Requests for comment/LTA database at meta

From Meta, a Wikimedia project coordination wiki
Jump to navigation Jump to search

The following request for comments is closed. There is no consensus for this proposal. --Mykola 20:43, 10 August 2022 (UTC)Reply[reply]


Rational[edit]

As of now, there are multiple long-term abuse (LTA) databases scattered across Wikipedia projects in different languages, but this leaves a problem: Wikipedia is not the only platform which can experience long term abuse. What about vandalism to Wikibooks, wikiversity, wikiquote, etc.? I don’t think creating a database for each project is a good idea since each one will have to be maintained separately. Also, the databases might not be at sync with each other which will make it harder to detect the vandals. I suggest to merge those databases to one large one at meta which will include all currently known serious LTAs (not in order to feed the trolls, but in order to give information on how to deal with serious ones who continue with their activities sometimes for over a decade). Also, I think there should be a usergroup with representives from all major wikis in charge of maintaining this list, that way it will also ease communication between the mods of different wikis and help fight crosswiki abuse. We should always avoid feeding the trolls, but if that strategy proves to be ineffecient, we have no choice but to add that user to the LTA database in order to make it easier to detect future activity. -GIFNK (TALK | CONTRIBS) 10:48, 5 February 2022 (UTC)Reply[reply]

Support[edit]

  • Strong support Strong support as submitter, per rationale. -GIFNK (TALK | CONTRIBS) 10:49, 5 February 2022 (UTC)Reply[reply]
  • Support Support and divide by sections of serious socks (normal LTA), banned users with multiple sockpuppets (example: C/Musee Annam) and allow translations. Also, we should reform SRG to become like WP:AIV in enwiki, where a person can report clear vandal cases, LTAs, bot sections separately. Thingofme (talk) 12:54, 7 February 2022 (UTC)Reply[reply]
  • Support Support This might make the oversighters job easier.--Yaratmayıcı (talk) 18:20, 8 February 2022 (UTC)Reply[reply]
  • Support Support Although we might want to discuss viewing access too.--Snævar (talk) 00:23, 9 February 2022 (UTC)Reply[reply]
  • Support Support This would be challenging to implement and require a lot of community conversation, but the alternative is continued inadequate defense against abuse. I support because I recognize the problem as stated and believe that the proposed solution would effectively address the problem. I do think that centralizing this on meta may contribute to misunderstandings or conflicts across Wikimedia platforms and language communities, and that having community conversation in advance would reduce future problems. Bluerasberry (talk) 01:42, 9 February 2022 (UTC)Reply[reply]
  • Support Support (conditional, tentative, provisional). I sometimes have to refer to LTA pages on frwiki, nlwiki, jawiki, and others (as well as enwiki), so it might make some sense to think about concentrating them in one place on meta. There's a number of challenges and questions this raises. Most importantly, I could only support if there continues to be public read access to this information. -- zzuuzz (talk) 09:10, 9 February 2022 (UTC)Reply[reply]
    Some challenges came around with translations, but it would be a benefit for users (who don't know the language) Thingofme (talk) 11:25, 9 February 2022 (UTC)Reply[reply]
    @Zzuuzz and Thingofme:, I also think the translation will be challenging but worth it, and that there should be public read access to the database. -📜GIFNK📖DLM💻MMXX🏰 (TALK🎙 | CONTRIBS) 11:39, 9 February 2022 (UTC)Reply[reply]
  • Support Support --Novak Watchmen (talk) 16:24, 21 February 2022 (UTC)Reply[reply]

Oppose[edit]

