Template:Redirect category shell is permanently protected from editing because it is a heavily used or highly visible template. Substantial changes should first be proposed and discussed here on this page. If the proposal is uncontroversial or has been discussed and is supported by consensus, editors may use ((edit template-protected)) to notify an administrator or template editor to make the requested edit. Usually, any contributor may edit the template's documentation to add usage notes or categories. Any contributor may edit the template's sandbox. Functionality of the template can be checked using test cases.

Redirect Template‑class

This page is within the scope of WikiProject Redirect, a collaborative effort to improve the standard of redirects and their categorization on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks. Note: This banner should be placed on the talk pages of project, template and category pages that exist and operate to maintain redirects. This banner is not designed to be placed on the talk pages of most redirects and almost never on the talk pages of mainspace redirects. For more information see the template documentation.RedirectWikipedia:WikiProject RedirectTemplate:WikiProject Redirectredirect pages Template This template does not require a rating on Wikipedia's content assessment scale.

This template was considered for deletion on 13 August 2016. The result of the discussion was "no consensus".

See also: Template talk:This is a redirect for discussions about the previous iteration of this template.

Given @Headbomb's comment here, I decided to create a template for automatically categorising redirects. The result is User:Qwerfjkl/sandbox/rcatCategoriser, which calls User:Qwerfjkl/sandbox/rcatCategoriser/shell. It works as far as I can tell, but I'd welcome any help with it. — Qwerfjkltalk 21:09, 18 November 2022 (UTC)[reply]

@Qwerfjkl, love it. EpicPupper (talk) 03:05, 4 February 2023 (UTC)[reply]

Elli attempted to have redirects from wikidata automatically be sensed more than a year ago. Inadvertently, I duplicated their work at /sandbox. We should really start discussing this again along with potentially automatically sensing namespacing per the fix suggested by Qwerfjkl above. We really don't need editors adding these rcats when it'd be more efficient to do so automatically. –MJL ‐Talk‐^☖ 18:54, 16 December 2022 (UTC)[reply]

@MJL: Yeah I agree with this. You might be interested in what I put together at User:Elli/rcat standardization. The main problem is that we'll need a bot to go over all current redirects with these rcats to remove them and if necessary add ((rcat shell)). Elli (talk | contribs) 19:14, 16 December 2022 (UTC)[reply]

@Elli: I don't think that would be as much of a problem if we implemented the entirety of that page. The only issue is making sure that we aren't leaving behind empty rcat shells. With something like ((R from subpage)) though, that might be a bit more challenging (though I guess we could just add ((R unprintworthy)) to fill it out).
I'm not sure how BFRA's work. Do we need to establish a new consensus here or...? –MJL ‐Talk‐^☖ 03:52, 17 December 2022 (UTC)[reply]

@MJL: we need to establish a consensus to do the work somewhere.

When I was making these changes previously, Paine Ellsworth expressed some concerns, so pinging them to see how they feel about things now (as it is like, over a year later). Elli (talk | contribs) 11:24, 17 December 2022 (UTC)[reply]

Yes, I'm just sorry it's taken so long to resolve, as it sounds like an idea whose time has come. From the [previous talk that has been archived] we see that my concern was that the edit was made to this template before the BRFA was approved and the work was done. And this because there were some redirects with two renditions of the ((Wikidata redirect)) template. There was also the discussion at Template talk:Wikidata redirect where the application of this Rcat shell without params came out. As long as these issues are dealt with, I think we should go ahead with this improvement, because we seem to have enough of a consensus to do so at this point. P.I. Ellsworth , ed. ^put'r there 18:18, 19 December 2022 (UTC)[reply]

Okay, so I guess we just need a solid actionable plan for how we're going to handle this update (to ensure we don't hit the same pitfalls). @Paine Ellsworth: How would you feel about creating a new maintenance category by applying something called like ((R untagged from bot)) as a failsafe to prevent overpopulating Category:Redirect category shell without parameters? That's probably been thrown around before, but it might be worth considering now. –MJL ‐Talk‐^☖ 23:17, 20 December 2022 (UTC)[reply]

