User:Retro/Templates/Article talk

My subpage for templates in.

( should probably soft redirect here.)

This category can include GA nomination templates, FAQ subpage templates, etc. Not all of these templates are header templates, so this category is probably a meaningful distinction.

This needs reorganizing, and possible rescoping; all of my GA-related additions have dilluted this page's original purpose. Perhaps many of these portions should be migrated to the parent page, and this page refocused to its original goal of article talk templates. After all, those comments are only steps on the journey to improving article talk templates. But some of the GA-related templates are extremely related to article talk, so maybe GA templates just need their entirely separate templates user subpage for organizing; I could then take advantage of main crosslinking between my user pages.

Article assessment classes
Importance class templates will probably be discussed here, but they are distinct, and should almost certainly remain two distinct parent templates.

Understanding Template:Class
I would like to understand Class entirely to see about eliminating the various "-Class" templates. I am motivated to do this because it takes the same amount of text to type class with a second parameter, and would be simpler to standardize on it. Splitting the "-Class" templates into separate pages just creates more independent code to maintain (though in this case, the code is probably rather straightforward).

At present, my understanding is that Class is dependent on these other pages. I would like to move the code from these subpages into in some manner, so the disparate pages become unnecessary. It could still retain a default setting to check for "-Class" pages for unrecognized names, so uncommon names could have support added, but having separate pages for the large majority of disparate templates currently seems unnecessary to me.

Template:Grading scheme/doc/table is a useful summary link of all the related templates. Other subpages of are also useful.

Even when Talk:Famine in India only transcludes, it still uses Unassessed class in the page preview. This is an oddity; I'm unsure why it would need to transclude Unassessed class, and this may lead to inflated transclusion numbers. I will want to look more closely at the inner workings of WikiProject India (and possibly WPBannerMeta).

Finding assessment class-related templates
I was searching for "class" in the template namespace and got many templates related to ships. These terms are commonly ship related: cruiser, frigate, destroyer, oiler, battleship, cruiser, submarine, gunvessel, ship, buoy tender, schooner, supertanker, battlecruiser, sloop, yacht, gunboat, trawler, helicopter destroyer, minesweeping sloop, escort carrier, monitor, torpedo boat, corvette, ironclads, ammunition ship, coastal buoy tender, brig-sloop, cargo ship, attack transport, tank landing ship

However, to properly exclude the search, one should probably include "class" before the target word in the title, since there may be assessments classes related to a specific WikiProject. Though WikiProject Ships is probably the most specific WikiProject, and anything more granular is a task force (I haven't investigated the WikiProject structure here closely; Template:WPSHIPS class and prefix templates is another interesting result.). I imagine my concern about this was unwarranted, so I can probably just use intitle more times, since I won't be carrying the repetitive "Class" prefix burden.


 * Original search: "class"
 * 3,233 (but now I'm getting 57,541; that's very curious. I wonder what I did different from these searches and the previous searches; it's not a namespace issue, but it appears to be related to /class-suffixed subpages. I can go back and look at my browser history to find out.
 * Revised search: "class" -intitle:"class submarine" -intitle:"class cruiser" -intitle:"class frigate" -intitle:"class destroyer"
 * 2,248
 * 3rd search: "class" -intitle:"class submarine" -intitle:"class cruiser" -intitle:"class frigate" -intitle:"class destroyer" -intitle:"class tanker" -intitle:"class battleship" *-intitle:"class corvette"

Other miscellaneous results: "first class cricket team", templates starting with "Template:GHSA Class"

Article assessment class miscellanea

 * {{code|hastemplate:"Unassessed class" insource:/\{\{Unassessed/}} (in all namespaces) turns up interesting results; only 2 results:
 * Template:WPBannerMeta/hooks/qualimpintersect
 * Template:WPBannerMeta/hooks/qualimpintersect/sandbox
 * Miscellaneous assessment class templates:
 * {{tl|-Class}}
 * {{tl|Chem A-Class}}, {{tl|Chem B-Class}}, {{tl|Chem FA}}, {{tl|Chem GA-Class}}, {{tl|Chem Start}}, {{tl|Chem Stub}}
 * {{tl|A-Class-Awake}}: Use?
 * {{tl|D&D/Unassessed-Class}}
 * {{tl|WikiProject Video games A-class assessment initialize}}
 * {{tl|WPBiography B-Class review}}, {{tl|WPFILM B-Class review}}, should search for "class review"

Talk header template priority
I think notices of previous deletion discussions should generally be at the very top of talk pages. That is where I expect to see them. I wonder how many are missing.

