for each article class, and that there definitely should be at least two examples for the "Evolution of an article" section. What are other people's thoughts on this matter? Kevin Baastalk 16:00, 16 July 2008 (UTC)


 * In principle, agreed. But finding the examples are not quite so easy as it seems (trust me:) ) G.A.S 18:32, 16 July 2008 (UTC)


 * I have not had the time to pursue this, but have long felt that perhaps the easiest and best way to construct a really effective example would be to select an existing stub, and then incrementally construct an article, deliberately pausing at each level transition. Most good, experienced editors would not normally do this, of course, but would create an article that is as good as possible in the first editing pass. Nevertheless I can't think of a more dramatic way to highlight the differences between levels, especially since the editor would have these issues clearly in mind as each edit is completed. Moreover, this approach would improve the corpus of live articles, rather than again dealing with instruction and meta-issues – a frequent complaint in these discussions. Spinality (talk) 19:17, 16 July 2008 (UTC)


 * Why not just go through the history of a few FAs and just find points where they satisfied each step? Sure some of them didn't begin as stubs but I'm sure some of them did. Adam McCormick (talk) 20:13, 16 July 2008 (UTC)


 * I'm not sure who's actually going to read those examples. There are a lot of different paths an article may take on the road to FA. Also, the criteria have changed over time. There probably are no FA's that ever passed through C Class (since it's so new), or that traversed the path from Stub to FA when the current criteria were in place. You'd probably find a lot of FA's that wouldn't have passed the current "B" criteria when they were actually rated B. It's a better use of scarce resources to try to better define the criteria. Marc Shepherd (talk) 21:14, 16 July 2008 (UTC)


 * I mean to locate points in their history that match the current criteria, not that matched the criteria at the time. And I would add that a lot of newbies read the examples when trying to figure out what the rating really look like. Adam McCormick (talk) 21:39, 16 July 2008 (UTC)
 * Examples more inline with this would seem to be the idea. Adam McCormick (talk) 21:42, 16 July 2008 (UTC)
 * I tried to do so with recently featured articles (just before the implementation of C class. Not so easy, but if someone is willing to find them... G.A.S 05:00, 17 July 2008 (UTC)
 * My suggestion above – to do this deliberately – is because so many FAs will not necessarily go steadily through the defined steps. A good editor will often skip a level or two in a single extensive rewrite. However, if someone were to take a small topic steadily through each stage, delaying incorporation of in-line citations for example, using cleanup tags, etc., then a good instructive set of examples might be produced in a couple of days. Obviously, if one can find FAs that have gone through stages that match today's levels, then great. But I did spend some time looking, and couldn't find any that really illustrated the sequence. Spinality (talk) 21:52, 16 July 2008 (UTC)

(Unindent) My short answer to this question would be YES, YES, YES! A major goal of this assessment revamp was to develop a set of examples, not just one. I had a lengthy phone call with User:Holon, whose academic research has shown that the implementation of an assessment scheme depends a lot on how good the examples are (i.e., lousy examples => lousy scheme). We agreed that we could provide multiple examples for each level, even though we may only have one main example in the top-level description. You can see his thoughts here. We will probably want to have some examples that sit right in the middle of the "grade", and others that sit on the border between grades (worse than this, it's Start, better than this it's C, etc). We will also want to have examples from different topic types - a biography, a science topic, a railway station, a town, etc. - so that people writing a particular type of article can find a relevant example easily.

I'd like to ask the folks here - can you each come up with a set of examples (include the actual version) from your own main subject area, and list them below so we can mull these over? I'll add some chemical ones later today. Thanks, Walkerma (talk) 17:49, 17 July 2008 (UTC)

Examples of B-class review?
Hi. I was wondering if anyone could point me to an example of a B-class review that has been carried out using the new B-class criteria? I'm thinking of doing some B-class reviews and would like to see how this is done, e.g. is a template used or just free text, what sorts of comments are given etc. Also, is there any way to distinguish between articles rated as B-class under the old system and those reviewed under the new system? I guess a lot of articles will need to be re-reviewed as they may not meet the new B-class criteria and will be downgraded to C-class, although likewise some high-end Start-class articles may be upgraded to C-class. Thanks. --Jameboy (talk) 21:40, 20 July 2008 (UTC)
 * Some projects, such as Military history and Films, have criteria parameters which are simply assessed as "yes" or "no"; when an article can attain "yes" for all of the criteria, then it stands as a B-class article. There's no need to discuss or comment unless a disagreement exists. Girolamo Savonarola (talk) 01:31, 21 July 2008 (UTC)
 * FYI, the Military History folks have a nice FAQ page to explain the process in more detail. We'll probably adapt that for our own purposes in time! Walkerma (talk) 05:38, 21 July 2008 (UTC)

Missing footnotes
The transcluded Good article criteria includes inline footnotes, but they don't link to anything here because we don't have a reflist in the grading scheme template. I happen to think we're better off without the notes section anyway -- it would double the length of the already-long transcluded GA criteria -- but in that case, we should disable the footnotes. (This can be done by putting "noincludes" around them on the Good article criteria page.)

Thoughts?--Father Goose (talk) 22:52, 23 July 2008 (UTC)
 * Yes, I agree with you. Thanks for spotting that. Walkerma (talk) 02:40, 24 July 2008 (UTC)

"Policy violations such as trivia"
I'd like to suggest rewording this portion of the C-class criteria. Strictly speaking, trivia (that is to say, trivia sections) are not a policy violation but an MoS violation. I'm not sure a short trivia section would not be enough to disqualify an article from being B-class, though a long one would diminish the article's readability. (I take it they're not allowed at all in Good articles, given the "list incorporation" criterion.)--Father Goose (talk) 03:38, 14 July 2008 (UTC)


 * I agree it should not be phrased as a policy violation, but I think trivia sections should disqualify an article from B-class for those Wikiprojects that have adopted C-class. Phlegm Rooster (talk) 04:38, 14 July 2008 (UTC)
 * I would blame that one on a drafting error. It would be better to say, "Trivia, bias or other policy violations", IMO. Tito xd (?!? - cool stuff) 05:33, 14 July 2008 (UTC)
 * Na, trivia is not a violation of policy. Technically, even an GA can have the trivia tag without risk being delisted. <b style="color:#0000FF;">OhanaUnited</b><b style="color:green;">Talk page</b> 06:26, 14 July 2008 (UTC)
 * Perhaps when given a better sounding appellation, for example; Britney Spears. Still, trivia sections should be discouraged as much as practicable, in my opinion. Phlegm Rooster (talk) 08:19, 14 July 2008 (UTC)
 * Well, "cultural references" sections aren't necessarily trivia sections at all. Several FAs have such sections, though none have trivia sections, i.e., unsorted information.  The question is whether a couple of trivia bullets would be enough to disqualify an article from "B" status.  I'd be more inclined to tolerate a short section containing "miscellaneous information" in a movie article than in a biography, for instance -- though in both cases all items would have to be sourced.--Father Goose (talk) 20:33, 14 July 2008 (UTC)
 * But complete sourcing is a requirement for GA. Are we making that a requirement for "B" as well, or can a couple of unsourced statements still exist in a "B"?  I should think so, since that's part of what makes it a "B" rather than GA or "A".  A "B" shouldn't simply be An A/GA that hasn't been through a formal evaluation. --EncycloPetey (talk) 20:39, 14 July 2008 (UTC)
 * GA is quite a bit more than B, with extra criteria, all applied more stringently, via a formal process. As for trivia bullets, they tend to be isolated facts which might or might not be true, so I personally would require sourcing for each before signing off on a B rating.  But a short trivia list itself wouldn't necessarily deter me from giving an article a B.--Father Goose (talk) 02:24, 15 July 2008 (UTC)


 * (outdent)Why not just acknowledge that articles with "substantial prose issues" (such as trivia or other inappropriate formatting) cannot reach the classification? Seems the simplest solution to me. Girolamo Savonarola (talk) 02:55, 15 July 2008 (UTC)


 * I do acknowledge it, but don't consider a couple of trivia items "substantial prose issues". A long list, yes.  Just as an article missing a couple of citations could probably reach B, a couple of on-topic and sourced trivia items ought not be a barrier, IMO.  If the items can be integrated into the rest of the text readily, then they should be, but some pieces of information, even when highly relevant, don't have much relation to other parts of the article.  I'd hate to think that information would be thrown out in pursuit of a "simple solution" -- ease of integration is not a good yardstick for inclusion-worthiness.--Father Goose (talk) 08:10, 15 July 2008 (UTC)
 * And actually, I think other parts of the criteria cover the issue of overgrown trivia sections: "need editing for clarity, balance or flow", plus at the top, "contains a lot of irrelevant material". Trivia sections can range from a big mess to barely an issue -- things like neutrality and sourcing are much more of a sticking point.  So I'd like to petition that "policy violations such as bias or trivia" be changed to "policy violations such as bias or original research".  I'm not suggesting trivia should be given a free pass, just that it be addressed in proportion to the problems it causes.--Father Goose (talk) 08:33, 15 July 2008 (UTC)


 * Why even have a C-Class if there are no criteria? Trivia sections might not be a big deal to you, but they are anathema to many editors. Who's going to be heartbroken if their article doesn't make it to B-Class because of a trivia section? Not the kind of editors who create trivia sections and/or add to them; they are rarely the same editors involved in Wikiproject article improvement drives. Phlegm Rooster (talk) 03:33, 16 July 2008 (UTC)


 * Wow, way to translate a discussion of trivia sections into "having no criteria". What I'm trying to suggest is that while trivia sections are not without their faults, they probably shouldn't be put on a par with WP:NPOV as a "policy violation".  Are you even paying attention to what I'm saying?--Father Goose (talk) 08:57, 16 July 2008 (UTC)


 * Would phrasing it as "articles should not include sections that are extensive lists of trivia, or any such list where the information is not supported by a reliable source." This restricts the form and the content, but dose not preclude a couple of bit of sourced info in a section waiting to be integrated into the main article. --Nate1481(t/c) 09:13, 16 July 2008 (UTC)


 * In my first response on 14 July I said that trivia should not be phrased as a policy violation. Maybe I should get one of those colorful signatures so I'll stand out better. Anyway, there has to be a difference or two between B and C, or there is no point to having a C-class. Phlegm Rooster (talk) 11:46, 16 July 2008 (UTC)

← Let's start over: do we all agree that avoiding trivia sections is not policy and therefore the current phrasing is unsatisfactory? I think we do. The question is: "What do we change it to?" Personally, I feel that a blanket disqualification will not do, and that these things should be judged on a case-by-case basis. Trivia sections are often a problem, but their identification is not without its difficulties, and the information they contain could or could not be useful and with a prospect of future integration into a better version of the article's text. I certainly disagree with Phlegm Rooster's statement that not including trivia in the criteria takes away all difference between B- and C-class. There are much more important and easy-to-determine factors to consider than trivia: a relative lack of sources, a complete lack of in-line citations, an obvious bias, a one-sentence lead, few and/or unhelpful images, language that is too technical, blatant errors in grammar and spelling... Any of these can keep an article in C-class. We have criteria. Trivia sections, as I see it, have more to do with article structure, and they could or could not be a problem. In any case, these things could be more easily caught in GAN; we don't want to resemble GA criteria too much, after all. And there is something else: this B-class checklist is a template for the more specialised WikiProject ones. If a project wants to eradicate trivia lists of all kinds, they might put that in their customised checklist. Waltham, The Duke of 15:06, 16 July 2008 (UTC)
 * I agree that WP:Trivia is a style guideline. I agree that the wording should be changed to reflect that fact.
 * Everything is always decided on a case-by-case basis anyway.
 * I did not state that "not including trivia in the criteria takes away all difference between B- and C-class", I said it takes away one criterion.
 * I argued against creation of a C-class because it would lead to too much "navel-gazing", by which I meant fuzzy decision making. Now that there is a C-class, we should retain as many sharp distinctions as possible. Is it so much to ask that somebody prosify/integrate the information in a trivia section into the text?
 * I don't feel that my argument is to be taken as a crusade against "trivia". To me, it's purely a style issue; some items in a trivia section are more than trivia, and should be intergrated into the main text. Others are evidence that something has a resonance in popular culture, and should stay in prose form. From what I have seen, trivia sections attract good faith edits by newer/younger editors who are afraid to mess with the main text. They should not be discouraged, but their contributions need to be cleaned up.
 * Since assessment takes place behind the scenes, I feel that shouldn't be used as a cudgel to eliminate trivia or trivia sections, if that is anybody's concern. Assessment is for advanced editors, who should be on board with Consensus. If an article is has a trivia section, but the consensus is that it be assessed as B-class or higher, that is the way it should be. But the default should be C. Phlegm Rooster (talk) 23:17, 17 July 2008 (UTC)


 * How about this: changing "policy violations such as bias and trivia" to "policy violations, such as bias and original research", and strengthening the "article structure" requirements in the B criteria with something like "The article has a defined structure. Content should be organized into groups of related material, including a lead section and all the sections that can reasonably be included in an article of its kind."  That, in combination with "contains a lot of irrelevant material" from the C criteria and "suitable references" from B addresses the underlying problems with trivia sections without turning the presence of one into an automatic fail.  (This is consistent with the current B criteria: "The Manual of Style need not be followed rigorously".  Such rigor is, however, expected at the GA level: "it complies with the manual of style guidelines for... list incorporation".)


