Requests for Comment/Lower requirements for global Interwiki administrator

Global Interwiki administrators are users with the ability to edit the local interwiki table of any (public) wiki. As of writing, in order to be one, you must not only be supported by users here on Meta, you must also meet 3 prerequisites:
 * 1) Have at least 1000 global edits.
 * 2) Been on Miraheze for at least 2 months.
 * 3) Been involved in some way on community matters.

It's no secret that I don't care about edit counts, so I'm already biased against such requirements.

First, 1000 edits is insane: Think about it, I, an editor who has edited almost everyday since joining, am not eligible for this role. As of writing, I have 922 global edits, and I bet if you were to analyze my edits, you would find out that the only reason I'm even close to 1000 is because of how translations work, in that every edit when translating is in reality 2, as every time you edit a translation you edit both the translation and a translation unit in the Translation: namespace (Special:EditCount/OrangeStar).

Second, the necessity of such a requirement is questionable at best: For example, such requirements don't exist for roles like Administrator in Meta (arguably a role that has more potential for misuse that interwiki-admin will ever have), yet you won't see any newbies in that role. The reason for this is that the people that vote in these requests for permissions already oppose requests from those that don't participate in the community, or that they don't know, which are, I suppose, the reasons for requiring such a high edit count in the first place. OrangeStar (talk) 14:55, 13 December 2022 (UTC)

Proposal 1: No edit count requirement
The first proposal is to drop any edit count requirement for this role. The other 2 requirements will remain, so if this passes, any editor who has been in Miraheze for at least 2 months and "been involved in some way on community matters" can request global Interwiki administrator. OrangeStar (talk) 14:55, 13 December 2022 (UTC)

Support

 * 1)  as proposer. OrangeStar (talk) 14:55, 13 December 2022 (UTC)
 * 2)  Reasonable. Ora &#38; D (talk) 16:00, 13 December 2022 (UTC)

Abstain

 * 1)  I support my own Proposal 2 and as such do not have any strong views on this matter. In general I believe that edit counts which are not too high or arbitrary can be a good thing to also avoid WP:SNOW and WP:NOTNOW requests from wholly inexperienced users. --DeeM28 (talk) 07:57, 14 December 2022 (UTC)
 * 2) I oppose the status quo but do not support this change. --DeeM28 (talk) 09:02, 14 December 2022 (UTC)