Yes, a solid, actionable plan sounds very good. My concerns so far are 1) not doubling up on the ((Wikidata redirect)) template on redirects such as asteroid names, which use the ((NASTRO comment)) template, and 2) ensuring that Category:Miscellaneous redirects does not get way overpopulated by the improvements. Editor Elli could probably better answer your question about a new cat applied by a new rcat template. P.I. Ellsworth , ed. ^put'r there 02:32, 24 December 2022 (UTC)[reply]

@Paine Ellsworth, 2) could probably be solved with something like

i.e. the automatic rcat template producing nothing, or else something that can easily be checked such as ((str find|<span style= "display: none">no rcats</span>)). — Qwerfjkltalk 04:50, 24 December 2022 (UTC)[reply]

To editor Qwerfjkl: is this something that can be tested? Is there some similar usage that can be observed and analyzed? I'm wary of solutions that cannot be tested until implemented. P.I. Ellsworth , ed. ^put'r there 05:57, 25 December 2022 (UTC)[reply]

@Paine Ellsworth, it can be tested with the /sandbox; how else would it be tested? — Qwerfjkltalk 09:10, 25 December 2022 (UTC)[reply]

It occurs to me that the easiest way to keep from doubling up on the ((Wikidata redirect)) templates might be to get the template that autocats it to stop doing that. Let me look into that. Won't actually try that unless I hear back from you, editor Elli. P.I. Ellsworth , ed. ^put'r there 10:11, 20 December 2022 (UTC)[reply]

@Paine Ellsworth, I've managed this with User:Qwerfjkl/sandbox/hasRcat. — Qwerfjkltalk 08:22, 28 December 2022 (UTC)[reply]

Currently it causes a template loop. I think this can be fixed if it is called through another template, by getting the raw text of the page, removing the autocatting template, and then expanding it. — Qwerfjkltalk 22:22, 28 December 2022 (UTC)[reply]

Suggest testing be isolated until you're certain that it works well. P.I. Ellsworth , ed. ^put'r there 20:39, 29 December 2022 (UTC)[reply]

I'll try to provide better implementations when I can. — Qwerfjkltalk 20:51, 29 December 2022 (UTC)[reply]

@Paine Ellsworth, I've done this in ((Automatic redirect categories/sandbox)), calling ((Automatic redirect categories/applyRcat)). It should work for every rcat except for Template:Wikidata redirect/hard; @Elli, is there any reason not to rename that template, as you suggest in the documentation of ((Automatic redirect categories))? — Qwerfjkltalk 07:45, 30 December 2022 (UTC)[reply]

Nevermind, I've almost managed it, sorry. My question to Elli stands. — Qwerfjkltalk 09:46, 30 December 2022 (UTC)[reply]

@Qwerfjkl: Yeah, it could be renamed. Elli (talk | contribs) 15:42, 30 December 2022 (UTC)[reply]

@Qwerfjkl: Wait, if this change is applied, won't it prevent templates from being sorted into Redirect category shell without parameters and Miscellaneous redirects if the page is protected? That effectively eliminates the ability for editors to intentional sort into the later if they are unsure about what to tag a (protected) redirect. Is that desired? –MJL ‐Talk‐^☖ 00:24, 31 December 2022 (UTC)[reply]

No. It needs a better implementation, but I'm currently working on other things. — Qwerfjkltalk 03:14, 31 December 2022 (UTC)[reply]

Ah, well, my original thought was to have a semi-temporary rcat called like ((R untagged auto)) which could be applied to redirects which would be left without any rcats after the conversion. Volunteers could then go through it and correctly apply the correct rcats where they need to be manually placed. –MJL ‐Talk‐^☖ 18:07, 31 December 2022 (UTC)[reply]

┌──────────────────────────────┘
@MJL, I agree there should be a placeholder rcat, which would possibly add a maintenance category. — Qwerfjkltalk 19:56, 31 December 2022 (UTC)[reply]

@MJL, per @Primefac's comment on my BRFA, I can probably run it after the autodetecting is added. — Qwerfjkltalk 15:35, 17 December 2022 (UTC)[reply]

The current sandbox should work, so long as a placeholder rcat is applied. — Qwerfjkltalk 20:08, 31 December 2022 (UTC)[reply]

@MJL, I've created a draft template at ((R tagged automatically)). I think a maintenance category would be helpful as well, to track redirects needing.categorisation. Any thoughts? — Qwerfjkltalk 05:35, 2 January 2023 (UTC)[reply]