 * If it were a trivial matter (no pun intended) to integrate trivia items, then I would agree with you that it is not "too much to ask". However, having done trivia cleanup, I've found that there are frequently a few trivia items that do not integrate easily with existing content, and it would take a disproportionate amount of effort to build a section into which it would fit.  Given that the alternative is disproportionately easy -- simply deleting it -- but not an improvement to the article, I'd rather the lower grades outline the general principles of organization, sourcing, and so on without edging into the rigidity more characteristic of GA and above.--Father Goose (talk) 22:43, 23 July 2008 (UTC)


 * This seems to tackle the issue quite nicely. It strikes the right balance between hazy and stringent criteria, and also addresses various aspects of how assessment should be done. What do you think, Phlegm Rooster? (Judging from the timing of the last few posts, you have six days to reply. :-D) Waltham, The Duke of 10:35, 29 July 2008 (UTC)
 * I made a small change to the Editing suggestions column; "...the article checked for general compliance with the manual of style and related style guidelines." Phlegm Rooster (talk) 15:55, 29 July 2008 (UTC)
 * Works for me.--Father Goose (talk) 21:20, 29 July 2008 (UTC)

B-List A-List
i think B-List A-List classes should be added.  Here  Ford 21:42, 1 August 2008 (UTC)
 * Why? <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 21:51, 1 August 2008 (UTC)
 * Because theres nothing in-between a start class list and Featured List Class. Here  Ford 00:55, 2 August 2008 (UTC)
 * Sorry, when I asked 'why' I meant "why do you think we need more gradation between 'List' and 'FL'?" <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 18:37, 2 August 2008 (UTC)
 * You could always assess an A-List as A-Class, and that's what many projects do. Tito xd (?!? - cool stuff) 01:11, 2 August 2008 (UTC)
 * Actually there is no start class for lists, only a general class for all lists except for those that are of feature quality. I see no reason for multiple levels for lists. It is either simply a list, or a Featured List. We never call a list a stub simply due to its nature. It does not take that much to move a list to Feature status unlike that required by articles. See the following fairly short list that is a Featured list: List of schools in Marlborough, New Zealand Dbiel (Talk) 01:48, 2 August 2008 (UTC)
 * For the most part, "list class" would refer to articles that can never be more than a mere list or indexes (as opposed to a categories), for instance List of symphonies by name; refer to Categories, lists, and navigational templates in this regard. I would rather that lists which are seen as articles follow the normal schema (Stub>Start>C>B>GA>A>FL), i.e. list of episodes, with descriptions, etc. Regards, G.A.S 20:47, 2 August 2008 (UTC)

Template:Grading scheme
Please update Grading scheme with the non-standard classes per Version 1.0 Editorial Team/Assessment. The current content there is outdated. G.A.S 11:16, 11 July 2008 (UTC)
 * I think you're right, but I think we should check up exactly how & where this template is used. My personal rule with popular templates like this is "Don't be too bold" because I don't want 900 angry messages that I've messed up WikiProject's assessment pages!  There are probably some projects that don't like the nonstandard grades (we at WP:CHEMS have never used them - though we never adopted GA-Class either!).  We also need to decide if we should add in all or just some nonstandard grades, whether we want an explanation section or not, and if we want them in a separate box.  Personally, I'd say SOME (as per this page), NO and YES to those questions.  Any thoughts? Walkerma (talk) 15:05, 11 July 2008 (UTC)
 * The easiest is to determine where these parameters are in use, and continue from there. Add a hidden category to the template that trigger when the parameters are used to determine this (One each). G.A.S 15:10, 11 July 2008 (UTC)
 * Comment please? G.A.S 05:55, 18 July 2008 (UTC)
 * Can you try testing this idea out as you propose? I don't understand about hidden categories, I'm unsure how best to determine where the parameters are in use! Once we know this, and we can be sure that it's OK, we can probably go ahead.  Cheers, Walkerma (talk) 06:36, 18 July 2008 (UTC)
 * I have found the correct way at Help:Template. Maybe that will be easier. Unfortunately I know nothing about coding templates, so I cannot implement this. G.A.S 06:44, 18 July 2008 (UTC)
 * OK, I'll ask around for help with this. Thanks, Walkerma (talk) 11:42, 18 July 2008 (UTC)
 * That method seems like overkill. We could simply pass a parameter to the template that will append a particular row to the bottom of the template if the parameter is true (e.g. <tt>{&#123;grading scheme | list=yes}}</tt>) and go from there. Tito xd (?!? - cool stuff) 07:55, 31 July 2008 (UTC)
 * Currently the grading template uses such a parameter, but the results is outdated, for instance, see "List class" on WP:ANIME/ASSESS, and compare it to WP:ASSESS. We need to know where  before considering updating it, per the above discussion. G.A.S 08:08, 31 July 2008 (UTC)
 * The templates there are almost impossible to understand, as they're a horrible mess of conditionals, parameters, and parser functions. Instead, I'll just add hidden categories to take advantage of the job queue, as that seems simpler. Tito xd (?!? - cool stuff) 08:58, 31 July 2008 (UTC)
 * Cool! G.A.S 09:03, 31 July 2008 (UTC)


 * See Category:WikiProjects using Portal-Class in their grading scheme‎, Category:WikiProjects using Dab-Class in their grading scheme‎, Category:WikiProjects using NA-Class in their grading scheme‎, Category:WikiProjects using Cat-Class in their grading scheme‎, Category:WikiProjects using Template-Class in their grading scheme‎ and Category:WikiProjects using List-Class in their grading scheme‎, in no particular order. Tito xd (?!? - cool stuff) 09:16, 31 July 2008 (UTC)
 * It seems we'd be affecting at most 20 pages. I'd say we're safe making the changes. Tito xd (?!? - cool stuff) 06:15, 4 August 2008 (UTC)

Assessment Chart
How do i get this chart to include the classes: list, features list, category, and template. They have all been added to the project, and have articles classified as such, but don't appear on the chart. Also, when does it update? Grk1011 (talk) 20:13, 3 August 2008 (UTC)
 * The WP 1.0 bot, which maintains those tables, does not recognise Category-Class or Template-Class pages. It should, however, recognise List-Class and FL-Class. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 20:59, 3 August 2008 (UTC)
 * I think the problem might be that it hasn't updated since those inputs went into effect. When will it update the chart again? Grk1011 (talk) 21:04, 3 August 2008 (UTC)
 * When the script for the 1.0 bot was originally written, it maintained a collection of just 37,000 articles. Now it collects data from 1.7 million pages, which means the script takes a very long time - several days, in fact - to complete a full circuit of the assessment system and start again.  Each project is usually updated every 3 to 4 days, so you'll have to wait until the next update. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 21:08, 3 August 2008 (UTC)


 * Actually, if you go to User:WP 1.0 bot, there is a link to "run the bot right away". That will let you update the tables without waiting for the next automatic update. I ran it on the Eurovision articles just now, so the table is updated now. &mdash; Carl (CBM · talk) 21:19, 3 August 2008 (UTC)
 * Thanks. Now, there is a discussion to rename the project. Will it be complicated to get the assessments straightened out if we do rename? Grk1011 (talk) 21:22, 3 August 2008 (UTC)
 * No, it isn't hard. You just need to repeat the steps you followed to set up the project in the first place - create the needed categories, and make the assessment template to use them. If you have trouble, ask me on my talk page. &mdash; Carl (CBM · talk) 21:24, 3 August 2008 (UTC)
 * I didn't set it up in the first place, but if the time comes, I will prob need your help, thanks. Grk1011 (talk) 21:28, 3 August 2008 (UTC)


 * Oh, i just checked, the chart does not include list and Fl, why are they missing? Grk1011 (talk) 21:30, 3 August 2008 (UTC)


 * They are missing because the corresponding categories are empty. Also, the bot only tracks the following assessments: FA, FL, A, GA, B, C, Start, Stub, List, Unassessed. This is why the template class does not appear in the table. &mdash; Carl (CBM · talk) 21:32, 3 August 2008 (UTC)

← If you look at The page is FL but its in the unassessed category. I think there is something wrong with the banner, but I don't know how to fix it. Grk1011 (talk) 21:34, 3 August 2008 (UTC)
 * I fixed the banner. You will need to wait for the software to update the contents of the FL category, then run the bot to update the table. The bot only looks at the contents of the category. &mdash; Carl (CBM · talk) 21:40, 3 August 2008 (UTC)
 * Thanks, Grk1011 (talk) 21:47, 3 August 2008 (UTC)

New color scheme proposal for quality and importance scales

 * Proposed color scheme
 * {|class="wikitable"

!colspan="4"|Quality !colspan="2" rowspan="2"|Importance !colspan="2"|Article classes !colspan="2"|Other classes !Class !Color !Class !Color !Import. !Color !style="background:#ff66ff"|FA !style="background:#cc6699"|Disambig !style="background:#ffccff"|Top !style="background:#ff66ff"|FL !style="background:#cc66cc"|Redirect !style="background:#ccccff"|High !style="background:#cc99ff"|A !style="background:#9966cc"|Category !style="background:#ccffff"|Mid !style="background:#9999ff"|GA !style="background:#6666cc"|Help !style="background:#ccffcc"|Low !style="background:#99ccff"|B !style="background:#6699cc"|Image !style="background:#ffffcc"|NA !style="background:#99ffff"|C !style="background:#66cccc"|MediaWiki !style="background:#ffcccc"|Unknown !style="background:#99ffcc"|List !style="background:#66cc99"|Portal !style="background:#99ff99"|Start !style="background:#66cc66"|Project !style="background:#ccff99"|Stub !style="background:#99cc66"|Template !style="background:#ffff99"|Current !style="background:#cccc66"|User !style="background:#ffcc99"|Future !style="background:#cc9966"|NA !style="background:#ff9999"|Needed !style="background:#cc6666"|Unassessed
 * colspan="2" rowspan="6"|
 * colspan="2" rowspan="6"|
 * }
 * Current color scheme (quality only)
 * This is taken from here.
 * {| class="wikitable"
 * {| class="wikitable"

! colspan = "6"|Article classes ! colspan = "6"|Other classes The current color scheme is really bad when all are seen side by side. Here is a new one that I would like to throw out here to see what everyone thinks. I hope that I got all of the quality types (there are at last count 22). I have added a possibility to expand even further. Please tell me what you think. - LA (T) 17:55, 2 August 2008 (UTC)
 * colspan = "3"
 * colspan = "3"
 * colspan = "3"
 * colspan = "3"
 * class
 * td
 * col
 * class
 * td
 * col
 * class
 * td
 * col
 * class
 * td
 * col
 * colspan = "3"
 * colspan = "3"
 * colspan = "3"
 * colspan = "3"
 * class
 * td
 * col
 * class
 * td
 * col
 * class
 * td
 * col
 * class
 * td
 * col
 * colspan = "3"
 * colspan = "3"
 * colspan = "3"
 * colspan = "3"
 * class
 * td
 * col
 * class
 * td
 * col
 * class
 * td
 * col
 * class
 * td
 * col
 * colspan = "3"
 * colspan = "3" class="assess-current " style="background: ; text-align: center; " | Current
 * colspan = "3"
 * colspan = "3"
 * class
 * td
 * col
 * class
 * td
 * col
 * class
 * td
 * col
 * class
 * td
 * col
 * colspan = "3"
 * colspan = "3"
 * colspan = "3"
 * colspan = "3"
 * class
 * td
 * col
 * class
 * td
 * col
 * class
 * td
 * col
 * class
 * td
 * col
 * colspan = "3"
 * colspan = "3"
 * colspan = "6" rowspan = "2"|
 * class
 * td
 * col
 * class
 * td
 * col
 * }
 * colspan = "3"
 * colspan = "3"
 * colspan = "6" rowspan = "2"|
 * class
 * td
 * col
 * class
 * td
 * col
 * }
 * col
 * }
 * }