Oppose

 * 1)  Per Agent's comments.  I think generally it's very difficult to determine whether someone is trustworthy enough to be appointed in a role. Especially with interwiki administrators, how are we really supposed to know they're trustworthy? Perhaps an acceptable compromise would be a higher threshold (maybe 90%?) to bypass the edit requirement or alternatively a reduced one, but I would not think eliminating it completely would be useful and it would likely lead to users easily getting interwiki admin because it would be perceived as a role which doesn't require that much trust. Reception123 (talk) ( C ) 15:27, 13 December 2022 (UTC)
 * We can never know with full certainty, but if there's something that will not tell you how trustworthy someone is is definitely edit count. Just look at the example I gave in this RfC, I have almost 1000 edits partly because I make translations, and I think no-one would call someone who only has 1000 edits via translations (as opposed to participating in all the various noticeboards in addition to translating) trustworthy. OrangeStar (talk) 15:46, 13 December 2022 (UTC)
 * That's why edits aren't the main criteria but are an additional safeguard. It could also be said that it shows that the user is willing to spend their time on Miraheze meaning that it's less likely they would "go rogue" and for example post a malicious link as that would pretty much give them a negative reputation in the community. --Reception123 (talk) ( C ) 16:24, 13 December 2022 (UTC)
 * I think the 2 months requirement already does that. It proves that they've been here for a while, and obviously those who have 2-months old accounts but none or almost none edits are going to be opposed because people don't know them. In fact, if i had to pick something to determine if a given candidate is trustworthy is seeing their comments and opinions on RfCs, RfPs and what not. It says much more about a given individual that account age, edit count and whatever other metric could ever say. OrangeStar (talk) 16:54, 13 December 2022 (UTC)
 * Having seen so many permission requests, I wouldn't be that certain that someone who is around 2 months but only has a few edits would be shot down, once again because interwiki administrator is viewed as a permission that doesn't require a lot of activity or experience (which it doesn't, but it requires trust). As for the comments part, I fully agree but as I say the whole idea of the edit requirement is just a safeguard to prevent someone who appears to be trustworthy or a "nice person" to be rapidly made interwiki administrator. The reason why it isn't the case for Stewards or Global Sysops is because there the community sets a high bar already. I will admit once again that maybe 1000 isn't the right number and I would, for example, agree to lower it to 500 but I wouldn't want it completely eliminated in order to allow a user with very few edits who makes themselves seem trustworthy to become interwiki admin. Reception123 (talk) ( C ) 18:30, 13 December 2022 (UTC)
 * Let's say an untrustworthy user gets interwiki-admin rights. So what. Remove them and move on with your life. The only thing this group does is give you the ability to edit interwiki tables. About the most scary thing it can do is use scary transclusion, which doesn't actually leave any damage in the wiki at all, and about all it does is a more inefficient version of trolling with CSS, which they can already do by just... editing an article. If they are abusing this right, just remove them and that's it, no-one reasonable will oppose an interwiki-admin being removed for obvious abuse. Wikis we're built from the beginning with the idea of people and bots from all over the Internet editing articles, and the idea of being able to revert to a specific version of an article. This includes the interwiki table. OrangeStar (talk) 20:54, 13 December 2022 (UTC)
 * In that case you could argue the same for every group but I think people prefer to know they can trust someone rather than just hope they won't need to be removed, and why should we risk even minimal damage being done (i.e. users being sent to phishing websites)?. The reason why interwiki admin exists as far as I'm aware is due to the fact that malicious links could be added and then presented as wikilinks by administrators or bureaucrats. Otherwise, I can't deny that there hasn't been a debate as to whether interwiki administrator should exist at all. I'm not saying I'd necessarily support that but as long as we preserve the group I stand by my arguments. In the future I would also recommend that you try to present a draft of an RfC or at least discuss with a few other users before as there may be other ideas and compromises (as you can see in the comments) that can be presented as proposals. Reception123 (talk) ( C ) 06:31, 14 December 2022 (UTC)
 * And I do argue the same for every group with very few exceptions generally, like CheckUser, as you can't unsee IP addresses, as well as interface administrator. Deletions can be undeleted, blocks can be removed, and suppressions can be unsuppressed (though I do have a higher standard for OS). No matter what access a bad-faith user gets, they'll, with very few exceptions, never be able to leave damage that can't be reverted. In my experience in wikis, vandals and other bad-faith users are kinda obvious if you've been around the wiki for long enough. If they haven't been blocked in the first month or so, have participated in the community in a non-trolling way, and just generally been doing good stuff, chances are they're a good-faith user, or not purely a vandal at least. OrangeStar (talk) 12:20, 14 December 2022 (UTC)
 * It doesn't necessarily have to be actual 'physical' damage, it can also be reputational damage. For Stewards or Global Sysops this argument is fully applicable as it's easy for Stewards or Global Sysops to not necessarily go "rogue" but to exceed their powers, interpret policies in an inconsistent manner, etc. For interwiki administrators, I'm sure the users who click on a malicious phishing link wouldn't be glad to hear that the community voted them because if something goes wrong they can just be removed. I think the argument that we should just approve people and then if they're not acceptable remove them is a dangerous one and in my experience it's always way harder to remove someone than to appoint them. Reception123 (talk) ( C ) 12:24, 14 December 2022 (UTC)
 * 1) Per above. I also think that the edit requirement should be lowered. --Blad  (talk • contribs • global) 20:34, 13 December 2022 (UTC)
 * 2) per above. However, I would also support lowering the edit count (maybe 500 or so should be enough). Tali64³ (talk) 00:00, 14 December 2022 (UTC)
 * 3) per Reception123's comments, along with Agent Isai's. I would support lowering the edit count within the 600-800 edit range, but as of right now, that's all I would consider. I get why the proposer thought this would be a good idea, as it sounds like something I would've proposed, but too many users have abused this position. Thanks - BrandonWM (talk • contributions • global • rights) 04:04, 14 December 2022 (UTC)
 * 4)  per above
 * 5)  Sorry, but that sounds like a horrible idea. --DarkMatterMan4500 (talk) (contribs) 14:42, 14 December 2022 (UTC)
 * 6)  I don't think this is a good idea. Switzerland  (talk༆) 19:06, 14 December 2022 (UTC)
 * 7)  Edit count is by no means a perfect metric, (and not all edits are equal in impact) but it does demonstrate a level of experience with the platform and a body of work that ideally demonstrates both commitment and judgement of the candidate. I vote against outright removal, though agree the 1k edit count may be a bit high. --NotAracham (talk • contribs • global) 20:41, 14 December 2022 (UTC)

