Wikipedia talk:Defending article quality

Untitled
I think you overrate featured article status. I've seen them approved with less than a dozen votes, with the toughest filter they have to pass being inane 32k limit.--Silverback 08:12, August 8, 2005 (UTC)

Choice of words
I think you should consider rewording. "Defend the status quo" is one of those phrases that provokes a negative response regardless of meaning. No comment on the actual proposal at this time. -- Cyrius|&#9998; 09:21, 8 August 2005 (UTC)
 * Maybe you should call it Defending the Good Quality or something like that.--Exir KamalabadiCriticism is welcomed! 11:36, August 8, 2005 (UTC)

Unwiki
Wiki is all about change. This proposal directly contradicts WP:BOLD.  Grue  11:54, 8 August 2005 (UTC)
 * Well, every policy has it's flip side ( see WP:BOLD ). &mdash; Ambush Commander (Talk) 14:00, August 8, 2005 (UTC)

My thoughts:
The first thing that has to go is the title. Defend the status quo, for one thing, immediately evokes everything that is unwiki. Defend makes it sound like we're obligated to win, and status quo is a really broad term when you think about it. Possible renames could be Don't change things just for the sake of change or something. Which logically implies if someone changed something and you can find no reason for that change, revert the change and leave a comment on the talk page.

You should add links to related reading, like: Be bold in updating pages or How to write a great article. I couldn't find anymore, but those two sorta talk about this proposal.

Finally, on to the content. First paragraph implies attaining featured article status means that few edits become useful and that featured article status can be a meter for determining when to invoke this rule, I know some users have objected to that. I understand that you actually mean that a characteristic of some featured articles is that they do not have many useful edits, so for the sake of clarity, remove the repeated references to featured articles and instead focus on "changes that do not do good or bad".

By writing this, I've found my POV! I think that it's not right to single out any particular article and then look for grey edits. I think that we should look for useless edits on all articles, and use featured article status as a sort of guiding light to where we should look.

By the way, the diff you provided doesn't help much. There's a lot of change on the article.

Otherwise, great proposal, I like it, I already know cases in my experience where this might be applicable, and I'll support it if you make it a little better. Cheers. &mdash; Ambush Commander (Talk) 13:58, August 8, 2005 (UTC)

2p from me
There's nothing with this editing guideline per se. The only thing I worry about is that if someone is entrenched enough in Wikipedia to get to this guideline page, then they already when to edit and when not to edit, so the guideline is "obvious" to them and ends up seeming like instruction creep. (This is true of several other policy/guideline pages btw). Pcb21| Pete 14:36, 8 August 2005 (UTC)

What they said
Yeh, what they said.

When I saw the link to this page in Pump, I came tearing over here looking to axe it, spike it, chop it up, thread it on a stick and turn it over a slow fire until evenly burnt around all the crispy corners.

What I read was actually mildly sane and somewhat in line with thoughts I've had myself.

The name should be changed or the content moved, perhaps put into a larger (or smaller) context. Also, consider a rewrite of the text, too. I know that by this time it sounds as if your entire effort has been shit, but don't be that way. The idea is not all bad. Let me restate it a couple of different ways.


 * Let there be a presumption that, as total edits to an article approaches some number N, quality increases (along some unspecified function); from that point on, quality tends to level or even drop with further editing. This is the form relative to article inception.

Another possible interpretation of your thoughts:


 * Let there be a presumption that Featured Article status, by definition, represents (nearly) the highest-quality possible state of an article.

(Of course, simply the fact that I may have read the page and be able to draw two different distillations from the same cask suggests a rewrite.) In either case:


 * ...Therefore, when an article has passed this point, further editing should be subjected to greater scrutiny and editors should hold themselves to a higher standard, or abstain.

This last bit of semantic juggling is very hard to enforce. It suggests that early edits ought to be slapdash -- which may in fact be the case, but hardly looks good in print.