Why not go for broke? I have now included a new color scheme for importance. I am a big fan of staying within the basic 256 colors. - LA (T) 18:19, 2 August 2008 (UTC)
 * I think your importance scale in particular loses a key attribute of the current scale: a continuous gradation from strong to weak in one colour, reflecting the parallel continuum from vital to inconsequential. What reasoning do you have for saying that "the current colour scheme is really bad"?? <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 18:39, 2 August 2008 (UTC)
 * My thoughts and color likes and dislikes...
 * Right now, when placed side by side, the colors are just jarring to my eyes. In some cases they just clash badly. Look how Category and Current clash with everything. List, Future, Disambig, and Image aren't much better. I have a personal color chart that I use when picking colors. I just picked a column and went with the colors in it to avoid clashing colors.
 * I see magenta as a soothing and good color while red is an irritating and bad color, so the importance scale goes from red (bad) to magenta (good), but when it comes to importance you may be right, it may be better to go with a gray scale (#000000 to #ffffff). (I don't like some shades of magenta.)
 * Project and Template colors are the same. (Plus I don't like that shade of pink.)
 * Not all of the colors are written in hexidecimal making templating a bit tricky when doing background colors. Also if one does try to put the # in front of the hexidecimal on a page just for the color, one would get an ordered list (# is the wikicode for ordered lists just like this one).
 * I am not a big fan of gray or white unless used in a gray scale.
 * Now, I know that all of this is just my personal preferences, but I had to ask. :) - LA (T) 19:34, 2 August 2008 (UTC)
 * Oh, and another thing, I went with bright colors for articles, and muted colors for non-articles. - LA (T) 19:44, 2 August 2008 (UTC)
 * Hmn, some of these are certainly valid complaints, although I don't think they justify throwing out the entire scale and starting again. I think colour is important and so the importance scale should probably not be greyscale; and remember that NA-Class is the only one to use white - Unassessed-Class and -Class are both transparent.  The colours of the non-1.0 classes (ie everything bar FA/FL/A/GA/B/C/Start/Stub/???) have never really been ratified or even centrally planned, so a redesign of these would certainly be a possibility.  I agree that the colour for Cat-Class is particularly jarring.  I also like the idea of using more muted, pastel colours for the non-article classes, although I don't like the 1-to-1 correlation between your colours - it might give the impression, for instance, that there is some connection between Portal pages and Lists, or between Category pages and A-Class articles.  A co-ordinated, subtle, scheme for non-article pages would be a nice development, however.  Why don't we concentrate on that area? Finally, I'm now in the process of adding CSS classes to all the  templates, so you can now override the default colours with your own preferences by adding code like the following to your monobook.css:


 * So if you have personal preferences or dislikes for particular colours, you can set whatever schemes you want for your own viewing. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 20:28, 2 August 2008 (UTC)
 * Could we go with the colors I came up with for that non-article pages? It is the non-article quality classes which need the most help at the moment anyway. The rest can wait until those are done, and if you would like I can go in and do some of the changes on those which are not protected or only semi-protected. I can change the color and add the classes, since I do know my way around html. I wasn't even thinking color correlation problem. - LA (T) 20:40, 2 August 2008 (UTC)
 * The problem is that they are all too similar. It's extremely hard to distinguish each one from another. I agree that the colours for non-article classifications need overhauling, but they should remain seperate colours to the article assessment colours. Besides, it's way too soon for something as radical as that. We've only just implemented the overhauled assessment scheme. Changing all the colours would be like starting a barnyard riot. -- .: Alex  :.  21:05, 2 August 2008 (UTC)
 * (ec)There's no rush or deadline, so no reason not to have a play around and come up with a colourscheme that pleases everyone. I think the colour correlation problem is serious enough to preclude the colourscheme you suggest above.  Why not try a collection that's a bit more varied from the article-class scale? I also think that the colours are generally quite dark - lighter shades would probably look better.  <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 21:08, 2 August 2008 (UTC)
 * The problem, as mentioned above, is that the colors in the proposed scheme are too similar. Currently, the quality assessment scale (FA-FL-A-GA-B-C-Start-Stub-Unassessed) uses a HSV scheme with constant saturation and values, just with different hues. The importance scale (Top-High-Mid-Low-None) uses a HSL scheme with constant hue and saturation, but with different lightness for each class.
 * So, if we ignore the non-standard classes, the scales are independent of each other, and work properly. The problem is that there is no real gradient between non-standard classes—we can't say that a template is "better" than a category or portal like we can say that A-Class is better than Start-Class, or that High-Importance is better than Low-Importance. We can't put unrelated classifications in a continuous color scale, as that would imply that they're related; so, I wouldn't support the proposed scale as is. I would prefer to have a non-standard scale that "gets out of the way" of the two standard scales. Tito xd (?!? - cool stuff) 05:54, 3 August 2008 (UTC)

I am going to try to answer all in one go, hence the outdenting.

The colors are so similar since I used the same column on the color palette I linked you to above. I will always use the same column, no matter which column I choose. I loathe colors that jump all over the place when dealing with similar items. And please stay away from gray; any hexidecimal color with the letters a, b, d, and e and the numbers 1, 2, 4, 5, 7, and 8; and any color that can't be typed with only three characters; and please, let's not use any of the names (they are evil). I also tried another group of colors, but the result was the same.

I seriously doubt the correlation would have been noticed had I done one long list instead of the groups side by side to save some from having to scroll. If any user is naive enough to believe there is a correlation between two extremely different classes, well, that we can not fix.

The colors I chose for the non-article items were done in alphabetical order by what the subject space was and started with magenta and worked my way to red in the column. (And HSV and HSL mean nothing to me, the article is a bit too dense for me at this time of the morning.) These were intentionally made dark, denoting their non-importance.
 * Article
 * Article-Disambiguation
 * Article-Redirect
 * Category
 * Help
 * Image
 * MediaWiki
 * Portal
 * Project (The only non-alphabetical one, since it points to Wikipedia space.)
 * Template
 * User
 * anything
 * NA (Can be anything.)
 * Unassessed (Can be anything.)

Also, there may be no rush or deadline, however, letting things drag on and on and on leads to nothing getting done and frazzled nerves. I would like to get things done as soon as possible so that I can move onto the next thing. I don't go away from a project until it is complete, or as complete as it can be. - LA (T) 07:34, 3 August 2008 (UTC)

If I had written the table as the one to the right, would you have seen any correlation? - LA (T) 07:40, 3 August 2008 (UTC)


 * Yes. -- .: Alex  :.  09:57, 3 August 2008 (UTC)


 * I would also have noticed the correlation quickly. To be perfectly blunt: you are extremely unlikely to gain support for changing the colourscheme of the article classes (FA/FL/A/GA/B/C/Start/Stub/?) or the importance scale (Top/High/Mid/Low/NA/?).  You are much more likely to make progress A) working only on the non-article classes, and B) actually listening to the comments and complaints that numerous editors are now throwing at your proposals.  When I say "there's no rush" I mainly mean that very rarely is the first idea presented the best: in 99% of situations, the ultimate best solution evolves from analysis and discussion of initial designs.
 * You seem to have very limiting personal preferences for colourschemes which I suspect the majority of wikipedia users do not share. Remember that you can now implement through CSS whatever colourscheme you personally prefer without any discussion; I think, however, that when making changes that will affect all users, you need to be a bit more open to compromise and allowing designs to evolve.  Now, three users have told you that your colourscheme is not practical, for a variety of reasons.  How do you propose to surmount those problems? <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 13:15, 3 August 2008 (UTC)


 * Actually, Happy-Melon, make these four. I too believe that it would be extremely impractical to change the colours for the long-standing and well-working article-quality and importance scales. On the other hand, I strongly suggest bringing some order to the non-article classes, and shall support an effort to do so.
 * In fact, I have a few ideas. Since the available colours are relatively few (and I am referring to colours easy to distinguish and sufficiently different from those already in use), I propose teaming up classes that have a relatively similar function (at least as far as the tagging WikiProjects are concerned), so that they can use similar, or in some case even identical, shades. Alternatively, some other pattern could still be based on such groups. Here's one idea of how such teams could be like—some of them already in existence, in a way:
 * Proper but unstable articles (Future, Current)
 * Navigational non-articles in the mainspace (Redirect, Disambiguation)
 * Content-organisation namespaces (Portal, Category)
 * Content-editing-related namespaces (Template, Project)
 * I'd put Image and List separately; I shouldn't change Needed and the "No assessment" classes at all, considering grey and white quite fitting for their respective purposes. The new classes proposed by Lady Aleena, namely Help, MediaWiki, and User, refer to namespaces which cannot be under the scope of any WikiProjects, and are therefore unnecessary. Waltham, The Duke of 21:00, 3 August 2008 (UTC)


 * My thoughts on what has been said above:
 * Both the current article assessment colours (including those for "Future" and "Current") and importance colours are perfectly fine as they are, IMO. Let's not try and fix something that isn't broken; the colours are well established, and changing them just for the sake of it is only likely to ruffle some feathers. That said, I do feel that the new colour for B-Class is too similar to that used for GA-Class (and I've seen the same complaint made by others).
 * On the other hand, I think we do need to re-evaluate the colours used for the non-article classes (and have said so elsewhere previously). Those used at present are a bit of a mish mash, and clash with others, Category/Start and Project/Template to name a few. A proposal has already been made here to change the colour of Portal-Class, and while I don't object, I do think we would be better doing the whole lot of them in one go.
 * Regarding the colours proposed above, I share the same concerns as those made by others. I won't go into it further, for the sake of not repeating what has already been said.
 * There are more classes than those listed above. I re-shuffled Grading scheme/doc/see also last night to include Merge-Class, but later found more. These include Bplus-Class, AfD-Class, MergeDel-Class, Merged-Class, Display-Class (this one is unused, so I've TfD'd it), Structure-Class, and WP-Class (a duplicate of Project-Class, which I've redirected). It's possible there are others, so if we are going to reassign colours, it would be helpful to know just how many we're dealing with. I also agree with the comment above that adding new classes for Help, MediWiki and User is unnecessary. PC78 (talk) 23:48, 3 August 2008 (UTC)
 * Another thing that might be worth asking, though, is what purpose are all of those extra classes actually serving, and is it possible to merge or delete some of them. It seems to be getting too detailed (all of the varieties of merge-involved classes, to name just a few). Girolamo Savonarola (talk) 00:22, 4 August 2008 (UTC)
 * Personally I would get rid of the lot and assess anything that isn't an article as "NA-Class", though I suspect that this is very much a minority opinion. :) None of these extra classes really have anything to do with article assessment, after all. But I think it's generally left to individual WikiProjects to use whatever grades they feel useful. PC78 (talk) 00:44, 4 August 2008 (UTC)
 * I have already conceded on the issue of both importance and article class colors. The only colors I would like to see changed are the ones for non-articles.
 * Yes, I have a narrow color palette, but I am trying desperately to make the colors easy to memorize. 12 3-digit number-letter combinations (with a limited choice) are easier to remember then 12 6-digit number-letter combinations especially if they go from one side of the spectrum to the other. There are so many colors, that if they are way too varied, then they would be hard to memorize. Sure, I have my land-line phone, cell phone, social security, and credit card numbers all memorized; but I really would like to have an easy to memorize list of color codes for at least the non-article classes. The basic colors are #f00, #ff0, #0f0, #0ff, #00f, and #f0f. Change the 0s to "c"s and you get the really light colors or change the "f"s to 3s and you get the really dark colors.
 * So everything from Needed-Class and up, I will grin and bear the colors, but from Disambig-Class down, I would still like the color scheme above or one very close to it. Though if Needed-Class can be anything, I would like to see it worked into the non-article classes color scheme.
 * I would really love to see other suggested color schemes. Please share. - LA (T) 08:04, 4 August 2008 (UTC)


 * I don't think it is any one's purpose to actually memorise the colours' codes. That's the purpose of using colours, after all: to provide a visual clue. Waltham, The Duke of 15:46, 4 August 2008 (UTC)

Assessment over time
I've just used a script to extract the numbers from every revision of WP:1.0/S, giving us a nice set of data tracking how the assessment project has evolved over time. I've got a couple of nice graphs of the results: This graph shows how the number of assessed articles (and unassessed articles after they started to be tracked in December 2006) has grown over time. It would be nice to be able to plot the total number of articles on the same graph for comparison. However, much more interesting is the second graph:



This graph shows how the percentage of all articles in each class varies over time. The late addition of the Unassessed articles rather screws with the data - the dotted lines show how the graph would continue if unassessed articles had never been included. It's a little depressing to see that for a few months, over 70% of our assessed articles were stubs - fortunately that figure is now steady at about 68%. In fact it's quite surprising that ''all' of the figures are quite stable.