Comments

 * 1) Interwiki administrator is a role of much less experience and competence, unlike Meta administrator. We don't impose an edit requirement on Meta sysops because it's very unlikely the community would vote in someone who perhaps doesn't have that much experience or competency required to handle situations. On the other hand, Interwiki administrator is much easier to obtain than Meta sysop despite it being an actual global role. We have had abuse in the role before and we've certainly had users who are the poster child of WP:CIR in the role. I don't think dropping the edit requirement would do us any good to be quite honest, it would only allow even more unqualified candidates to apply. Agent Isai  Talk to me! 15:02, 13 December 2022 (UTC)
 * Your comment would only convince me if the community that votes for local roles and the one that votes for global roles were different, which is obviously not the case, they're both the Meta community. Just as it's unlikely for randos to become sysops, it's equally unlikely that randos will become interwiki admins (randos meaning unknown to the Meta regulars), the archives of Requests for global rights is proof of this. it would only allow even more unqualified candidates to apply then vote against them. Global Interwiki admin is already an obscure role that relatively few know exists, let alone apply for it. The last person to become one was Bukkit (who retired) back in September this year, and before them it was... yourself, back in September 2021. With the exception of Ugochimobi, the other interwiki admins are all either Stewards or SRE, who obviously have a lot more things to do than process interwiki requests. My intention with this request, other than allowing me to run for interwiki-admin (COI detected) is to make the process of requesting rights more inviting to other users. The process for getting rights here is very similar to Wikipedia's RFA, which is infamous for a reason. I think dropping stuff like this will help make anyone who thinks they could help in this area make the jump to posting the request. OrangeStar (talk) 15:43, 13 December 2022 (UTC)
 * Looking at the archives, there are a few randos/CIR users who have become interwiki-admin. It's so much easier for someone to become interwiki-admin versus sysop because the requirements are so much more lax. The reason our process of getting rights is similar to Wikipedia's RfA process is because historically, we've been very liberal about handing out rights like interwiki-admin, CVT, wiki creator, etc., and that has caused so many headaches to the community in the past in terms of having to deal with role abuse, lack of competency, lack of standards for enforcement, along with so many other things. I'm rather split when it comes to this. It would be great to have more users in roles but having the wrong users in them will cause more issues than staffing shortages could ever. Agent Isai  Talk to me! 16:12, 13 December 2022 (UTC
 * That's what the community is for, to vet candidates for advanced permissions to ensure only somewhat-competent people get rights. OrangeStar (talk) 21:07, 13 December 2022 (UTC)
 * That's true, but these requirements were also introduced by the community itself. There are many other areas where the community has provided for clear standards or rules that 'override' it in a way (like support ratios, minimum participation, voting eligibility). Reception123 (talk) ( C ) 06:34, 14 December 2022 (UTC)
 * 1) While this distinction may seem to be not so clear this is the main reason for why we have edit requirements. The truth is, an interwiki administrator's job doesn't require as much discretion as GS or Steward, you just need to not add malicious links. For that reason, the main factor isn't necessarily to see how the user handles situations or responses to users but is really almost fully based on whether the user is trusted or not. The fact that a user has many edits (as I mention above) indicates that they have spent a lot of time on Miraheze and are unlikely to give that all away. I'm not saying the actual number of edits is ideal but I think a fixed requirement is also useful to prevent people from easily supporting users who may not be trustworthy. What stops someone from requesting IW after a few weeks then? The community is likely to be less careful with IW which has a very simple and clear task. --Reception123 (talk) ( C ) 16:24, 13 December 2022 (UTC)
 * What stops someone from requesting IW after a few weeks then? The second requirement, being here for at least 2 months. I think the community does a good-enough job of keeping away obvious trolls from advanced permissions. OrangeStar (talk) 20:58, 13 December 2022 (UTC)
 * Fair enough, at the time when I wrote that comment I forgot about the requirement. In either case that's where I feel the issue lies, the word "advanced permissions" and the fact that as I say, IW won't necessarily be seen as something serious enough to oppose a user who appears like a trustworthy person. Reception123 (talk) ( C ) 06:33, 14 December 2022 (UTC)