  • Oppose Oppose I'm in the school of thought that these LTA pages do more harm than good. It's remarkably easy for LTAs to evade detection when their M.O. is documented out in the open. It can also encourage copycats. We shouldn't glorify these individuals or their behaviour. There's of course value in having this documentation for other patrollers to reference, so in my opinion a private wiki (with a relatively low barrier to access) would be better, or invent something new. But, if we're going to continue having these pages in the open, then sure, it makes sense to centralize them on Meta. MusikAnimal talk 22:00, 13 February 2022 (UTC)Reply[reply]
    I think privacy problems are a problem. We can see that to stop vandalism, CU and OS log is kept privately, and some private abuse filters are designed to make vandals not know how it works. Even OS requests must be made privately. However, we should make the database to the public, as not everyone are patrollers and they may occasionally patrols (due to some busy works) Thingofme (talk) 15:32, 14 February 2022 (UTC)Reply[reply]
    @MusikAnimal and Thingofme:, to begin, I don’t think the privacy issues are really an issue as no really private information is being disclosed. If the lta edits without creating an account he’s the one exposing the information. MusikAnimal, do you suggest creating a wiki abuse.wikimedia.org that not everyone can access but the bar for entry is not high? IMO the only problem with the current databases is that there are multiple of them and it’s far easier to maintain one that it is to maintain many. -📜GIFNK📖DLM💻MMXX🏰 (TALK🎙 | CONTRIBS) 17:25, 14 February 2022 (UTC)Reply[reply]
    I'm sure it depends on the wiki, but English Wikipedia puts lots of very descriptive information about LTAs out in the open. This isn't "private" information, per-se, but it makes it easy for the LTA to know what no to do in order to get caught again. My belief is this kind of documentation shouldn't be visible to them. A private wiki could serve as a central database, too. I realize however not everyone agrees with me that this should be private, so we'd need a broader RfC for that idea by itself before you could move forward with a request to create a new private wiki. MusikAnimal talk 23:39, 14 February 2022 (UTC)Reply[reply]
    But if it is private, we can easily request account on this wiki with requirements (good patroller, no blocks in 6 months, not a vandal, have more than 2000+ global edits) I think it would make this better. Thingofme (talk) 00:34, 15 February 2022 (UTC)Reply[reply]
    @MusikAnimal:, this is an interesting idea and it depends on one question. Is there evidence of LTA evading detection due to the detailed information being public? I think if the LTA continues abusing he will get banned anyways. Btw, I’m just curious, what will be the licence of the content in abuse.wikimedia? -📜GIFNK📖DLM💻MMXX🏰 (TALK🎙 | CONTRIBS) 11:55, 15 February 2022 (UTC)Reply[reply]
    Also there’s CU to check is address. -📜GIFNK📖DLM💻MMXX🏰 (TALK🎙 | CONTRIBS) 12:18, 15 February 2022 (UTC)Reply[reply]
    I'm not going to dig up any diffs but there are a number of LTAs I deal with regularly, and I can say with confidence they know about their respective LTA pages and are likely to check them. It's the same reason we use private edit filters for LTAs. Banning these LTAs accomplishes nothing since they continually create throwaway accounts, but once they know we've caught on their M.O., they'll change it. Going by their public LTA pages is a good way for them to evade detection, hence why I argue it should be private. abuse.wikimedia.org doesn't have to be CC-BY, I don't believe. The CheckUser wiki attributes all contributions to copyright (c) Wikimedia Foundation. MusikAnimal talk 22:29, 16 February 2022 (UTC)Reply[reply]
    @MusikAnimal:, I see but does this make it significantly harder to detect them? I mean you can find them anyways. -📜GIFNK📖DLM💻MMXX🏰 (TALK🎙 | CONTRIBS) 16:08, 17 February 2022 (UTC)Reply[reply]
    There is a private checkuser wiki, also Thingofme (talk) 03:53, 20 February 2022 (UTC)Reply[reply]
  • Oppose Oppose strongly, per MusikAnimal. If a LTA database has to exist, it must be hosted elsewhere and not in the clear for anyone to see (WP:DENY & WP:BEANS, among others). A private wiki is absolutely impractical and overkill, and a huge maintenance burden. What about a Toolforge or Cloud VPS project with access controled via OAuth to specific user groups meeting some predetermined criteria? Would that be possible? An attempt to create a "LTA Knowledgebase" on Toolforge existed (e.g. https://phabricator.wikimedia.org/project/profile/2434/) but it was discontinued without further notice (that I'm aware of). I'd be interested in knowing the reasons as to why it didn't go forward. —MarcoAurelio (talk) 16:36, 21 February 2022 (UTC)Reply[reply]
  • Oppose Oppose While the intent would be to create a central repo for information on LTA to be used to protect all projects, what we would be actually doing is creating an LTA Hall of Fame. Don't feed 'em. Operator873 connect 20:41, 21 February 2022 (UTC)Reply[reply]
  • Symbol oppose vote oversat.svg Strong oppose Only vandalisms who keep cross-wiki abusing should be add to LTAs of meta. For me, most LTAs like to edit on different wikis with different characteristics, and adding them to meta seemed not useful. Pavlov2 16:57, 5 March 2022 (UTC)Reply[reply]
  • Symbol oppose vote oversat.svg Strong oppose It doesn't make sense to document patterns used by the LTAs. We should be transparent, but everything has its limits. We can't be cutting the branches we're sitting on. --Martin Urbanec (talk) 22:59, 10 March 2022 (UTC)Reply[reply]
  • Oppose Oppose per MusikAnimal --Ameisenigel (talk) 15:03, 27 March 2022 (UTC)Reply[reply]
  • Symbol oppose vote oversat.svg Strong oppose per above. --Victor Trevor (talk) 11:44, 26 June 2022 (UTC)Reply[reply]
  • Symbol oppose vote oversat.svg Strong oppose I question the value of maintaining LTA lists centrally because LTAs will use this information adversely (WP:BEANS) and adapt their M.O. very quickly, which means the information in the LTA list will be outdated almost as soon as it is written down. ArticCynda (talk)