These data are really just food for thought, but they do throw a few ideas: if we were going to devote our efforts to just one task, I would suggest working on cutting that huge percentage of unassessed articles. Bored? Find a random subcat of Category:Unassessed articles and get evaluating! <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 13:47, 4 August 2008 (UTC)
 * I'm pleased to announce that WikiProject Video games has no unassessed articles! it happened about a week ago I think, when some users assessed them all. The graphs also show the growth of subject-related articles I suppose. You might want to fix that link though... -- .: Alex  :.  14:02, 4 August 2008 (UTC)
 * This is excellent, thanks! I would draw a different conclusion, though; I think a big reason for things being fairly static is that often an article, once assessed, remains "stuck" because there are fewer people working on re-assessing articles (which is less interesting anyway).  This was our problem at WP:CHEMS - our initial enthusiasm for assessment got us to assess all of our core articles, but we have shown much less enthusiasm for re-assessing.  I think quite a few articles haven't been re-assessed since 2005, and yet many of them have improved since then.  So, Alex (and also myself!), we still have (re)assessment work to do!


 * Another thing - we have around 2.5 million articles on en:WP, but only 1.4 million assessed. Of the missing 1.1 million, I would predict that nearly all are Start or Stub (since they tend to be the more obscure topics that remain untagged), and as these come into the system they "bring down" the % values in the second graph.  Fortunately we have a steady stream of Starts going to Cs going to Bs to help counter that, and overall things are roughly balanced.  I think that if we had all articles assessed, you would see a slight but steady improvement of article quality over time.  Of course, all this is just my speculation! Thanks again for some great stats & graphics! Walkerma (talk) 15:31, 4 August 2008 (UTC)


 * Looking again at the graphs, it would really interesting to see what a graph of JUST B/A/GA/FA would look like, so we can see the detail on this important subset. Also I meant to mention that if you look at the French project listing, their bot is set up to show what % of articles each project has assessed with a very nice graphic - I can imagine that many projects strive to reach that magic 100% that Video games has reached.  (Alex, fr:Projet:Jeu vidéo has only made 92%!).  Walkerma (talk) 15:38, 4 August 2008 (UTC)
 * Your wish is my command:


 * <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 15:59, 4 August 2008 (UTC)
 * This sounds eerily familiar to Image:WPTC assessment summary.png et al. :) It's also very interesting to see what the French project is doing: the percent completion stats are something that we should definitely consider doing ourselves, and they came up with a category intersection scheme based on the toolserver that people keep requesting, and that we were planning to do. Tito xd (?!? - cool stuff) 18:15, 4 August 2008 (UTC)

Future class and Current Class
Why are and {{Template:Current-Class)) included as valid assessments?
 * This template is used in assessment summaries for the WP1.0 project; see Wikipedia:Version 1.0 Editorial Team/Assessment.

See the following pages: I thought that they were considered invalid classes
 * Template:Current-Class
 * Template:Future-Class
 * Template:Cat class/doc
 * Template:Cat class

The following are some articles using these classes:
 * Talk:Chinese Democracy
 * Talk:John Carter of Mars (film)
 * Talk:Next United Kingdom general election

Projects using one or both of the classes
 * WikiProject_Films/Assessment
 * Exclusive class for this project only are:
 * {{Film|class=Future}} - For Future films

I have been working on this edit for several hours now and haved developed a better understanding of the use of classes in the process. I have edited Template:Current-Class and Template:Future-Class to reflection their non standard use. I believe that Template:Cat class/doc needs to be edited as well, but I am not sure how best to do it. The page currently implies that they are approved for general used by all projects and endorsed by this project, which I do not believe to be the case. I hope I have not made this post impossible to understand. Any feed back? would be appreciated.

Well my edit to Template:Current-Class was a failure and had to be reverted Dbiel {{sup|(Talk)}} 20:11, 2 August 2008 (UTC)
 * The classes are not invalid per se; we just don't use them Wikipedia-wide, but we don't discourage their use. WP:WPTC uses these scales also, for pages like 2008 Atlantic hurricane season (ongoing, hence {{tl|Current-Class}}) and Post-2008 Atlantic hurricane seasons) (hasn't occurred yet, so {{tl|Future-Class}}). As for the text present in the pages: That was just boilerplate copied from older versions of templates such as {{tl|A-Class}}. The edits are fine. Tito xd {{sup|(?!? - cool stuff)}} 06:01, 3 August 2008 (UTC)


 * Let's not single out {{tl|Future-Class}} and {{tl|Current-Class}}; if you look at the documentation for {{tl|Disambig-Class}}, {{tl|Category-Class}}, {{tl|Template-Class}} et al, they all use the same boilerplate text. I also don't see a problem wit {{tl|Cat class/doc}}, though I assume you actually mean {{tl|Grading scheme/doc/see also}}: it's just a list of the various assessment grades used across Wikipedia, it doesn't really imply anything IMO. PC78 (talk) 22:50, 3 August 2008 (UTC)

Here's another question, then - is there a good reason why these classes cannot be included in the bot-generated tables? I only ask because at present, when articles are assessed as such, the bot notes them as "removed" and does not track them actively. This can create problems if the tag (or task force parameter) is mistakenly added/removed or if the page is renamed improperly, because we can't see it in the logs. (It would also be nice to have a full tally of the number of articles within these classes, although that is less important and of course can be checked manually through a look at the category itself.) Any thoughts? Girolamo Savonarola (talk) 21:53, 3 August 2008 (UTC)


 * I would assume they aren't included because they aren't widely accepted by WikiProjects; it's only a small minority of projects that actually use them, as opposed to List-Class. That said, I do think it would be useful if the bot did track them, since they relate to actual articles in the article space, though we don't need to trouble the bot with the whole gamut of non-article classes. PC78 (talk) 22:57, 3 August 2008 (UTC)


 * Yes. I should clarify that I wasn't implying any usage at non-article-space (where any assessment is by type and therefore is unlikely to need changing). Girolamo Savonarola (talk) 23:26, 3 August 2008 (UTC)
 * I would say they haven't been included because nobody has asked for these classes to be enabled WP1.0-wide, so we've never had the discussion of whether these would be useful. It really is trivial to add these classes, but we need to have more-or-less of a consensus to do so. Maybe we will end up waiting for this and doing it in v2 of the bot, but discussing it now wouldn't hurt. Tito xd {{sup|(?!? - cool stuff)}} 04:18, 5 August 2008 (UTC)
 * Is it really something that requires a wide concensus, though? The addition of these two classes would only be seen by the handfull of projects that use them. PC78 (talk) 10:19, 5 August 2008 (UTC)

Wow, I didn't know these classes exist, nor did I know Merge, Needed, and Redirect class. But what's the point of having Needed-class? If you insert it into a talk page, that means the article exists and doesn't need to be "created" <b style="color:#0000FF;">OhanaUnited</b><b style="color:green;">{{sup|Talk page}}</b> 04:22, 6 August 2008 (UTC)
 * I would assume it's for redirects that should be expanded into full articles. PC78 (talk) 10:19, 6 August 2008 (UTC)

Improvements to the WP 1.0 bot under discussion
Hi, all. As we know, the Version 1.0 Index stores assessments for over 1.7 million articles. Originally, the bot was designed to process about 10,000 articles; we never actually thought that 70% of Wikipedia was going to be covered under some sort of assessment. That has slowly caused the bot to take longer to run, as bot runs that used to last about four hours now take about four days. To make the bot more efficient, changes to the way the bot framework operates are being discussed, and simultaneously, we are discussing which features it might be worthwhile to add as we are recoding everything. We really would like to have your participation at User:WP 1.0 bot/Second generation and its talk page as we do this. Thanks, Tito xd (?!? - cool stuff) 16:44, 7 August 2008 (UTC)

C-Class support
When we deployed C-Class, we did a thorough job on the category structure and then sat back, patted ourselves on the back, and thought we'd pretty much finished. Not so, aparently. I just did a quick test (here) and have discovered that, of the first 100 WikiProject banners in CAT:WPB, a full 63 of them do not currently support C-Class (and one treated C as equivalent to category!!). I think I have found the reason why the number of C-Class articles is growing so slowly!

So I would like to propose that a priority of the 1.0 Assessment team be to clear the list of WikiProject banners that do not support C-Class. Every project banner should handle C-Class in some fashion, either treating it as C-Class, or as Start-Class if the project has opted out of the C grade. I have been working by converting the worst-coded cases to, which has (IMHO :D) very good support for C-Class. I'm planning to document a way to not support C-Class, so projects that have opted out can still be converted. Alternatively, of course, we could just add C-Class support to the existing code, although some of the code I see in these banners leaves me physically ill :D!

So is anyone else interested in working on CAT:WPB, or a more selective list of 'broken' banners, to help find our missing C-Class articles? I guarrantee we'll find quite a few! <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 19:22, 4 August 2008 (UTC)