Proposal 1A: 500 edit count requirement
The edit requirement is lowered to 500 edits

Support

 * 1)  Only if Proposal 2 fails. I have decided to create this as a lesser alternative to the initial one as it preserves the requirement but makes it less strict as 1000 edits is difficult to achieve. I myself only have 400 and have been here for a long time and would consider myself eligible for interwiki administrator otherwise even though I do not have any interest in the role. --DeeM28 (talk) 09:02, 14 December 2022 (UTC)
 * 2)  This is much better than the original proposal. --Blad  (talk • contribs • global) 12:32, 14 December 2022 (UTC)
 * 3)  only if Proposal 1 and 2 fails. OrangeStar (talk) 13:38, 14 December 2022 (UTC)
 * 4)  Very reasonable.  Agent Isai  Talk to me! 13:44, 14 December 2022 (UTC)
 * 5)  The initial edit requirement was probably not given much thought and given the arguments made by DeeM28 and OrangeStar I think it's perfectly fair to lower the requirement which admittedly is quite unreasonable and difficult to attain. Reception123 (talk) ( C ) 14:03, 14 December 2022 (UTC)
 * 6) I would like to see the edit requirement a bit higher, maybe at 600ish, but this is better than 1000 edits required. Thanks - BrandonWM (talk • contributions • global • rights) 14:23, 14 December 2022 (UTC)
 * 7) Coupled with the other requirement of 2 months tenure and community election to the role, this seems like a reasonable adjustment given scope and intent of role responsibilities. --NotAracham (talk • contribs • global)

Proposal 2: Abolish interwiki administrators
Foreword: I am aware that due to the responses to Proposal 1 this proposal is very likely to fail but I would nevertheless wish to present it as an alternative. From the comments above it appears that the reason for interwiki administrators existing is weak one. It appears that they only exist to prevent "malicious" links from being inserted or scary transclusion. Comparatively, Interface administrators have as far as I can understand more power to do damage than interwiki administrators but they are not regulated in any serious way. Is there really a large enough difference to justify keeping them between linking like this or like this?? I would argue that there is not and if anything interface administrators are the ones who should be regulated but that is not up for debate in this RfC. My proposal is the following. If successful I would like to request that this RfC page is renamed as otherwise it is liable to confuse people who later read the title. --DeeM28 (talk) 07:55, 14 December 2022 (UTC)

PROPOSAL:
 * Global interwiki administrators and local interwiki administrators are abolished
 * Local bureaucrats automatically have interwiki rights. These can only be delegated to administrators (no other groups)
 * A feed is set up on Discord and/or IRC to monitor all additions to interwiki tables
 * Any user who purposefully inserts a malicious link into the interwiki table is to be treated severely and may be globally locked