I agree that it should be more difficult to edit well-developed pages, but as a gradualist, I support a continuum of increasing difficulty. As a technocrat, I support hard coded solutions; as a humanist, I urge that all such systems be built to human scale, with human needs and capacities in the fore. &mdash; Xiong &#29066; talk * 14:40, 2005 August 8 (UTC)

Devil in the details
Hm. There's some good stuff to be picked out of this proposal -- starting with the improved name -- and I don't mean only words modeled on my comments, either. (Actually, those are the weakest on the page, being mere sketchy talk talk; if they're to remain, they should be rewritten to a higher standard.) But being a grumpy old man in a grumpy old mood, I'd rather find more stuff to bash.

Defending FA quality?
Does this concern only appear when an article passes FA? I'm a gradualist; I'm suspicious of human-oriented processes with sharp-edged boundaries. It's an analog world. I prefer (over the digital switching model) processes modeled on analog feedback (in the technical sense). To speak more plainly, I would prefer to see a proposal that ranked all articles along one or more continuous scales; articles further along some of these scales would require progressively greater "effort" (a completely undefined term here) to edit.

In this conception, arrival of an article at some designated benchmark along some of these scales would automatically confer FA status -- or at the very least, FAC status.

Who will bell the cat?
I favor hard-coded solutions. That means I dislike -- intensely -- drudgery such as RC patrol. I can't endorse yet more makework. Whenever possible, the machine should be made to do all routine.

This is (at least in theory) always an available method. For, given an action, there either does or does not exist explicit criteria. If these exist, the machine can be set to follow them. If they do not, and subjective judgement is required, then any process based on humans exercising it to determine when to take action will fail in a large community -- as we can no longer trust others to exercise good judgement in the absence of explict criteria.

Therefore, let George do it.

Bit rot concern?
The thesis must be substantiated that bit rot is a concern. This is the process wherein a code segment (or in this case, an article) does indeed degrade with time after some apex of quality is reached.

Although I support this thesis, it can be argued that bit rot is at best of secondary concern; viz: If a process exists to improve the text when it is below the apex, and bit rot acts to degrade the text past this point, then soon the text will again be below the apex and the first process will again operate. At worst, quality will oscillate between the apex and some threshold point immediately below.

This argument must be trounced soundly in order to justify a program intended to counteract bit rot.

&mdash; Xiong &#29066; talk * 08:10, 2005 August 14 (UTC)

Article Guardianship or Adopt an Article
It's an alternative to the Featured article patrol you propose at the end of the this proposal. Instead of formalizing the whole thing, encourage users to take articles "under their wing" and follow them closely. For controversial articles, this is a given, but for the more bland ones, they could do with a watchful eye upon them. No need to say you're patrolling that article. &mdash; Ambush Commander (Talk) 22:39, August 14, 2005 (UTC)

Instruction Creep
This proposed guideline is unnecessary instruction creep. Ideally, all edits should be subjected to scrutiny. This may not be possible in all cases, but it is already a goal of the wiki community. In all cases, edit summaries should explain what is being changed, and at least implicitly, why. People already understand this. If an edit does not improve the article, it should be reverted or changed so that it does. This too, is obvious to almost all editors. People should feel free to improve or attempt to improve all articles. That is the point of WP:BOLD. This guideline adds nothing to the above; it's only result would be to make people fearful of improving articles. I can never support that on a wiki. Superm401 | Talk 04:22, August 16, 2005 (UTC)

Isn't this done already?
I 'patrol' all FAs I contributed to. I'd have thought that most users would do the same. Thus only FAs created by missing Wikipedians are in real need of patrol.--Piotr Konieczny aka Prokonsul Piotrus Talk 02:13, 15 January 2006 (UTC)

Repairing the dead link for the Stvilia paper
Besiki Stvilia is now at Florida State University. His home page links to his "Information Quality Discussions in Wikipedia," and I believe this is the same paper as that with the dead link. Dgorsline (talk) 19:36, 14 January 2013 (UTC)

FA watchlist
could be mentioned in this essay. Toccata quarta (talk) 21:11, 23 January 2013 (UTC)