 * I would think it best to first try and establish whether or not these projects have an interest in using C-Class, before adding it to their banners. PC78 (talk) 19:39, 4 August 2008 (UTC)
 * Throughout the C-Class discussions, we made the assumption that the grade would be 'opt-out': projects could elect to not use it, but the default position would be to incorporate it. So while it will be necessary to check that each project hasn't made a decision to opt-out, the majority of projects just haven't bothered to do anything about the new class.  Right now we have the untenable situation whereby marking an article as C-Class actually removes it from the assessed categories in most project banners.  If a project has opted out of C-Class, then setting C should be handled in the banner the same way as Start... but it being the equivalent of cheesecake is not acceptable :D <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 20:41, 4 August 2008 (UTC)
 * We should probably figure out why new projects are adding code without C-Class—they're probably copying code from an example that doesn't have the support built-in. Tito xd (?!? - cool stuff) 00:56, 5 August 2008 (UTC)
 * Yes, I think in most cases it's probably just an oversight, as here (thanks HappyMelon!). We should be careful to avoid adding C-Class to projects listed as opposing it, based on this feedback, though HappyMelon's idea of making it read as Start is a brilliant one.  The main ones that have refused to adopt it are Military History, India and US Supreme Court Cases.  At WP:Chemicals (ironically!), many of the assessment people opposed it so it's de facto unused (but then we've never even used GA-Class!); however, the parent WP:Chemistry DOES use C-Class.  I'm "technically challenged" when it comes to reading template code, but this is an important piece of work, and if there are any ways I can help please let me know. Walkerma (talk) 03:53, 5 August 2008 (UTC)
 * It may seem minor compared to the missing C-Class, but I've been ploughing through lots of project tags tonight, and I think it's something that is worth addressing at the same time as this C-Class issue. Some projects such as MILHIST and WP:Former Countries do a nice job of including the B-Class checklist in the banner (see the Former Countries banner here as an example).  IMHO, it would be nice if this were a standard feature in most banners; we can't force this onto projects, but could we at least add it into the meta banner? Walkerma (talk) 05:14, 5 August 2008 (UTC)
 * That would most definitely need an opt-out option. Some WikiProjects have a "forced" checklist that automatically reassesses articles tagged as B's as C-Class if they didn't pass one or more of the B criteria (I coded WP:ANIME and WP:WPTC's banners to do this, and WP:MILHIST has done it since time immemorial...) While showing the checklist is something more or less minor, adding the forced reassessment code is something that should be done either on an opt-in bases, or widely advertised before deployment. Tito xd (?!? - cool stuff) 06:16, 5 August 2008 (UTC)
 * It would be 'opt-in', at least to start with - I'll see what I can do. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 09:36, 5 August 2008 (UTC)
 * Not sure I like the idea of having C-Class read as Start-Class for those projects that have opted out (as suggested above); best leave it alone, IMO. While I'm thinking about it, a lot of Projects also don't seem to have FL-Class. PC78 (talk) 10:17, 5 August 2008 (UTC)
 * I disagree. Currently the code in many banners means that when an article is 'upgraded' from Start to C, it is actually lost from the assessment system entirely, and categorised as "unassessed".  This is untenable - all project banners need to do something sensible when given C. Those projects that have opted out of C-Class have essentially said "we don't think we need more gradation between Start and B": fine, for them, we can stick C- and Start- back together and not lose any sleep over it.  But leaving a 'black hole' in the middle of the assessment scale where projects can 'fall out' of the grading structure, is not tenable. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 10:30, 5 August 2008 (UTC)
 * If the project doesn't use the class, then the banner doesn't need to do anything; treating C-Class as Unassessed is perfectly fine, IMO. Not all C-Class articles are upgraded Start's - many are downgraded B's. There should be no "black hole" with regards to these projects. PC78 (talk) 10:44, 5 August 2008 (UTC)
 * &larr; Sorry, I'm not being clear. The majority of project banners, when given the parameter C, display "this article has been rated as C-Class on the assessment scale" but categorise into Category:Unassessed Foo articles. See User:Happy-melon/sandbox1 (or the permalink above) for examples: look at the huge number of "unassessed" categories at the bottom. This is the 'black hole' I'm talking about, because most assessors are not going to check that the banner is handling their change correctly: they have the right to assume that it's going to deal sensibly with the parameter, not tell them it's dealing with it but actually pretend it doesn't exist.  If it told them "this article has not yet been rated...", that would be a start, because it would cause them to go back and readjust, but that's still rather counterintuitive.  So we have to go through all WikiProject banners and adjust so that they are at least honest about what they're doing with C.  Why shouldn't we, at the same time, be making sure they're doing something useful with it? <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 11:39, 5 August 2008 (UTC)
 * Yes, we MUST deal with this "black hole". With projects that voted not to use C-Class, in almost all cases a Start-Class designation would be the "correct" one; these projects are typically ones that had the stricter B-Class definition already.  And it's disastrous if people are assessing & tagging articles as C-Class and the articles are disappearing into article purgatory, to be lost from the assessment-universe for years! Walkerma (talk) 12:59, 5 August 2008 (UTC)
 * Right, OK, and perhaps my initial concern was also unclear. Ensuring that WikiProjects have the necessary support for C-Class is something that should be addressed (and should have been taken care of before C-Class was implemented, IMO). Just make sure that a project hasn't opted out first. But I still don't see why these few (I assume) projects should be treating C-Class as Start; it shouldn't be displayed in their banners at all, and should therefore be treated as Unassessed. As I said, the "black hole" you refer to should not extend this far.
 * To repeat somethinhg I said above (in case it got lost in the discussion), a lot of WikiProjects also don't fully support FL or List-Class; perhaps this is something else you should look into at the same time. PC78 (talk) 13:30, 5 August 2008 (UTC)
 * Fortunately, the number of projects explicitly opting out of C-Class is small, so it's a much smaller issue than the big 'black hole' we need to deal with. I too am depressed by the lack of wide support for FL and List classes.  Conversion to WPBannerMeta would solve these problems too :D. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 14:22, 5 August 2008 (UTC)

Just thought I'd inform, since I can't seem to recall seeing anyone else announce it here: WP Films has also opted out of the C-class. The banner currently shows them as unassessed if the class is attempted, and I personally prefer this, since I'd rather not have code which says one thing ("class=C"), but then behaves another way (displaying a Start class). This is especially confusing for users who are trying to upgrade the article, as it appears to them that nothing has happened, and maybe it's a site, cache, or template problem in their eyes. The Film project monitors unassessed articles on a daily basis usually, so these are easy enough to spot and rectify.

Another point is that the number of WikiProjects that decline the class is less important than the number of articles within those projects which will subsequently not be assessed at C-Class. It seems that the larger the project is, the less likely they are to adopt the class. This makes sense, however, as they tend to have well-developed assessment guidelines which often go into greater specific detail than the site-wide scheme, and may be disinclined to re-vet thousands or tens of thousands of their Start and B class articles. Nonetheless, I think that this is something that needs to be openly acknowledged and considered, implications-wise, because if the larger projects are declining the class, they may be better indicators as to how to assess articles, especially in large numbers. Just a thought. Girolamo Savonarola (talk) 04:43, 6 August 2008 (UTC)


 * Or maybe, as you said, they're just disinclined to reassess a large number of articles. The other thing is that if their standards are already different from the general assessment scale (by being more specific or stringent or whatever), they're not going to be in a rush to change them to match the changes we made.  They might embrace C-class in time, and there's no harm in their ignoring it.--Father Goose (talk) 06:40, 6 August 2008 (UTC)
 * I note that WP:AVIATION, a project spanning 30,000 articles, uses the general criteria and has embraced C class.--Father Goose (talk) 06:46, 6 August 2008 (UTC)
 * As far as the "black hole" is concerned, how about adding an "undefined" class (and category) to the templates to catch typos along with C classes in projects that don't support C class? That way they could find the misrated articles easily.--Father Goose (talk) 06:54, 6 August 2008 (UTC)
 * So much extra work when the current system works rather efficiently! :) That would only make sense if there were a large quantity of unassessed articles; however, I would argue that if that were the case, then this really is not going to alleviate anything anyway. Keep it simple. Girolamo Savonarola (talk) 07:12, 6 August 2008 (UTC)
 * What I'm suggesting is adding an "undefined" class to the templates (just to the templates) separate from the "unassessed" class (which is used for articles with a null assessment in the template). "Undefined" would flag the erroneous assessments.  I don't see how it would be much extra work.--Father Goose (talk) 07:36, 6 August 2008 (UTC)
 * It's work because you have to implement it - that's a lot of template banners to edit, project assessment scales to revise, another category that needs regular checking, a new class template presumably, etc. Why create extra work for oneself? Girolamo Savonarola (talk) 07:53, 6 August 2008 (UTC)
 * Because we've already got the C class, and this is a way to fix some problems it has caused with some assessment templates. Better to fix the templates than beg off the work.  And it sounds like Happy-melon is willing to do the work.--Father Goose (talk) 08:46, 6 August 2008 (UTC)
 * I can't honestly see "erroneous" assessments being such a big problem; it would be a redundant feature, IMO, and what Happy-melon has already suggested will take care of the real problem. With regards to what you said above, WP:BIOGRAPHY (the largest WikiProject, I believe) also use C-Class, not because it was "embraced", simply because it was ushered in without any discussion from project members. I suspect C-Class acceptance has a lot to do with apathy. PC78 (talk) 10:16, 6 August 2008 (UTC)
 * I wouldn't say I'm "willing to do the work": there are over 850 banners in CAT:WPB, of which my tests suggest about 2/3, or 560, do not properly handle C-Class. I'm certainly not going to fix all of those myself, although I'll certainly be working on it. I really started this thread to request some help in dealing with that large number of templates.  I could fairly easily write a tutorial on "converting banners to WPBannerMeta" if that would help.  It's not particularly difficult if you know a reasonable amount of template code.
 * I'm also not keen on an "Unassigned-Class" - it's absolutely no better than treating them as Unassessed-Class or Start-Class, because it would still have to be implemented in each template; while we're doing that, we might as well be fixing the problem completely. If projects that have opted out of C-Class can set their banners to handle C as saying "this article is unassessed" rather than saying "article is C-Class" but categorising into Category:Unassessed Foo articles, then I'm happy with that outcome.  An HTML comment in the wikicode to note that the project has opted out wouldn't go amiss, either, or you might well get someone like me coming around 'fixing' your banner to a state you don't want it in!
 * For me, the priority is to get all or at least the majority of project banners working of a central codebase so we can make changes like this centrally in future, and not have to worry about editing hundreds of individual pages. Really this situation is only one step (a large one, but only one) above having to edit each individual talkpage. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 14:04, 6 August 2008 (UTC)
 * OK, I think we should follow Happy-melon's lead here. Should we sign up to do certain letters of the alphabet?  I suggest that we just leave alone the few projects that we know about that don't use C-Class (Military History, India and US Supreme Court Cases, Chemicals, Films), and they can resolve the issue with their own templates (they probably already have).  Happy-melon, is there a simple way to fix most of the faulty templates?  I don't want to force a project to use the meta template against their will, and they may have little custom features that would get lost. Some small projects may be happy to switch, but we'd need to ask them first.  But is there a simple way to edit the majority of the broken templates and get them working?  (BTW: I should say, personally I'm currently very busy off-wiki writing the selection criteria for articles, and the clock is ticking on that, so I can't help until later this month - but I will definitely help then if needed.) Let us know what we should do next, Happy-melon!  Walkerma (talk) 15:41, 6 August 2008 (UTC)

Well I guess the most important thing is to plug the 'black hole', so a quick (and easy) fix would be a good way to proceed. Most templates contain a block of code like this: ... |FA|Fa|fa= |A|a= |GA|Ga|ga= |B|b= |Start|start= |Stub|stub= ... Which is generated by. This generates a very small list of acceptable parameters that actually cause a categorisation; unlike the fairly loose code that controls the message that is displayed (to use my favourite example, passing undefined would usually result in the banner displaying "this article has been rated as Cheesecake-Class on the assessment scale"!!). To fix the 'black hole' we need to add the following line: |C|c= In between the lines for B- and Start-Class, producing: ... |FA|Fa|fa= |A|a= |GA|Ga|ga= |B|b= |C|c= |Start|start= |Stub|stub= ... This won't solve the other issues raised above, but it will fix the C-Class issues. I'll start compiling a list of templates that don't handle C-Class so we can focus our efforts. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 16:25, 6 August 2008 (UTC)

List now going up at User:Happy-melon/sandbox2. Grab a group of ~20 templates and go for it. I'll grab a few myself once I've finished the list and fixed the redlinks. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 16:32, 6 August 2008 (UTC)
 * Screw that, this is an ideal job for a bot. I'll finish the list and then get a bot-run done on it. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 17:44, 6 August 2008 (UTC)

One point that I find interesting about this discussion is that there are a lot of other classes that are not supported by some, or even the majority of projects but are supported by the primary templates (sub templates) used by the project banners. These class include: redirect, future, current, needed, portal, etc The following is a "complete" list as I understand it (possibly not complete)

It seems that most, if not all, project banners accept all of the above classes as valid input regardless of the project's acceptance of the class or not. If an invalid class is entered then it shows up outside of the banner as a red template link such as Template:Redirecte-Class td Sorry for rambling, but it seems to me that the bigger question is how should classes not supported by individual projects be handled. Remember that articles are at times accessed by individuals outside of the specific project. Just thought I would toss in my 2cents worth of confusion. Dbiel (Talk) 20:14, 11 August 2008 (UTC)
 * ! The metatemplate handles all of these gracefully: if a project doesn't use the extended scale, the grade can't be given. Further improvements to the handling are of course also much easier if we only have to modify one metatemplate to implement any changes.  If you want to help solve this problem, I recommend converting a few in CAT:WPB to use the template. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 11:45, 12 August 2008 (UTC)

NA-Class articles vs NA-Class pages vs Non-article pages
There seem to be three different name styles for NA categories, with projects having some or even all - eg just to start with, Abortion. Is NA-Class articles the current standard, and should pages and templates be changed to use it only, or are variations acceptable? cat class recently got support for Non-article pages. TRS-80 (talk) 16:12, 16 August 2008 (UTC)
 * If the goal is to eventually integrate these by default into WP 1.0 bot, they should all be "NA-Class". &mdash; Carl (CBM · talk) 16:19, 16 August 2008 (UTC)
 * Like many other features of the WikiProject system, there is a clear preference (for NA-Class) but nothing ubiquitous. NA-Class is preferred for both the 1.0 bot and .  <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 18:24, 16 August 2008 (UTC)

Default Opt-in to C-Class
What is going on with this bot adding C class to the templates? There was nothing said about projects having to opt out of this during the discussion. All the discussion led me to believe it would be up to the project to decide to implement C-Class. Just to remind you of one of Happy Melons many comments during the discussion;


 * There is absolutely no requirement for existing projects to adopt the C-Class if they think that the moderate effort required to reassess their Start- and B-Class articles does not yield sufficient benefits. This proposal is simply to give them the ability to use a C-Class within the 'official' 1.0 scale if they want to. The biggest change this proposal implements is to make using C-Class the default position for new WikiProjects, and there, none of these "extra effort/work/hassle" arguments apply.

