Wikipedia talk:High-risk templates

Protect templates and images

I have a quick question. If we protect high-risk templates from editing, should we not also consider cases of protecting the images transcluded by the template? Doing the former allows us to prevent someone some inserting "My roommate is gay!" into the text of a template, but doesn't stop someone from uploading a photo of genitalia in place of wikiproject's logo on a talk page banner transcluded tens of thousands of times. Imzadi 1979 → 04:10, 6 November 2010 (UTC)

You're absolutely right. And we actually already do protect such high-risk images. But we haven't written it up as an editing guideline yet. (Unfortunately we have a problem with well meaning admins that keep deleting/unprotecting the protected high-risk images, so we keep adding more and more protection methods for those images.) You can read more about this at Wikipedia talk:Cascade-protected items#About high-risk images.

--David Göthberg (talk) 18:56, 3 October 2011 (UTC)

Template:Portal talk

Was just editing Template:Portal talk and noticed its not protected in any way. 308 transclusions found to page and fewer than 30 watchers. Moxy (talk) 06:00, 31 December 2010 (UTC)

318 transclusions is not that large number. And none of them are articles. So the template is not high-risk, why do you think it should be (semi-?)protected? (If you don't think it should, why do you mention it here?) Svick (talk) 22:10, 6 January 2011 (UTC)

Requiring consensus for changes

This guideline has always said that changes to fully protected "high risk" templates should be discussed on the talk page. Currently it says: "If fully protected, so that they can only be edited by administrators, these templates should be changed only after consensus for the change has been established on the template's talk page."

Before that it was[1]: "These templates should be edited only by administrators and only after a consensus to change the template has been established on the template's talk page."

Originally it said[2]: "These templates should only be edited by administrators after a consensus to change the template has been established on the template's talk page."

Fully protected "high risk" templates are often changed without any consensus or even any discussion. I could give hundreds of examples, but I'll point out only two atm. The history of Template:Pp-template, a template about this very issue, has many changes not on its corresponding talk page. The other example is Template:Asia topic, where myself and others have made changes without any discussion.

Perhaps there should be consensus for large changes, but this part of the guideline has never reflected reality, and I don't think small changes should require discussion on the talk page. If the underlying reason for this "must be discussed first" is to avoid the performance hits of bad template programming, I think that the problem is over stated and should be deprecated or diminished per WP:PERF. The other reason is to avoid vandalism, and admins arnt vandals so there is no reason to slow them down by requiring they get consensus on the talk page. John Vandenberg ^(chat) 11:40, 11 November 2011 (UTC)

I like the current wording, but think it could be tweaked to reflect that uncontroversial changes may be made without discussion. My philosophy is that administrators do not have any additional authority when it comes to editing, so they should not be using the admin tools to make changes in a way that non-administrators can not. I admit there are some fine lines between WP:BOLD, WP:PROTECT and also sometimes WP:INVOLVED and it might be a good idea to try to clarify these issues. — Martin (MSGJ · talk) 16:41, 11 November 2011 (UTC)

I support adjusting the guideline to actual practice. I believe for templates that have been protected only because they are high-risk we typically have the following situation: (1) They have not been protected because of any disputes, so reasonable edits should be allowed with as little red tape as possible. Basically, BRD should apply. (2) They are used very widely or very promintently, so anyone editing one of them should know precisely what they are doing to prevent breaking something, and should be extra careful not to get involved in an edit war. If there is any doubt on either point, the talk page should be used first. Hans Adler 23:45, 11 November 2011 (UTC)

I strongly concur. — SMcCandlish   Talk⇒〈°⌊°〉 Contribs. 22:59, 4 February 2012 (UTC)

I oppose any loosening of the restriction. The purpose of the full protection is to make sure that all changes got discussed on the talk page first, not only to prevent vandalism. A well intentioned but poor edit on a high-risk template can break many thousands of pages. Having a few people go over changes first decreases this risk. This is not vandalism, but the effect is bad enough anyway. These changes are rarely so urgent that mentioning them on the talk page first isn't possible. Letting admins change these templates as they see fit only increases the gap between admins and non-admins in a totally unnnecessary manner. Fram (talk) 10:44, 14 November 2011 (UTC)

The point of the restriction is to make sure that admins won't become first class citizens, so to speak. Just think what would happen if there were a dispute at Template:Asia topic between an ordinary editor and an admin. Allowing admins to freely edit such templates while anyone else is technically forbidden from doing so creates an inequality in our user base that only leads to worse things. --Conti|✉ 17:59, 15 November 2011 (UTC)

It's causing other problems, though. I've had {{editprotected}} requests turned down, for trivial tweaks, on the basis that "I don't see a consensus on the talk page for it", or because "this hasn't been sufficiently sandboxed and testcased". Not every change is dangerous, and not every change needs to have a talk page thread about it. The idea that admins aren't already "first class citizens" is naive; innumerable admins "take shortcuts" all over the place just because they can, and this is mostly tolerated because most of them are smart enough not to do so in ways that directly piss people off. That said, I agree with the rationale that we don't want to make the situation worse and widen the gulf.