Support

 * 1)  Per the foreword above. It does not make much sense to regulate interwiki additions but instead allow interface admins to be freely chosen when they can do even more damage with CSS and JS scripts. --DeeM28 (talk) 07:55, 14 December 2022 (UTC)
 * 2)  Well said! Ora &#38; D (talk) 08:10, 14 December 2022 (UTC)
 * 3)  No issues, I agree with this proposal. OrangeStar (talk) 13:36, 14 December 2022 (UTC)
 * 4)  for the gist of this. After discussing it with a few people I've come to be very unconvinced that this role really adds much. Stepping stone to other volunteering opportunities? Too minor and out of the way, not to mention failing this in nearly all examples so far. Safety? I think the use case is simply far too minimal to excuse the inconvenience to liberty. Combined with actual monitoring and consequences I believe the impacts would be negligible and the actual handling would still save time compared with the hubub of actually maintaining this system. If anything with a reliable interwiki feed we could catch up on something like this more quickly than other forms of abuse and so it would be an easy flag for something we would have to deal with anyway. I'm willing to support this, see how it goes, and if I'm around should it be a problem, support a restoration of this group if the need is truly demonstrated. But I am thoroughly unconvinced of its merits now. --Raidarr (talk) 14:41, 14 December 2022 (UTC)
 * 5) Exactly! Since I came to Miraheze, I think this group is useless and just another HAT.--MrJaroslavik (talk) 18:20, 14 December 2022 (UTC)

Oppose

 * 1) Opposing for "Local bureaucrats automatically have interwiki rights. These can only be delegated to administrators (no other groups)". I would not trust everyone who requests a wiki to automatically have these rights. Submitting a malicious link would be as easy as requesting a wiki. --Blad  (talk • contribs • global) 12:39, 14 December 2022 (UTC)
 * I can do the same with no need for  rights. In fact, I can also make it look like an interwiki link, as well as a regular link to another article in the wiki. The reason of preventing  malicious links from being added to articles is a weak one. I'm gonna say though, it is easier to use abuse filters to automatically stop non-interwiki links compared to interwiki links, but if the IRC feed for interwiki changes is implemented, we won't need to use abuse filters. OrangeStar (talk) 13:35, 14 December 2022 (UTC)
 * It would be hard to have to add every malicious link possible into an AbuseFilter. Perhaps if an extension existed to automatically scan links, that would be helpful but otherwise I don't see how to implement that. Agent Isai  Talk to me! 13:43, 14 December 2022 (UTC)
 * Fair point, however, assuming that the link is in an unprotected page (which, if a random guy came along and posted a malicious link, this is likely the case), that link can just as easily be removed. If someone is posting in a protected area, worst case scenario, a Steward/GS comes along and removes it, upon someone's request. --Blad  (talk • contribs • global) 00:01, 15 December 2022 (UTC)
 * I should also note that users can host a local election for local interwiki administrator rights. I prefer this as it requires at least 5 active users, and a decent level of trust. --Blad  (talk • contribs • global) 16:15, 14 December 2022 (UTC)
 * 1) While indeed interface-admin should be regulated, there would be too much global backlash for that to ever occur. We should be proactive in trying to avoid any sort of future abuse and issues rather than just saying "Well, X can do worse so Y should be allowed." If anything, I would support giving Stewards and Global Sysops (Global Rollerbackers too?) the right rather than outright giving it to all bureaucrats to be assignable to any groups.  Agent Isai  Talk to me! 13:44, 14 December 2022 (UTC)
 * You think interface-admin should be regulated? Well, I can kinda see the point, but still... Would you then support limiting that access to something like Wikimedia's Global Interface administrators? In Wikimedia, that's a global group with access to the interface-admin toolkit + the  right in all wikis. The only formal requirement for that role is that the candidate must have 2FA enabled. OrangeStar (talk) 14:10, 14 December 2022 (UTC)
 * 1)  per above
 * 2)  Opposing based on technical challenges noted by others along with creating a new avenue for needed vandalism/malicious activity monitoring.  I'd prefer the platform not to become known as Scamaheze.  --NotAracham (talk • contribs • global) 20:49, 14 December 2022 (UTC)
 * 3)  Admittedly the security issues aren't that serious compared to interface admins they still do exist. I'm unsatisfied with the proposal because for technical reasons it isn't strictly possible to limit the rights to bureaucrat/sysops and they'd be able to bypass that. Additionally, for private wikis the wiki feed wouldn't really be possible. I also think that while this shouldn't be the case at all, in practice users always seem to be more motivated to volunteer and help if they have a 'hat', even as minor as it can be so interwiki admin (similarly to Global Rollbacker) is a minor role but which can be a first step to other roles in the future and more visibility. Reception123 (talk) ( C ) 20:56, 14 December 2022 (UTC)