The project I am involved with, WikiProject Electronics, has had no discussion on this and the assesment procedure does not include C.  Adding this to our template is a nonsense. I am adding our project to your opt out in the absence of a discussion and reverting your bots change.

This proposal actually ended up with more oppose votes than supports but still got implemented. Whatever the arguments, it certainly cannot be claimed that this has consensus. Trying to make projects default to this scheme is beginning to look like steamrollering by a minority.  Sp in ni ng  Spark  19:48, 10 August 2008 (UTC)


 * This is exactly the kind of thing I was fearful of when I posted my initial comment in this discussion. However, it's fair to point out that your project has C-Class in it's assessment scale, and that nowhere (at least that I can see) has there been any discussion there about opting out of C-Class - are you really in a position to make this decision on behalf of all your project members? I'm not thrilled about it either, but it is my understanding that WikiProjects (such as yours, and indeed, most others) who asssess articles under the WP:1.0 program get C-Class by default, they don't opt out by default. PC78 (talk) 20:14, 10 August 2008 (UTC)


 * I fail to see how the actions of myself or the 1.0 Editorial Team are in any way in contradiction to the ideas, which I still firmly hold, presented in that quote. When we deployed the C-Class infrastructure we endeavoured to make it as easy as possible for projects to "use a C-Class... if they want to" (I'm not entirely sure why a Category:C-Class electronic articles category wasn't created for you).  Upon looking closer we realised that we had overestimated the technical template-coding ability of the majority of wikiprojects, hence we have agreed to take further measures to make it even easier for WikiProjects to use C-Class if they want to.  While I second PC78's concerns that you appear to be speaking for an entire WikiProject without any prior discussion, if WikiProject Electronics does decide to opt-out of C-Class, then we will accept and support that decision.  Remember that the real 'decision' is made by the project members in its actions: whether or not it assesses articles as C-Class.  Everything else is just infrastructure, housekeeping to support that development.  I could just as easily (to use my favourite example) add code to every WikiProject banner to allow articles to be assessed as "Cheesecake-Class".  Even if I were not reverted and sent to the asylum, my alteration would be useless because no project is ever going to make an effort, concerted or spontaneous, to assess its articles as Cheesecake-Class.  But the proof of the pudding, when we see which projects have "decided" to use C-Class, is when we look in Category:C-Class Foo articles and we see articles in there.  This whole thread has been about articles not appearing as such when they have been assessed as C-Class, an unacceptable situation.  I suggest you read the rest of this thread, and consult with your WikiProject before making such sweeping statements. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 20:30, 10 August 2008 (UTC)


 * A few comments:
 * There was a clear majority of support votes for this proposal (152 to 107), as well as an overall consensus, as you can see here. We had considered several proposals for change, but this was the only one where there was a clear consensus for change.
 * Every project was informed of the change, and was given the chance to discuss the issue and to opt out. As PC78 explains, the default is now to include C-Class.
 * Thanks, Happy-Melon, for your hard work! Many, including myself, are pretty clueless when it comes to the more complicated templates, and I'm sure that many people are pleased to see their C-Class assessments finally showing up in the tables! Walkerma (talk) 00:59, 11 August 2008 (UTC)


 * The quote plainly says that the default will be C-Class for NEW projects and implies that EXISTING projects can positively choose to use C-Class if they wish. This was said not just this once during the discussion but many times.  I invite you to provide a diff from the original discussion where it is categorically stated this proposal implies a default for using C-Class for all projects.  I am certain that many people where persuaded to support the C-Class proposal on the basis that it would not be imposed on them.
 * Attacking me on the grounds that I do not have the support of my project is the last refuge of an argument that has no foundation and is verging on personal attack. I think I am entitled to speak for the project for two reasons.  Firstly, because I have carried out most of the recent project assesments I am the person most directly affected by this decision.  Far more importantly, the project has an assesment procedure and assesment criteria that most definitely do not include a C-Class.  Assessing to C-Class will make no sense whatsoever until the project criteria are rewritten to include C-Class.  Until the project chooses to do that the default no C-Class position pertains.  Rewriting the criteria and re-assessing the articles will involve us in additional work, something we were promised would not happen during the discussion.  Sp in ni  ng  Spark  19:25, 15 August 2008 (UTC)
 * You are missing the point completely. A project "using C-Class" does not occur when the category Category:C-Class Foo articles is created.  The tipping point does not occur when the code is added to the project banner to allow them to use C-Class if they so wish. It doesn't even occur when a project has a discussion on the issue and reaches a conclusion one way or another (although I note that after five days with my note on the project talk page there still hasn't been that discussion at WPElectronics).  A projects starts using C-Class when its members start assessing or re-assessing articles as C-Class; when they apply the grade to real articles.  The 1.0 team can, and has, done all of that preliminary work, because it makes absolutely no difference whatsoever to any project whether it has been done.  We cannot, and will not, ever try to force an editor or a project to change their processes to accomodate C-Class if they do not want to.  If you don't intend to use C-Class, why do you care whether your project banner can handle it? Why do you care whether the necessary category exists? If you don't want to use C-Class, as a project or as an individual editor, don't!  That is what we mean by "opt-in", "choose to use", "not compulsory", and all the other words-to-that-effect: if you don't want to adapt to use C-Class, we have no intention whatsoever of forcing you to do so.  We're just trying to make it as easy as possible for you to make that change, to minimise the work that you or your project has to do iff you/they start using C-Class.  I fail to understand how anything we've done could possibly be construed as "forcing" anything on anyone. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 19:51, 15 August 2008 (UTC)


 * No personal attack was made against you, and you're clutching at straws if you truly think that was the case. I count 78 members in your WikiProject. I see no evidence that you have raised the issue with these members, and I see no evidence that you have been placed in a position to speak for the project as a whole. Your project's quality scale includes C-Class. PC78 (talk) 19:56, 15 August 2008 (UTC)


 * It is in the quality scale only because the 1.0 Editorial team have changed the template. Nothing to do with any action on the part of the project.  Sp in ni  ng  Spark  20:15, 15 August 2008 (UTC)


 * Of course it is: that is the scale that your project has elected to use. PC78 (talk) 20:22, 15 August 2008 (UTC)


 * No one from the 1.0 Editorial Team has made a single edit to WikiProject Electronics/Assessment - all the content there, including its transcluded templates, has been selected by the WikiProject. If you disagree with the content of the  template, you should not be using it.  <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 18:28, 16 August 2008 (UTC)
 * I would encourage you to discuss the issue openly on the WikiProject. If the consensus is not to use C-Class, that's fine - just don't use it!  (And if possible tell us here, it helps us keep track).  And for an assessment table you can use something like this. If you need any help, let us know.  Thanks, Walkerma (talk) 02:56, 18 August 2008 (UTC)

Since people tend to speak-up when things don't go the way they want to and remain silent when everything works fine - it may be misleading from this discussion that the C-Class change has not been well accepted. So, just for the record, I would like to say that I (and I believe many other wikipedians like me) find the new C-class useful. I also believe this change has been adequately discussed and properly rolled-out. Hats off to the people who managed such a big "technical" cut-over.  Arman  ( Talk ) 06:17, 18 August 2008 (UTC)

C class category

 * The category for C-class science fiction articles exists, but it is not being auto-populated when i rate articles as C. They go to "unassessed" instead. How can this be fixed?Yobmod (talk) 21:03, 6 August 2008 (UTC)
 * I updated Science Fiction Project template, should work now. Ruslik (talk) 14:13, 19 August 2008 (UTC)


 * It does indeed, thanks! Now back to down-grading all those unearned Bs... ;-D. Yobmod (talk) 09:04, 26 August 2008 (UTC)

Can someone with the technical know-how have a look at Template:WikiProject Unionism - this similarly isn't supporting C-class yet. Timrollpickering (talk) 12:28, 27 August 2008 (UTC)
 * <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 12:49, 27 August 2008 (UTC)
 * Thanks! Timrollpickering (talk) 12:53, 27 August 2008 (UTC)

Stub to B-class
I believe a stub may only fill 1, 2 or even none of a B-class checklist, a start article 3 -4, a c-class 5, and obviously B-class should fulfill all. Who else agrees? Gaia Octavia Agrippa Talk | Sign 21:51, 28 August 2008 (UTC)


 * I think this is far too specific and inflexible for a project of Wikipedia's size and nature. You have to take into account things like partially meeting one or more criteria, unusual or difficult to classify articles, uneven article development, and a whole bunch of other factors when assessing an article. Human intuition with the aid of the B-class checklist works well enough to separate stub, start, C, and B without tying the checklist into a strictly quantified system at every level of assessment. Below B-class, assessment is meant to be fairly informal, and it's best to keep it that way. Pyrospirit  ( talk  ·  contribs ) 03:44, 29 August 2008 (UTC)


 * I agree with Pyrospirit; at WP:ANIME we use the B class checklist for B and C class articles (If all = y then B class, else C), but it is quite tedious to fill them in, assessing these articles takes about 5 minutes each, while stub/start takes about 30 seconds each; so in our case the distinction is whether (1) you add value by completing the checklist, i.e. telling the editor what should be looked at, and (2) whether it is worthwhile to do so by completing the checklist.
 * Secondly, in or case, a lot of stubs – and start class articles – does not fulfil any of the criteria (yes, they are that bad), while "high" C-class articles fails ~3 of them, but can be fixed with a minimum amount of work (e.g. by writing a proper lead, formatting the references, maybe expanding one section a bit).
 * Regards, G.A.S 05:02, 29 August 2008 (UTC)


 * I think I agree with GAS and Pyrospirit, but I also think that the proposal is OK as a rough rule-of-thumb within a small project. I'd oppose writing it up formally here, though, because on WP such things have a habit of becoming RULES and policy, which would be a bad thing. Walkerma (talk) 18:55, 29 August 2008 (UTC)

How many A-class assessments...
If an article is tagged by (say) three wikiprojects and is aiming for A-class, would it need three A-class assessments? Or if it passes one A-class assessment, can/should it be rated A-class by the other wikiprojects without further review? Thanks. --Jameboy (talk) 18:41, 29 August 2008 (UTC)
 * Great question! I'd say that it would need three separate assessments, because the different projects bring different expertise to the table.  However, I think we could say that an A-Class assessment by project X should be considered as an automatic NOMINATION for A-Class at projects Y and Z. Walkerma (talk) 18:57, 29 August 2008 (UTC)

There should be another label - advert
There are oh so many articles on wikipedia that are plain advertisments, with some info attached to it. The same way that some companies are shipping cans of food with a recipe and a nice photo as "serving recommandation", the exact same way these articles look like. All "good" companies erase the criticism and already have a team of followers who reinsert the questionable parts. Politicians are "cleaning up" their biographies, either them or their staff and barely a few things that are over-publicised still remain. IT and electronics firms put only the good reviews in their products' article. And so on. Probably you can get an idea after reviewing some articles in each mentioned category.