But this argument largely only applies to genuinely high-risk templates. A very large number of low-risk templates that don't even have talk pages, or have ones no one reads or GAFs about, have been protected under [mis?]interpretatons of this guideline, simply because they have some code connection to an actual HRT. In those cases there is no one to form a consensus with. Most of them are one-editor templates, randomly protected for no genuine reason. One upshot of this is that the purpose of the full protection that is put into place on trivial templates like {{strongbad}} is emphatically not "to make sure that all changes got discussed on the talk page first", since no one cares about them but their developers and sometimes a tiny handful of other editors. It's to protect them from vandalism because they're allegedly "highly visible". As I get into in more detail in the #Stepping on developers' toes thread below, this is often a faulty assumption (it's usually only true when the actual HRT calling the trivial template is one that is always subst'd).

What needs to happen here is a sharp division between the concepts of

1. High-visibility templates (generally also implies high-use)
2. High-use but low-visibility templates (often metatemplates and other "guts" code that isn't immediately user-facing)
3. Incidental connection to either of the above (often just like #2 but only conditionally deployed, so both low-visibility and low-use)

and different protection criteria for each.

On that last point here's a starting suggestion, numbered as above:

1. Hi-vis: Always full permprotect
2. Hi-use only: Always partial permprotect, but don't full protect unless and until they are repeatedly vandalized, or when consensus [i.e., an actual discussion] determines on a case-by-case basis that the fallout from vandalism would be very severe
3. Incidental: Never protect at all unless and until vandalized, or when consensus determines on a case-by-case basis that the fallout from vandalism would be very severe; and absent that condition, only protect to the level necessary, just as we would with articles (e.g., temporary partial protection to stop IP vandals, temporary full protection to stop more widespread vandalism, escalating terms of protection to deal with longterm patterns of vandalism, etc.)

It's getting exceedingly frustrating to keep developing templates, only to be locked out of my own work for no real reason, when no one else is working on the templates or cares if I twiddle with their details. I doubt a single person on the entire system cares if I make some minor code tweak like add an if-test to {{strongbad}} (I actually do need to do exactly that Update: an admin unprotected it long enough for me to fix what I needed to), which has recently been added to {{collapse top}} as a replacement for a long string of bare CSS. — SMcCandlish   Talk⇒〈°⌊°〉 Contribs. 22:50, 4 February 2012 (UTC)

Stepping on developers' toes

It would be nice:

1. If not every minor template that happened to be transcluded in some obscure section of code in an actual high-risk template was treated as if it were itself a high-risk template. Unless obscure templates are used inside genuinely high risk templates that are regularly subst'd and thus expose their code to, e.g., noob vandals, there's no reason to full-protect such utility templates, as they are really at no more risk and no more highly visible than any other random template. I have no objection to their permanent semi-protection, to ward off monkeying by anons. A case in point is {{strongbad}} (which is incidentally used by the highly-visible {{collapsetop}}, if and only if it the latter is called with |warning=y), which is one of the templates I'm still working on; for one thing it needs a ns:0 test just like the xt family of templates Update: Fixed. But, someone just now willy-nilly permprotected it, so I'm locked out. (Yes, I've been around 6.5 years and am not an admin; too much politics and my tongue is apparently too sharp.)
2. If admins look at template history, and if a template has a) largely or entirely been developed by one or two coders and/or b) shows signs of significant recent reworking by anyone, notify that/those editors of the intent to full-protect the template, and ask them if they need any time to complete their immediate development plans for the thing. It's like notifying people that you've tagged their article for deletion, or that they've been mentioned at WP:ANI; just a matter of courtesy and respect, and procedure appropriateness.
3. If WP:HRT regulars took some time to think about how filing {{editprotected}} requests to fix a suddenly-protected template someone is still halfway through developing, and hoping it takes less than a week Update: try two and a half weeks, lately for it to be responded to, and hoping that the responding admin even understands the code, and hoping that they consider the coder's sandboxing and testcasing to be sufficient, and hoping they don't have some bee in their bonnet about a trivial change not being "discussed for consensus", and, and, and... is a massive pain in the backside. It's getting to be so much of one I'm considering throwing up my hands and not bothering to develop templates any more except when utterly necessary (which would be a net loss for WP, frankly; I'm among the more skilled, active and need-fulfilling template coders). HRT folks might consider re-prioritizing a bit and not being so protection-happy. Not to mention ensuring that there are template-savvy admins processing the {{editprotected}} backlog in the "Template talk:" namespace, most of which is literally being caused by HRT being over-protective. It's not helpful to fix one problem by causing another. Another way to look at it: If you lock 50 cats in your house, you have to feed them. It's lately been taking almost a week Update: almost three weeks for template editprotecteds to get dealt with, in my experience (maybe random bad luck, but it looks like a pattern to me).