@Paine Ellsworth, I think I've got it working with ((Redirect category shell/sandbox)). See User:Qwerfjkl/sandbox/applyRcat/testcase (2). — Qwerfjkltalk 05:17, 2 January 2023 (UTC)[reply]

Not sure the Rcat shell is ready for such an application. Seems to be off the track of this discussion, which has to do with application of the ((Wikidata redirect)) template. Also concerned that printworthiness in mainspace is not addressed. There are so many redirects that require a thinking mind to determine which category(ies) is or are appropriate for them. So there is only limited bot or botlike viability. I appreciate the work put into this, but I am not yet convinced that this is the way to go. P.I. Ellsworth , ed. ^put'r there 07:05, 3 January 2023 (UTC)[reply]

@Paine Ellsworth, not sure what you mean by concerned that printworthiness in mainspace is not addressed - do you mean if an rcat categorises as (un)printworthy but an editor tags it with the opposite rcat?
There are so many redirects that require a thinking mind to determine which category(ies) is or are appropriate for them. So there is only limited bot or botlike viability. That's why this is limited to the rcats that can automatically be applied, by ((Automatic redirect categories))
I think this is appropriate for this discussion, which is about applying rcats automatically, even if it was originally suggested just for Wikidata redirects.
Is there are reason not to do this? — Qwerfjkltalk 08:32, 3 January 2023 (UTC)[reply]

Yes, what I mean is that most of the rcat templates cannot really be applied automatically. ((R printworthy)) and ((R unprintworthy)) are prime examples of this. Looking at [this page] and [this page], the subpage rcats are also examples. Aside from my not knowing how you manage to get both subpage rcats, ((R to subpage)) and ((R from subpage)), by applying just ((R to subpage)), there are many Foo/bar pages where "bar" is not necessarily a subpage of "Foo", but instead is an actual project page in its own right. See WP:Naming conventions/Geographic names for one of many examples. Editor SMcCandlish and others might shed more light on that for you. Usage of the subpage rcats to be automatically applied in all cases of "Foo/bar" shows that we still have much to learn about this idea. We cannot just start applying bot or bot-like editing to most of the rcats. ((R from move)) has been automatically applied for a long time, and ((Wikidata redirect)) appears to be another good candidate to be automatically applied; however, those are the only rcats I know that are or might be adaptible to bot or bot-like executions. P.I. Ellsworth , ed. ^put'r there 21:28, 3 January 2023 (UTC)[reply]

On the "/" point: Yes, there are many pages that are guidelines in their own right and are not just subpages, but which have "/" in their names. E.g. most (but not all) pages that begin with "Wikipedia:Manual of Style/".  — SMcCandlish ☏ ¢ 😼  19:10, 4 January 2023 (UTC)[reply]

@Paine Ellsworth, it uses the mw.title.getCurrentTitle().isSubpage method in Module:RedirectData to check, which is an built-in method, so is presumably reliable. Other than that, just the namespace rcats are applied - which is just a case of detecting the namespaces of the redirect and its target - and ((R to disambiguation page)), which is Foo (disambiguation) → Foo.
Aside from my not knowing how you manage to get both subpage rcats, ((R to subpage)) and ((R from subpage)), by applying just ((R to subpage)) - after spending far too long on this, I managed to write some rcat-detecting code, so it suppresses applying rcats if the page already has them. — Qwerfjkltalk 21:44, 3 January 2023 (UTC)[reply]

"Presumably reliable"? No, I don't think that can discern between true subpages and /bar pages that are not true subpages. Also noticed your auto template senses dab redirects. How does it sense that the page name does or does not have "(disambiguation)" in the title? Because if it cannot make that distinction, then it will tag ((R from ambiguous term))-type redirects incorrectly. P.I. Ellsworth , ed. ^put'r there 22:05, 3 January 2023 (UTC)[reply]

@Paine Ellsworth,

if string.match(titleObject.text, "(disambiguation)") then -- title contains (disambiguation)

— Qwerfjkltalk 22:32, 3 January 2023 (UTC)[reply]