The point is: wikipedia needs that publicity sign, as this is starting to look more like Yellow Pages than an encyclopedia. 93.126.141.234 (talk) 12:57, 13 September 2008 (UTC)


 * Yes, indeed, can you help us deal with this? You should join WikiProject_Spam, I'm sure they would love to have another person helping them!  Thanks, Walkerma (talk) 15:41, 13 September 2008 (UTC)

c class
It is quite difficult on wiki to retrospectively enter a debate, which is absurd considering the fluid nature of the beast. So I am starting a new section simply to coment that the C class is a VERY GOOD IDEA and badly needed, particularly as ideas about how to grade the classes above seem to become steadily stricter and stricter. I would say, however, that FA GA and A remain confusing high standards with a poorly defined hierachy.The B level seems to have widely become a 'very good article ' standard, which meant the whole assessment scheme was making fine distinctions merely between the best, and being wholly unhelpfull about the big lump of articles below B.Sandpiper (talk) 08:31, 22 September 2008 (UTC)

A and B class
I'm sure this has been argued about, but honestly I can't really make out the difference between A-class 'Very useful to readers. A fairly complete treatment of the subject. A non-expert in the subject matter would typically find nothing wanting', and B-class 'No reader should be left wanting, although the content may not be complete enough to satisfy a serious student or researcher.' If anything, the B-class criteria seems slightly stricter, since it says 'no reader should be left wanting', whereas the A-class critera simply says 'a non-expert would find nothing lacking'. It seems to me a 'student or researcher' might very well already be the 'expert' implicitly differentiated in the A-grade description? This may be a situation where those who have haggled over the fine print understand the words, through familiarity, rather better than those of use who occasionally pop in and wonder how on earth to interpret such similar statements. Sandpiper (talk) 09:09, 22 September 2008 (UTC)

Completeness
I have big difficulties with assessing whether an article is complete. Experience suggests it is never complete, so reliance on this as a defining criteria is problematic. I am absolutely certain that impartial external assessors are normally very badly placed to determine article completeness, lacking detailed knowledge as they normally do. I acknowledge that pointing out the absence of qualified assessors is not necessarily helpful when the criteria is more or less sensible in itself, but it can make the system unworkable. Sandpiper (talk) 09:29, 22 September 2008 (UTC)
 * Speaking of chemistry, my own area, I would say that a complete article is one that a group of expert chemists could agree covered the main topics. For example, for sulfuric acid, you would have to discuss its physical & chemical properties (especially its acidity), its history, its uses, its methods of production and its hazards - and few chemists would see much to add or remove from that list.  With many B-Class articles, you will see good coverage of many such subtopics, but little or no coverage of one or two key areas; you might have 600 words about uses, but nothing on history, for example.  I would say that one of the best ways to judge this is to look at FAs on similar topics - for a film, look at a film FA etc., allowing for the fact that there may be aspects specific to that film that warrant special attention (for example, the special effects in Star Wars).
 * We were beefing up the assessment definitions over the summer, but now Version 0.7 has come along and we've got busy on that - but we need to finish the job we started! Walkerma (talk) 12:49, 22 September 2008 (UTC)
 * It may be easier if an article is one of a class, so the expected components can be identified, but I have seen assessments made by people who are dedicated assessors but which have missed what seem to me clear omissions. The difficulty is that the people who worked up the article are often the only experts interested in assessing it. Sandpiper (talk) 16:16, 23 September 2008 (UTC)

Question about reassesment
What is the process for requesting an article be re-evaluated after significant changes have been made? I was thinking of just removing the assessment grade from the talk page template so that it'd show up in the relevant "unassessed" category, but I wanted to make sure there wasn't a better way first. Kurt Weber ( Go Colts! ) 14:49, 24 September 2008 (UTC) (Reposted by Coren on behalf of Kmweber 17:36, 24 September 2008 (UTC))

Wikipedia:Version 1.0 Editorial Team/Chicago articles by quality log not running again
Version 1.0 Editorial Team/Chicago articles by quality log has not run in a week and Version 1.0 Editorial Team/WikiProject Illinois articles by quality log has not run in eight days.--TonyTheTiger (t/c/bio/WP:CHICAGO/WP:LOTM) 03:02, 25 September 2008 (UTC)


 * Thanks for pointing this out. I looked at some bot logs, and the issue is a combination of
 * An interface change by the mediawiki devs that broke the bot for a day
 * Ongoing server problems that cause the bot to quit in frustration
 * I have run the bot again by hand, hopefully the server problems are fixed enough for it to make some progress. &mdash; Carl (CBM · talk) 03:22, 25 September 2008 (UTC)

A class vs. GAN
I'm asking around to see if there is a certain core set of style guidelines pages that just about everyone thinks of as helpful and inoffensive, roughly along the lines of (1b) at WP:WIAGA: "". Is there anyone involved with A-class reviews who don't like any of those style guidelines? - Dan Dank55 (send/receive) 16:26, 30 September 2008 (UTC)
 * Have you seen what WP:MILHIST has done with regards to its A-class review system? I think this is what you're asking about: WikiProject Military history/Assessment/A-class FAQ. -MBK004 16:32, 30 September 2008 (UTC)
 * Yep, MILHIST has a great A-class process. You guys say: "in line with style guidelines", meaning it wouldn't take a lot of copyediting to comply with most of them, and in fact, most high-quality MILHIST articles I've seen are easily above and beyond GAN standards.  I'm wondering if there are some A-class review processes out there where they feel fairly strongly that the GAN style standards are not helpful. - Dan Dank55 (send/receive) 16:56, 30 September 2008 (UTC)
 * The starting point here is Version 1.0 Editorial Team/Assessment/A-Class criteria. The key difference between A-Class and GA is focus - content vs style. An A-Class article should be complete and comprehensive in terms of content, and one can forgive a few style problems; a GA-article has not necessarily had any review by a subject-expert, and so it might not be complete, but it's often held to higher standards on style issues.  Walkerma (talk) 20:45, 30 September 2008 (UTC)
 * Good reply. I've added that more or less verbatim to the Milhist A-Class FAQ. Feel free to tweak! -- R OGER D AVIES  talk 21:43, 30 September 2008 (UTC)

C-class help
Hi. We've begun assessing articles related to Wikipedia talk:WikiProject Crime and Criminal Biography and WikiProject Criminal Biography/Serial Killer task force, but apparently I'm clueless about how to revise our settings in order for the C assessments to show up in our summaries. Could someone help? Wildhartlivie (talk) 17:25, 5 October 2008 (UTC)
 * Should be fixed now. <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 17:41, 5 October 2008 (UTC)
 * Thanks! Wildhartlivie (talk) 18:54, 5 October 2008 (UTC)

Articles having two quality assessments
It makes no sense to me for an article to have two different quality assessments. If you look at Talk:My Sassy Girl it is rated as both 'C' and 'start'. This is one article so it should have a single quality rating! The problem here is that one project does not support using 'C' class ratings. Yet the other one does. To me this anomaly is stupid. If the article is correctly rated in the film template it shows up as unrated. That may be better then using two ratings. How is this project working to eliminate this issue? I know that in the most simple terms if all of the project templates used the basic coding templates this would not be a problem. But many project still use custom code to deal with the area of assessments. Vegaswikian (talk) 20:44, 18 October 2008 (UTC)


 * It happens for a few maritime history articles where WP:SHIPS rates it as "C" and WP:MILHIST rates it as "Start" (becuase they don't use C-class). However, this isn't going to change because MILHIST doesn't want C-class, and the 1.0 team can't force them to use C-class. But it's not too big of a deal, is it? &mdash; Ed 17   for President  <sub style="color:#2F4F4F;">Vote for Ed  20:50, 18 October 2008 (UTC)
 * I'm not sure how big of a deal it is. However it is clearly wrong to have two different quality ratings for an article.  If I run into a difference when I'm looking at the talk page, I set them to the one that is correct.  Maybe this is where the problem lies.  Why should all editors be required to know which projects 'support' a specific quality rating?  Given that many (most?) articles are supported by multiple projects, I don't see why if you are using a quality rating you are not using the same ones that everyone else is?  I understand that some articles might need to be reevaluated and the rating changed, but wrong ratings are minor in my mind to conflicting ratings. I realize that some of this was discussed in the past, but the confusion is now upon us.  BTW, I'm raising the question here because of a discussion about deleting a category.  The problem there as I do more digging is that the article above belongs to two WPs.  One supports C and the other does not.  The one that does not, categorizes the article into both parents.  So we have a breakdown in the templates.  Not sure how to address that without redoing the project templates. Vegaswikian (talk) 22:59, 18 October 2008 (UTC)


 * There's nothing at all wrong with separate projects having different assessment levels - not the least because articles which have overlapping scope between two projects may be adequate for the area one project covers, while being grossly underserved in the area the other one does. Assessment is a judgment of the article's quality in relation to the WikiProject's goals. Many of these will be common throughout all articles, such as adequate citations and factual accuracy. But some things will be isolated to a particular project's style guidelines and content requirements, and therefore it is not unreasonable to suppose that in some cases where multiple projects are being served, some will be less adequately covered than others. Girolamo Savonarola (talk) 00:12, 19 October 2008 (UTC)


 * Yes, I have to second GS's view here. Often I will find a B-Class article on a chem-related topic where the chemistry portion of the article is terrible or non-existent, but someone from the Physics/Medicine/Aromatherapy/Geology etc project won't be able to judge just how much chemistry "should" be there.  It will affect what is selected, but I think this will only be in an appropriate way, other than in cases where the project is not using C-Class.  Walkerma (talk) 01:05, 19 October 2008 (UTC)

Any tools?
I liked User:Outriggr's script because it allowed me to spend more time reading and thinking about the article and less on the tedious business of updating the templates. That script has been deleted; are there any other useful assessment tools out there?-- Kubigula (talk) 05:26, 25 October 2008 (UTC)
 * Go to User:Pyrospirit/scripts and select metadata.js. Walkerma (talk) 05:35, 25 October 2008 (UTC)
 * Actually, Outriggr had two scripts relating to article assessment: metadata.js and assessment.js. The former is the one that I now maintain and can be installed as a gadget; it is documented at User:Pyrospirit/metadata. The latter is the one that has been deleted, and I don't know of any existing replacement for it. The difference is that metadata.js displays information on the article's assessment, featured/good status, and active peer reviews, while assessment.js displayed similar (though more limited) information but also provided an interface for viewing and changing individual project assessments. Pyrospirit  ( talk  ·  contribs ) 00:21, 27 October 2008 (UTC)
 * Thanks for the info. I am a fan of the metadata script, but I was hoping for a replacement for the assessment script.  I had it just long enough to get used to using it.-- Kubigula (talk) 03:21, 27 October 2008 (UTC)

List Class
Shouldn't List articles be added to the Grading scheme? Simply south (talk) 20:22, 26 October 2008 (UTC)
 * It's already in there; see the template documentation. It doesn't show up by default, you have to use a parameter. PC78 (talk) 20:34, 26 October 2008 (UTC)


 * Why is it only optional though? It is a main article type. This is about grading articles mainly. The other things are not as important and have all got a prefix. Simply south (talk) 21:36, 26 October 2008 (UTC)


 * I don't know. I'm guessing it's because not all WikiProjects use List-Class, though the same could be said for FL-Class and C-Class. It would be an easy job to make it show by default, though. PC78 (talk) 21:39, 26 October 2008 (UTC)


 * List-Class is one of the extended options. The ones shown should be the defaults everyone uses, generally.  §hep   •   ¡Talk to me!  21:40, 26 October 2008 (UTC)


 * Should List-class be or beome one that users use generally? Simply south (talk) 02:05, 27 October 2008 (UTC)

Nope, many projects rate "lists" by the amount of prose they have. Little bit of prose = stub, some more = start and so on. §hep  •   ¡Talk to me!  21:26, 28 October 2008 (UTC)

C-class articles
Hi! We at WP:INDIA are debating the introduction of C-class articles for our assessment. Could someone involved in assessment please weigh in on Wikipedia talk:Noticeboard for India-related topics? We need to know:


 * 1) If the implementation of C-class has been effective in general?
 * 2) Does C-class articles involve more red tape?
 * 3) Are the lines of distinction between Start and C, and Start and B classes blurred?
 * 4) Does C-class complicate the assessment process?