Like this: Talk:Justin Bieber on Twitter. Then again, here, it seems the edit request notice is more important. This same article is missing a deletion review from its history.


 * File talk:Pink Floyd - all members.jpg: All ffd-delete, afd-delete should be at the top (I wonder what the guideline/policy for this is.)

I need to get more empirical data here. I suspect for pages with more notices, it being at the top is less consistent, but if it's not, then that would make a stronger argument.

But this also brings up an interesting point: Article history also has an option for noting deletion discussions. I generally prefer a separate template, since I few a deletion discussion as important (but then again, I view GA and FA confirmation discussions as important as well, but they can be summarized better in . This leads into my next section:

Talk header template redundancy
It's very easy to become overwhelmed by the number of notices. This is especially common on large, contentious topics like Talk:Homeopathy. I think it's important to merge notices related to the articles history, and I will list such notices below. Perhaps not all such notices should be merged, but there are some obvious cases of redundancy that could be eliminated.

The category hierarchy in this area is as follows: Category:Talk header templates contains Category:Article talk header templates contains Category:Non-subject-matter-related article-talk header templates. I don't particularly prefer the name "Non-article-subject-matter-related article-talk", and would prefer "Article history talk" (as in Category:Article history talk header templates). This is not a clean rename; it might require splitting the original category up. Actually, I'm wondering if an inversion should be done; general templates noted in the "Article talk header templates", and more specific ones in a separate "Article talk header templates by subject" (Category:Article talk header templates by subject; need to do a title search for the phrase "by " to see how regular this title will be.) Category:Wikipedia talk page templates, Category:Talk page templates might be good soft redirects somewhere. Category:Talk templates should already exist.

As mentioned previously, some of the templates I'm interested in combining (at minimum in the sense of eliminating multiple ones on the same talk page) could have their own category Category:Article history talk header templates. These templates should all have warnings and a tracking category for when they're in the wrong namespace (those in the broader non-article category should also have these). Here's an example of when a tracking category could have helped. Counter to this, I seem to recall either the Article history or the DYK talk page template tacitly endorsing some use in articles by mentioning how the transclusion would differ.

Multiple article history statuses shouldn't take up unnecessary real estate, especially on talk pages that are already filled with notices. I would like to propose an semi-automated cleanup that combines multiple occurrences of redundant templates into a single Article history template (maybe even self-initiated, if I can pull a bot together). If I advance far enough, it might even be useful to have a TfD for merging talk templates (either for select templates or for the entire concept of article history; my preference is currently the latter).

A standardized anti-redundancy warning template header notice should probably be at the top of every article history-related template's page (with a few exceptions). If there is concern about missing important information, couldn't Article history have a selection of important history events that it displays by default (like AfD and promotion events, sans demotion). I'm sure this is tricky LUA coding, though.

Other links that might be useful here: Category:Wikipedia good articles templates, Good articles/Templates


 * Article history
 * FailedGA
 * I am currently clever enough to craft some searches for this:
 * (40216+ pages)
 * (86 pages)
 * {{sl|hastemplate:FailedGA insource:/\{\{FailedGA.*?\{\{FailedGA/||ns1}} (74 pages)
 * Ideally, this could be more efficient, but there's something I don't understand about Mediawiki's regex engine that prevents this from working:
 * {{sl|hastemplate:FailedGA insource:/ \{\{FailedGA(?>(?:(?>[^={]*)(?:(?!==)=|(?!\{\{FailedGA)\{)?)+)\{\{FailedGA /}} (this search link template doesn't work because something is preventing it from properly parsing my search to send it to Special:Search. But the search query itself also doesn't work if it was copy-pasted into the search bar.)
 * It seems the issue is that Mediawiki doesn't currently support atomic groups {{code|(?> ... )}.