Okay, stepping forward, if those redirects are in mainspace they will also require ((R unprintworthy)), and of course if they are not in mainspace, then printworthiness can be ignored. Also, many of those dab redirects resulted from an old page move and have not yet been tagged with ((R from move)). Can that also be sensed and corrected? P.I. Ellsworth , ed. ^put'r there 22:40, 3 January 2023 (UTC)[reply]

@Paine Ellsworth, if those redirects are in mainspace they will also require ((R unprintworthy)) - which ones? Most of these rcats automatically apply printworthiness, so I don't think this will be necessary.
Can that also be sensed and corrected? - probably, but not by a template. This would take a bot task. — Qwerfjkltalk 22:45, 3 January 2023 (UTC)[reply]

There's the rub in terms of printworthiness. While having rcats like ((R to disambiguation page)) automatically apply unprintworthiness was a definite step forward, it only satisfies one of two requirements. All rcats have two purposes; in the case of the printworthiness rcats they are: 1) sort to either Category:Printworthy redirects or to Category:Unprintworthy redirects, which is done by some rcat templates, and 2) inform editors who see the redirect about why such categorization was made, which is only done by actually installing a printworthiness rcat template on the redirect. So it definitely will be necessary to include printworthiness rcats which, in the case of such dab redirects must include ((R unprintworthy)).

And if the page move can be sensed by a bot, then why can't it be sensed by a template? Submit that if it cannot be done by a template, then it probably cannot be done by a bot, either. P.I. Ellsworth , ed. ^put'r there 00:00, 4 January 2023 (UTC)[reply]

@Paine Ellsworth, I thought the point of the ((R (un)printworthy)) templates was that they were applied manually, i.e. an editor has decided the printworthiness.
Bots can query the API, see the page logs, history etc. Templates can't. — Qwerfjkltalk 00:37, 4 January 2023 (UTC)[reply]

Firstly, since ((R unprintworthy)) should always accompany ((R to disambiguation page)), there is no reason why it cannot be handled the same way as the latter template. Secondly, I find that very interesting because I always check the history page(s) to find out if the redirect is the result of a page move. I wonder how a bot would do it? The word "move" is used in histories to explain category "moves", section "moves", and other kinds of moves, to include page "moves". How would the bot be able to discern a page move from other different kinds of moves found in a page's edit history? P.I. Ellsworth , ed. ^put'r there 03:05, 4 January 2023 (UTC)[reply]

┌─────────────────────────────────┘
@Paine Ellsworth, by querying mw:API:Logevents. — Qwerfjkltalk 05:10, 4 January 2023 (UTC)[reply]

That probably won't work for redirects older than the move log, but I imagine it's a good start. –MJL ‐Talk‐^☖ 07:09, 4 January 2023 (UTC)[reply]

I guess older than that you'd have to manually check each revision, or maybe check for Conversion script? This seems a bit off-topic however. — Qwerfjkltalk 07:28, 4 January 2023 (UTC)[reply]

@Paine Ellsworth: Aside from my not knowing how you manage to get both subpage rcats, ((R to subpage)) and ((R from subpage)), by applying just ((R to subpage)), there are many Foo/bar pages where "bar" is not necessarily a subpage of "Foo", but instead is an actual project page in its own right. See WP:Naming conventions/Geographic names for one of many examples.

So, that definition doesn't exactly square with the definition provided by Wikipedia:Subpages (which is a content guideline, I guess).

I'm not saying you're wrong because that's pretty much exactly how Template:R to subpage/doc says to use the template, but the two pages contradict one another in that regard. Either a subpage: (1) includes pages independent of their parent as long as they contain a / in their title and are outside mainspace or (2) is only a child page which is actually dependent on its parent page outside mainspace. If the former is true, then the template should be changed. If the latter, then a bot cannot accurately apply this rcat. –MJL ‐Talk‐^☖ 07:29, 4 January 2023 (UTC)[reply]

@MJL, what do you mean by (in)dependant? — Qwerfjkltalk 07:37, 4 January 2023 (UTC)[reply]

@Qwerfjkl: Like subpages relying on another page are dependent vs MOS:LEAD which is an independent subpage. –MJL ‐Talk‐^☖ 23:34, 5 January 2023 (UTC)[reply]

@MJL, I see. The template currently handles case 1; if subpage is defined differently, this should be removed from the template. — Qwerfjkltalk 03:30, 6 January 2023 (UTC)[reply]

@MJL, @Paine Ellsworth, in the interest of moving things along, we can (initially, at least) prevent the template from adding ((R to subpage)) — ((R from subpage)) shouldn't have such a problem. Are there any other problems? — Qwerfjkltalk 21:52, 14 January 2023 (UTC)[reply]

((R from subpage)) shouldn't have such a problem.