Eagerly awaiting feedback. Thanks, =Nichalp   «Talk»=  09:09, 2 November 2008 (UTC)
 * Please see my question to User:G.A.S's reply here -- GP Pande   talk!  10:12, 3 November 2008 (UTC)
 * Now I've commented, sorry it took a while. Hopefully others can provide some further perspective. Walkerma (talk) 17:50, 3 November 2008 (UTC)

assessing quality of articles differently to the rating given by other wikiprojects?
Hi I'm in Wikiproject Alternative medicine and I wondered if you could give me the answer to this. I'm trying to put the wikiproject banner on all the category's talk pages and assess them for quality. If articles have been assessed for quality by other wikiprojects, can I give them a different rating or do I always give the same? The reason being that some I may disagree with, and also some such as Dualism (philosophy of mind) the philosophy coverage is great but I don't know if it explains the importance of the idea to alt med much, and so the coverage in our subject area isn't so good and would need a little more editing. Hope you can tell me the right thing to do.:) <b style="color:#FF8C00;">Sticky</b> <b style="color:#FF8C00;">Parkin</b> 18:14, 6 November 2008 (UTC)
 * If it covers philosophy well, but not alternative medicine there is no reason it shouldn't have two ratings. Classes can and in most cases are different project by project. That's atleast what I've come across. §hep   •   ¡Talk to me!  19:50, 6 November 2008 (UTC)
 * Very true, Shep! Also - it may have been a year or more since the other project assessed it, and so the other assessment may no longer be valid anyway. Thanks for doing the assessments, Walkerma (talk) 23:28, 6 November 2008 (UTC)
 * c.f. Wikipedia talk:Version 1.0 Editorial Team... Tito xd (?!? - cool stuff) 01:02, 19 November 2008 (UTC)

A-Class
Is GA-Class a prerequisite for A-Class? MrKIA11 (talk) 18:25, 18 November 2008 (UTC)


 * No, though an A-Class article should be able to attain GA status Class without any problems and be very close to FA. §hep   •   ¡Talk to me!  23:23, 18 November 2008 (UTC)


 * I think a WikiProject could require GA status as part of their A-Class standards if they wanted, but I don't know of any projects that have actually chosen to do so. Pyrospirit  ( talk  ·  contribs ) 23:50, 18 November 2008 (UTC)

New Users & Confusion
Does it say anywhere that WP is short for Wikipedia?

WP:GA (Wikipedia Good Article), WP:FA (Wikipedia Featured Article)... A short description after an abreviation or code would help me and maybe other newbies, too. I would list the range of ratings in the topic sentence. Maybe; WP rates articles A through C class, but has FA, FL, GA, stub, start and some non-standard classifications. :) Romanfall (talk) 03:12, 27 November 2008 (UTC)Romanfall
 * There you go: Namespace. I would recommend turning on navigation popups if you are not sure what abbreviations stands for (Preferences \ Gadgets); this would allow you to point to a link, and a popup window will appear which will resolve the redirect.) G.A.S talk 04:46, 27 November 2008 (UTC)

Script-assisted assessment, the new generation?
After Outriggr's script went byebye I started searching for something to replace it with. And I think I might have struck something at the French Wikipedia (Projet:Wikipédia 1.0). The whole thing is detailed under Évaluation and since I know we've worked with frwiki on other projects I was wondering if someone here could open up a discussion of getting this in action over here. For those of you like me, who don't have a grasp on the French language here's the site translated. (Google and Windows Live) Thoughts? §hep  •   ¡Talk to me!  01:23, 14 December 2008 (UTC)
 * Yes, excellent idea, I'll email User:Kelson who coordinates that project. I'm in regular contact with him, but I wasn't aware of that script.  Can you clarify for me - which specific features are you looking for in the script?  Thanks, Walkerma (talk) 02:32, 14 December 2008 (UTC)
 * Pretty much just what outriggr's did, modify assessments (class, importance) and add new banners with assessments. The ability to remove bad banners sounds interesting too.  As much functionality as it has on frwiki would be awesome, but whatever they can muster would be a great help.  §hep   •   ¡Talk to me!  02:34, 14 December 2008 (UTC)

Question
How do I request an article for assesment? I wish for Palm Lake Elementary School to be assesed.  Royal  Mate1  20:45, 14 December 2008 (UTC)
 * You don't need to ask anyone, you can do it yourself! Have a read of the assessment criteria at Template:Grading scheme and work out which class the article belongs in.  Then go to Talk:Palm Lake Elementary School, click "edit this page" and replace class with WhateverClassYouDecide, click save and you're done!  <b style="color:forestgreen;">Happy</b>‑<b style="color:darkorange;">melon</b> 20:52, 14 December 2008 (UTC)

A-Class template
Hi there, I figured this would be the place to ask. I'm wondering where the discussion for this modification to the A-Class template was at? Based on that change there's an editprotected request at B-Class and I couldn't find any discussions on the changes. I don't mind what the template looks like, just curious to read other's thoughts on the proposal. Thanks, §hep   •   ¡Talk to me!  20:17, 22 December 2008 (UTC)
 * After those it makes sense to add the icons to C, Start, and Stub as well? §hep   •   ¡Talk to me!  20:19, 22 December 2008 (UTC)

Article length and assessment
Following a discussion with Parsecboy over the assessment of the MV Pelikan article, the question of article length was brought up. Whilst a stub article may be of any length, there is no guidance given as to whether there is a minimum length for an article to make Start class. Is this something that needs to be stated in the criteria for the class, or is there no minimum lenght requirement as long as the article meets the basic criteria for Start class? Mjroots (talk) 06:00, 23 December 2008 (UTC)

A-class questions
1) "Good article status is not a requirement for A-Class." Why not? 2) "The article [has] been reviewed by impartial reviewers from a WikiProject, like military history" - this needs to be clarified. Mihist has an internal review process; I am not familiar with any other projects which do that. On the other hand, I am familiar with many instances when an editor just gives an article s/he likes A-class, without any visible involvement from a WikiProject. I think that no article should be given A-class without being first listed on a WikiProject discussion page (if a project has no dedicated subpage for A-class reviews, general talk notification will do). Thoughts? --Piotr Konieczny aka Prokonsul Piotrus 13:20, 27 June 2010 (UTC)
 * Because A-class is a WikiProject rating, and theoretically a WikiProject could use any criteria that they want to judge it. Therefore there is not necessarily much in common between GA- and A-class.
 * I am also not aware on other WikiProjects which have active A-class review processes though there may be a few semi-active ones (e.g. WP:WPM). Yes, I believe the current understanding is that A-class may not be awarded without some kind of WikiProject involvement. You may like to read the discussion on Wikipedia talk:WikiProject Council/Assessment working group/Archive 3.
 * &mdash; Martin (MSGJ · talk) 14:01, 27 June 2010 (UTC)


 * If the concerned project don't use A-Class like the WP:ANIME/ASSESS then revert and trout that editor. Other cases contact the concerned project Assessment department, requiring the article to be re-assessed properly. In the case that there is no Assessment department then drop a line in that project talk page. --KrebMarkt 19:55, 27 June 2010 (UTC)

Trying out a new assessment method
WikiProject United States Public Policy, as part of the effort to quantify the effect of Wikimedia's Public Policy Initiative as progresses over the next year and change, is trying out a new assessment method. We'd love some help from people on the 1.0 editorial team to improve it (and it might be something of interest to other WikiProjects as well). The banner for the WikiProject is now live using this optional rating method as an alternative to the standard system. See Talk:Immigration and Nationality Act of 1952 for an example.--Sross (Public Policy) (talk) 13:17, 6 July 2010 (UTC)

The discussion above about whether or not the assessment system should be more apparent to the reader brings up several issues that the Public Policy Initiative must deal with in order evaluate its impact in improving article quality. We would appreciate input from Wikipedians involved in that discussion to help out on the project. Part of the research directly addresses many of issues that some of you raised with regards to the current assessment system (inconsistency in lower class articles, differences between expert and wikipedian assessment, difficulty for new contributors to assess). The | Public Policy Initiative  is recruiting Wikipedians to assess article quality improvement. We are testing the metric for consistency and to see if there are differences between Wikipedian scores and subject matter expert scores. We are looking to identify the strengths and weakness of the current assessment system since we are using that system to evaluate article quality improvement through the project. This project will produce quantifiable results evaluating many of the issues you mention above, your insight is needed, so please join the assessment team, if interested check out WikiProject: U.S. Public Policy. Thanks! ARoth (Public Policy Initiative) (talk) 23:42, 7 September 2010 (UTC)


 * This looks needlessly complicated: six separate criteria, with four different ranges for scores.  It sounds like the kind of thing that people could waste a lot of time arguing over.  Compared to the current process -- I can reliably identify a page as a stub or start in two seconds -- I'd say that you're doing a lot of unnecessary work for no additional benefit.  WhatamIdoing (talk) 19:20, 8 September 2010 (UTC)
 * The standard assessment system is actually quite complicated itself, if you look at the documentation. It's tough for newcomers to learn, because there are detailed definitions of each class of articles.  The goal of this system is to unpack that, to make it a little more explicit what the requirements for the different classes are.  The extra detail will probably also lead to a little more consistency in assessments, since it reduces some of the "gut feeling" approach to rating articles that is prevalent in the standard system.  You're right that it's extra work without much benefit for start and stub articles.  Note, however, that's it's compatible with the standard system, so instead of rating a start or stub with the detailed method, one can always just add the standard class directly.--Sage Ross - Online Faciliator, Wikimedia Foundation (talk) 19:54, 8 September 2010 (UTC)
 * Thanks for raising this - I was hoping we could discuss this in detail. The "hardcore" 1.0 team people (and there aren't too many!) are busy with the Version 0.8 preparation at the moment - we should have a list or articles with "safe" revisionIDs by about tomorrow - and this is why I personally haven't promoted discussion on this so far.  Regarding the scheme, I agree somewhat with WhatamIdoing that simple is good - the current system became popular for that reason.  Also, I think that WP:USPP is an unusual WikiProject, where measuring article improvement is a much higher priority than it is for most WikiProjects.  We've had the current system analyzed by a university prof. who studies assessment schemes, and other than us adding some more exemplars (still need to do that!), he concluded that our system is pretty robust and achieves what it is supposed to.  HAVING SAID THAT, I am fascinated to see if this will take off; clearly there is one WikiProject that likes it.  The new scheme essentially quantifies what we already do but in a more rigorous way.  I think a critical feature is the compatibility with the existing system.  If it were widely adopted, less active projects could stick with the simple system, and obvious stubs/starts might end up being tagged without a lot of rigmarole; after all, the assessments mainly get important once you get to C and above.  I would note, though, that for B and A there is already a more detailed rubric that goes well beyond the "gut feeling" we use for lower classes.  Is this new scheme really SO different from what we do already at those higher levels?


 * You should be aware of an earlier attempt that was similar to this proposal - see Article assessment. However, the similarity with this new proposal is not as great as it may seem.  The old scheme had good intentions, and a rigorous approach to assessment - so why did it fail?  Personally, I believe it failed for three reasons - (1) The results of all the work didn't really go anywhere especially useful; (2) WikiProjects weren't involved, so it relied on "drive-by" visitors doing assessments, often on topics they had little interest/knowledge on; (3) the assessment process took some time (perhaps 5-10 minutes per article).  Although #3 is still relevant, 1 and 2 can be avoided in this new scenario.  Still - it's always good to learn from the mistakes of the past, so we don't repeat them!


 * My suggestion would be to wait and see how the new system continues to work at WP:USPP. If it falls out of use as people get busy and move onto new things, then clearly it won't scale well to the wider community.  If it persists, and grows, and perhaps a "neighbouring" WikiProject copies the idea too, we should have a formal discussion here.  We spent six months using the current WP:ASSESS scheme over at WP:CHEMS in mid-2005, and we found the system was popular with many WikiProject members, so when we started to use the same scheme here at WP:1, we already had an idea it would scale.  So I'd like to talk to WP:USPP people in January and see how things look then. MANY thanks for raising this.  Walkerma (talk) 21:49, 8 September 2010 (UTC)