WP:HRT "enforcement" is seeming increasingly paranoid and singleminded to me, with decreasing regard for the fallout all this protectionism is causing, and I doubt that I'm alone in this perception. Yes, there's no policy that says that the open editing afforded to articles, which are not full-protected unless absolutely necessary, should be applied to templates, and obviously we have to be less lenient with templates, because one incident of vandalism can be seen on thousands of pages via template abuse, than we are with articles, but the pendulum has swung too far, away from the concept that this is a wiki, and people edit it, and that's the whole point. This isn't some firewalled development environment with staging servers and NDAs and security badges and 4 rounds of QA checks for every code alteration. WP is not going to fall apart if some joker goes and changes {{strongbad}} to purple or makes it italic for the 5 minutes or 5 hours it would take before someone reverted and blocked the vandal. And it's a radically unlikely vandalism candidate. So are a large number of templates this guideline or process or project, whatever you want to call it, is full-protecting just because of their incidental connection to one of "the big ones". — SMcCandlish   Talk⇒〈°⌊°〉 Contribs. 22:20, 4 February 2012 (UTC)

• I realize that this comment is being made months after McCandlish posted this, but I have to say that I agree with it completely. It seems as thought the impulse to protect templates is way too high, generally.
  — V = IR (Talk • Contribs) 21:24, 13 April 2012 (UTC)

• • Just to echo Ohms law. The amount of protection done is exceesive. Protecting templates that have never been incorrectly edited is just protectionism that is alinating to editors and ultimately harmful to Wikipedia itself. Regards, SunCreator ^(talk) 07:34, 18 May 2012 (UTC)

Asleep at the switch?

It's been the better part of a month since I raised serious issues, twice, with how this page is being [mis]interpreted to call for extremist protectionism of templates high and low, and not one person has responded in any way. I'm inclined to interpret this as agreement with my observation that the pendulum has swung too far. If people, admins especially (since they're the ones dealing with these templates), don't even notice that discussion, including criticism, of this guideline and its practical effects is taking place, why does it even have {{guideline}} on it? (Technically it's {{subcat guideline|editing guideline|...}}, but you get the point.) — SMcCandlish   Talk⇒〈°⌊°〉 Contribs. 13:12, 21 February 2012 (UTC)

It's possible that the TL;DR aspect is drowning out the message that you were trying to get across with these couple of posts.
— V = IR (Talk • Contribs) 21:25, 13 April 2012 (UTC)

I am just passing, I didn't even know that this guideline existed. My own take on this is that sometimes there is disagreement that leads to edit wars on templates much like on ordinary pages and in cases like that then full protection is needed and probably should remain if the template is used on many pages. There are also core templates such as {{Citation/core}} which have complex structures and really should not be edited without thorough testing in a sandbox first and agreement for the change to be reached on the talk page. But for may other high usage templates, to protect them from vandalism (for which they are a prime target for malignants (rather than passing graffiti sprayers who just spray on wherever page they happen to read) it is often a better solution to use "Block new and unregistered users" as that still allows long time responsible editors to edit the template (a requirement of assume good faith). -- PBS (talk) 09:51, 4 May 2012 (UTC)

I think edits to the words in templates aren't likely to cause breakage and we shouldn't worry about it much, while edits to complex template code is normally only done by knowledgeable editors, who should follow good practices about testing in a sandbox etc. Maybe with the forthcoming Lua templates, we can figure out a way to separate the words from the code, and then allow protecting the code while leaving the words editable. 66.127.55.46 (talk) 05:19, 18 May 2012 (UTC)