Certainly would have a problem. Suppose the target has a solidus or slash (/), but is not a subpage. Then the redirect that also has a solidus would not be a true subpage, either. I would also like to move forward, however we should insist that it begins as correctly as possible, because this will not be an uncomplicated task. P.I. Ellsworth , ed. ^put'r there 16:54, 16 January 2023 (UTC)[reply]

@Paine Ellsworth, like Wikipedia:Manual of Style/Command-line examples? Then let's not apply either.
Any other problems? — Qwerfjkltalk 18:01, 16 January 2023 (UTC)[reply]

Hard to say at this point without a thorough check of the other rcat templates in this index for adaptability. P.I. Ellsworth , ed. ^put'r there 18:23, 16 January 2023 (UTC)[reply]

In the interest of being able to move forward, I think it'd be best if we focus on just adding the automatic detection of ((R from unnecessary disambiguation)), ((R to disambiguation page)), the namespace rcats, and ((R with Wikidata item)). ((R unprintworthy)) and ((R printworthy)) can be added by humans or via software in the future if that's desired (for example, by somehow making rcats with |printworthy=yes make ((Redirect category shell)) transclude ((R printworthy))). For now though, it should just be added by humans.
@Paine Ellsworth and Qwerfjkl: Sound good? That seems like where we are at. –MJL ‐Talk‐^☖ 17:54, 17 January 2023 (UTC)[reply]

@MJL, I have no problems. — Qwerfjkltalk 18:47, 17 January 2023 (UTC)[reply]

Perhaps the dab rcat templates can be applied; however, the ((Wikidata redirect)) and ((R with Wikidata item)) seem redundant, so I'm unclear about the readiness there. P.I. Ellsworth , ed. ^put'r there 06:50, 18 January 2023 (UTC)[reply]

@Elli? — Qwerfjkltalk 07:04, 18 January 2023 (UTC)[reply]

The point of having this as two templates is that ((R with Wikidata item)) should be used on hard redirects, within ((redirect category shell)), while ((Wikidata redirect)) should only be applied to soft redirects i.e. to Wiktionary. Using either in the other context looks off-putting and ugly and while we could have the template itself detect as such it's much easier to just never manually apply ((R with Wikidata item)) (and instead have it always be applied by ((redirect category shell)), so ((Wikidata redirect)) is always only used on soft redirects. Elli (talk | contribs) 16:08, 18 January 2023 (UTC)[reply]

Really have no preference either way, but with nearly 33,000 hard redirects presently tagged with Wikidata redirect and R with Wikidata item not yet tagging anything, it seems the latter should be used to tag soft redirects and the former to remain on hard redirects. And this knowing that the minor-planet redirect loving ((NASTRO comment)) template already auto-applies Wikidata redirect to a lot of asteroid hard redirects and similar. I do see the need now for the hard WD vs. soft WD application. P.I. Ellsworth , ed. ^put'r there 16:52, 18 January 2023 (UTC)[reply]

The reason why I wouldn't swap the names is that other soft redirect tags are named similarly, i.e. ((Wiktionary redirect)) (and the same is true of rcats for hard redirects). Elli (talk | contribs) 20:26, 18 January 2023 (UTC)[reply]

That does make more sense. Many many years ago I set out to categorize redirects. It was a daunting task, and not just because there were already millions of uncategorized redirects, but also because some editors didn't think redirects should even be categorized, let alone using templates to do so. Now I sort of find myself in their shoes. Auto- and bot-categorization should be used whenever possible, so it's time to move forward. P.I. Ellsworth , ed. ^put'r there 16:01, 19 January 2023 (UTC)[reply]

@Paine Ellsworth, If it helps, I could submit a BRFA for fixing these, i.e. detecting if the redirect is hard and if so switching out the template. — Qwerfjkltalk 16:59, 19 January 2023 (UTC)[reply]