{{tlp|Article history}}, {{tlp|FailedGA}} (now there's a whole collection of such templates in Category:GA talk message boxes, though most of them are subtemplates.)

Is {{tlf|FailedGA}} really necessary? As far as I can tell, {{tlf|FailedGA}} only supports one review listing, while Article history is more versatile and supports multiple. But is {{tlf|FailedGA}} even worthwhile to keep when there's only one occurrence?

(Obviously further investigation is worthwhile before replacement)


 * {{tlf|FailedGA}}: 2311 transclusions
 * {{tlf|Article history}}: 40107 transclusions, and increasing.

A related question is how many occurences of {{tlf|Article history}} only have one parameter.

And of course, there remains the immediately more important question of how many articles that have been GA reviewed (through the article's talk /GA# subpages) or FA reviewed (through the FAC subpages; perhaps with differences for an older process) are marked as such. I might be able to search for this using Quarry now. There's almost certainly a gap, and given the inconsistency I've seen, I'd imagine it's pretty sizeable, though the 40000 {{tlf|article history}} transclusions may say otherwise. I'll bet there's just plain unintentional gaps, too.

GA header template redundancy table

 * Template name regex (TnameRegex)
 * Multiple occurences:
 * Article history -
 * GA -
 * FailedGA -
 * DelistedGA -

I wonder if I could do it through a LUA module. Probably, but I'd want to get a good table creation library. A template could probably format the Tnameregex, but a LUA module would be required for the table to be able to generate an arbitrary number of rows and columns.


 * 01:02, 20 May 2019 (UTC): The last fix was actually only a partial fix. But I'm fairly sure it's completely fixed now. But there are still false postives; numbers in parenthesis denote false positives. The false positive I've seen so far is using the header template in a non-heading context, i.e. to congratulate/denote that an article has passed its GAN.
 * 01:02, 20 May 2019 (UTC): The last fix was actually only a partial fix. But I'm fairly sure it's completely fixed now. But there are still false postives; numbers in parenthesis denote false positives. The false positive I've seen so far is using the header template in a non-heading context, i.e. to congratulate/denote that an article has passed its GAN.
 * 01:02, 20 May 2019 (UTC): The last fix was actually only a partial fix. But I'm fairly sure it's completely fixed now. But there are still false postives; numbers in parenthesis denote false positives. The false positive I've seen so far is using the header template in a non-heading context, i.e. to congratulate/denote that an article has passed its GAN.

{{sl|||ns1}} {{sl|hastemplate:"" hastemplate:""||ns1}} {{sl|hastemplate:"Tname" insource:/\{\{ *TnameRegex *[{{!}}}].*?\{\{ *TnameRegex *[{{!}}}]/||ns1}}

{{cot|My edits fixing this}}
 * Special:Diff/897887973
 * Talk:Franklin's electrostatic machine (no fix necessary)
 * Talk:Ikuhiko Hata (no fix necessary)

{{cob}}

Talk header template redundancy miscellanea

 * A bot could also search for article talks that should have templates but don't by searching article talk GA subpages and FAC archives.
 * I don't think Talk page header works as well as the creator intended it to. The idea that experienced will glance at it and remember guidelines just overlooks basic human psychology; after you see enough of the same thing, you just start to filter it out. I do see it being potentially useful to new editors and experienced editors who have taken a break from editing Wikipedia, but I don't think it serves the often editor very much.
 * But this is hard to address (since there's no easy way to automatically and systematically filter editors based on how often they visit talk pages) and a bit of a personal problem.
 * Talk:Christopher Columbus: This article is listed as multiple quality classes (B and C class) The classes be consistent.

Links in header templates
In templates, I think the most specific link to a discussion should be linked first, with less specific links later.

An example is the deletion template; I find it annoying that the general deletion criteria is linked long before the specific discussion regarding deletion. It's annoying to have to spend more time hunting for a link when the link could be sooner. As it is, it seems that sometimes templates are word-linked cleverly in places that don't also match the utilitarian places to link, and I find that problematic when there would a more utilitarian place to link that is still semantically understandable.

To put my previous sentences into plain English, I notice that templates often try to link words related to their targets, which I consider a good practice.

As an editor, I find it more important to be able to access the relevant links quicker.