Overboard with protection

I have noticed a shocking number of templates, particularly WikiProject templates, that have been fully or semi protected due to being a high use template. Some only have a couple of hundred articles associated. This really is counter productive to the pedia. Even if someone did vandalize the page these are going to be discovered and reverted nearly instantly so there is no need to over protect these templates. Kumioko (talk) 11:20, 27 May 2012 (UTC)

The above two sections say basically the same. Seems agreement to your point. Regards, SunCreator ^(talk) 15:40, 27 May 2012 (UTC)

I agree with you and SMcCandlish. — Lfdder (talk) 10:29, 7 August 2013 (UTC)

Ten years later, and this is still an ongoing problem. And my initial hopes for the templateeditor right have pretty much vanished. —Locke Cole • t • c 16:43, 20 February 2023 (UTC)

Certain articles that are high-risk include...

• GoAnimate (edit | talk | history | links | watch | logs) 112.209.86.65 (talk) 05:59, 15 November 2015 (UTC)

This page is about high risk templates, not articles. I see that the GoAnimate article is currently protected. To ask for other articles to be protected, post at Wikipedia:Requests for page protection. -- John of Reading (talk) 07:58, 15 November 2015 (UTC)

Frequently substituted

I was just asked to fulfil a request to add template protection to a user subpage that was intended to be substituted as part of that user's signature. This kind of usage could be argued to fall under clause #3 of this guideline, "It is substituted extremely frequently on an ongoing basis", if that user made a lot of talk page posts, but I don't think the intent of the guideline was to call this kind of usage "high-risk". Should we tighten up the wording here somehow? I'm thinking of something like "It is substituted by multiple users extremely frequently on an ongoing basis", although there are probably better ways to put it. — Mr. Stradivarius ^{♪ talk ♪} 01:27, 30 April 2016 (UTC)

As the guy who made the aforementioned request, and realized that it wasn't legitimate, I do support a change of wording to "substituted by multiple users extremely frequently on an ongoing basis". — Andy W. (talk · contrib) 01:37, 30 April 2016 (UTC)

Userspace pages can be protected per WP:UPROT, regardless of whether they technically qualify as high-risk or not, "as long as a need exists" - Evad37 [talk]

@Evad37: Question is whether template protection is a valid level, then? The need "exists"-- it's a substituted template in my userspace that could be a vandal target, as the signature page itself points out. For me, semi didn't cut it, (and probably shouldn't), and full wouldn't help me. On the other hand, "template protection" seems designed more for transclusion/visibility risk, right? (The admin reason is "highly visible template", which my signature template is technically not.) So I understood and backed out. — Andy W. (talk · contrib) 02:31, 30 April 2016 (UTC)

I think there's enough precedence for that type of template-protection in userspace, see the list of template-protected userspace pages (from Special:ProtectedPages). I think WP:UPROT is intentially meant to be a lower bar to pass than general requests for protection – otherwise what would be the point of having the WP:UPROT provision? - Evad37 [talk] 02:49, 30 April 2016 (UTC)

Haha. @Evad37: Gave it some thought. I'm inclined to agree with you at this point, especially with policy trumping guideline. The two signature examples I see are this and this. [edit]Here is one with the reason (User request within own user space). @Mr. Stradivarius: would you be willing to undelete the page I referred to then? :) — Andy W. (talk · contrib) 03:47, 30 April 2016 (UTC)

The problem is that the protection policy is also quite specific about where template protection may be used. From WP:TEMP-P: "This protection level may only be used on high-risk templates and modules and possibly in rarer cases where pages in other namespaces become transcluded to a very high degree." I think some of those template-protected userspace pages are in violation of the policy, although personally I wouldn't object to loosening the wording to include uses like those. That's probably something that needs to be proposed on the policy talk page, though. — Mr. Stradivarius ^{♪ talk ♪} 05:44, 30 April 2016 (UTC)

Continued the discussion here. Thanks — Andy W. (talk · contrib) 06:59, 30 April 2016 (UTC)

Should */sandbox, */testcases be unprotected

I think most sandboxes and testcases should be unprotected and not protected by the bot. They can be protected manually by an admin in case of misuse of sandbox or used in live templates. See also Template talk:Portal#Module:Portal/images/c/sandbox. Thingofme (talk) 02:16, 28 June 2022 (UTC)

Should some templates be fully protected?

Are there any "very" high risk templates or modules which need full protection or is template protection adaquate? The guideline is not clear on this matter, and there is an ongoing discussion about the protection of Module:WikiProject banner — Martin (MSGJ · talk) 21:59, 11 August 2023 (UTC)