Seems like the vast majority, around 30,000, can be swapped just by swapping the templates in the ((NASTRO comment)) template. Not sure if a bot would be needed for the rest of 'em. P.I. Ellsworth , ed. ^put'r there 17:54, 19 January 2023 (UTC)[reply]

Courtesy pings to editors Tom.Reding and Rfassbind. Looks like we should change ((Wikidata redirect)) to ((R with Wikidata item)) in the ((NASTRO comment)) template. The former is to be used only for soft WD redirects, while the latter is to be used for hard WD redirects. P.I. Ellsworth , ed. ^put'r there 17:59, 19 January 2023 (UTC)[reply]

@Paine Ellsworth, edit request made. — Qwerfjkltalk 18:03, 25 January 2023 (UTC)[reply]

@Paine Ellsworth, the template has been modified. I searched insource:"Wikidata redirect" -insource:"Wiktionary redirect" -insource:"short pages monitor" and got no hard redirects, so I think it's safe to assume that the remainder are soft redirects. It may take a while for the pages to update. — Qwerfjkltalk 20:59, 29 January 2023 (UTC)[reply]

Thank you, editor Qwerfjkl! Moving right along. P.I. Ellsworth , ed. ^put'r there 23:35, 29 January 2023 (UTC)[reply]

┌─────────────────────────────────┘

bot conversion

@Paine Ellsworth, sorry, I've done a more detailed search and I missed around 6000. I've written some code to convert them to ((R with Wikidata item)). I'll run this in a day if there are no objections. @Elli, can the rcat handle a parameter? — Qwerfjkltalk 19:05, 31 January 2023 (UTC)[reply]

List at User:Qwerfjkl/sandbox/WD redirects. — Qwerfjkltalk 19:06, 31 January 2023 (UTC)[reply]

@Qwerfjkl: not sure what you're asking? Elli (talk | contribs) 19:33, 31 January 2023 (UTC)[reply]

@Elli, I'm about to convert a large number of ((Wikidata redirect))s to ((R with Wikidata item)). Can ((R with Wikidata item)) handle ((R with Wikidata item|Qxxxxxx)) properly? — Qwerfjkltalk 20:22, 31 January 2023 (UTC)[reply]

@Qwerfjkl: Yes, it should be able to. Elli (talk | contribs) 20:23, 31 January 2023 (UTC)[reply]

Great. — Qwerfjkltalk 20:24, 31 January 2023 (UTC)[reply]

I've started running this. — Qwerfjkltalk 10:55, 1 February 2023 (UTC)[reply]

doc page issues

The template doc at Template:Wikidata redirect/doc is wrong or at least, misleading. I reverted one of the bot edits, because the doc says, "Wikidata redirect may be used on all redirects". I'll self-revert, but someone should have a look at the /doc and rephrase it appropriately. Mathglot (talk) 22:22, 1 February 2023 (UTC)[reply]

@Elli. — Qwerfjkltalk 22:25, 1 February 2023 (UTC)[reply]

Continued

@Elli, @Paine Ellsworth, are there any further issues? — Qwerfjkltalk 16:40, 14 February 2023 (UTC)[reply]

Best of everything to you and yours this holiday season! and...

Happy New Year!
Thank you for your contributions to Wikipedia and its redirects, and
have a prosperous, productive and enjoyable New Year, Best Ever, in 2024!

I propose that the redirect Template:Rcat shell simplify its wikitext from

#REDIRECT [[Template:Redirect category shell]]

((Redirect category shell|
((R from high-use template))
((R to redirect template))
((R from template shortcut))
))

to:

#REDIRECT [[Template:Redirect category shell]]((Rcat shell|((R from high-use template))((R to redirect template))((R from template shortcut))))

Can this change be made to simplify the wikitext? Thank you! 2003 LN6 06:37, 14 February 2024 (UTC)[reply]

No. Primefac (talk) 12:07, 14 February 2024 (UTC)[reply]

That does not simplify the wikitext. — Qwerfjkltalk 12:35, 14 February 2024 (UTC)[reply]

The documentation explains in the Notes section why the present readability format should not be compromised. P.I. Ellsworth , ed. ^{put'er there} 01:15, 4 April 2024 (UTC)[reply]

I understand the procedures now. Thank you! 2003 LN6 16:01, 14 February 2024 (UTC)[reply]