Neutral

 * 1)  I got nothing to say on the matter other than maybe saying that this will likely cause issues if that were to be taken away. --DarkMatterMan4500 (talk) (contribs) 14:44, 14 December 2022 (UTC)
 * Please explain what issues that you personally believe will (likely) be caused. --Raidarr (talk) 13:26, 15 December 2022 (UTC)

Comments

 * 1) I haven't looked (yet) at the source code of ManageWiki, but I don't think the second point is technically possible currently. The closest we can get to implementing that proposal is probably to only allow those with  (Stewards and sysadmins, basically, assuming managewiki works like I think it does) to assign that right to local admins, so giving interwiki rights to admins would require a request at the SN. I also would've like to add an additional point to this proposal, that the   prefix shouldn't be deleted. OrangeStar (talk) 13:44, 14 December 2022 (UTC)
 * Yeah, it's not possible to only restrict one right to be assignable to only one group. It's also not possible for Stewards to assign any 'restricted' rights, that would have to be done via LocalSettings.php as those rights don't appear in ManageWiki. I don't know if the mh prefix can be overridden (it's not an actual interwiki prefix but rather a psuedo-interwiki prefix) and I also wouldn't like the "m" and "meta" prefix overridden as that's used to link to Meta in many pages like the footer links and such. Agent Isai  Talk to me! 13:48, 14 December 2022 (UTC)

Proposal 2.1
If Proposal 2 is successful, Global Sysops are given  permissions and may remove any malicious links.

Support

 * 1)  While I have not yet voted for Proposal 2, I think if we are to lower standards in that way we definitely should allow Global Sysops to remove malicious links as they might be able to respond more quickly than Stewards. I apologise for adding a proposal this late in, but it seems that that has already been done and as I mentioned before it would've been much better if a draft would've been presented before publishing this RfC. Reception123 (talk) ( C ) 14:20, 14 December 2022 (UTC)
 * 2)  Per Reception123. Global sysops should absolutely be trusted with interwiki access, they're just a level below Stewards. Thanks - BrandonWM (talk • contributions • global • rights) 14:26, 14 December 2022 (UTC)
 * 3)  as a useful supplement to the above. --Raidarr (talk) 14:41, 14 December 2022 (UTC)
 * 4)  They should've had this right from the start. --DarkMatterMan4500 (talk) (contribs) 14:45, 14 December 2022 (UTC)
 * 5)  Sure. OrangeStar (talk) 17:26, 14 December 2022 (UTC)
 * 6) I wanted to suggest the same thing but couldn't find the courage.--MrJaroslavik (talk) 18:22, 14 December 2022 (UTC)
 * 7)  Not a fan of proposal 2, but this is fine if it passes. --Blad  (talk • contribs • global) 18:53, 14 December 2022 (UTC)
 * 8)  While I disagree with proposal 2, proposal 2.1 would be needed (and might be a decent idea to empower Global Sysops with this anyway) to provide additional support.  GS is a trusted role and seems like a fine place to put such capabilities. --NotAracham (talk • contribs • global) 20:51, 14 December 2022 (UTC)