Neutral[edit]

  • Neutral Neutral Judgements of LTAs may be different on every wikis due to their benefits, providing a global sock master database, however, could be useful. --Liuxinyu970226 (talk) 04:09, 6 February 2022 (UTC)Reply[reply]
  • Neutral Neutral Most LTAs have a different behaviour on a different wiki, but the formatting used in w:ja:LTA:SLIME could work. --SHB2000 (talk | contribs) 08:47, 6 February 2022 (UTC)Reply[reply]
@SHB2000: What is the meaning of the term "slime" in Japanese Wikipedia? Bluerasberry (talk) 01:44, 9 February 2022 (UTC)Reply[reply]
@Bluerasberry I'm not a Japanese speaker, so unfortunately I have no clue. SHB2000 (talk | contribs) 06:58, 9 February 2022 (UTC)Reply[reply]
This user uses "slime" just as its English meaning. However, this user is a cross-wiki abuse like LTA:QCHM and LTA:HMGY from zhwp.Pavlov2 17:02, 5 March 2022 (UTC)Reply[reply]
  • Neutral Neutral: I don't think we should build a database with translations and all. But a set of categorized redirects from meta to their primary wiki, for cross-wiki LTAs, and brief disambiguation pages for LTAs with documentation pages in multiple wikis could be handy. MarioGom (talk) 19:00, 16 May 2022 (UTC)Reply[reply]

Other comments[edit]

  • Comment Comment I have seen some people want access to this database to be restricted only to admins. I believe it should work like the current databases, only improve the format, and have it all in one place easier to maintain. I think read access should be open to all. write access only to admins and the user group in charge of maintaining. -📜GIFNK📖DLM💻MMXX🏰 (TALK🎙 | CONTRIBS) 11:42, 9 February 2022 (UTC)Reply[reply]
  • Comment Comment If you want to create a LTA database at Meta, just do it. Why is there a need to ask for permission through a RFC? Nguyentrongphu (talk) 02:41, 13 February 2022 (UTC)Reply[reply]
@Nguyentrongphu:, I thought this was the best way to gain consensus & bring users together for this task. Do you feel it can be created now? -📜GIFNK📖DLM💻MMXX🏰 (TALK🎙 | CONTRIBS) 12:27, 13 February 2022 (UTC)Reply[reply]
Content that is currently scattered thoughout multiple wikis needs to be imported and unified into one database. I think it’s no small task therefore needs consensus & a community willing to maintain it. -📜GIFNK📖DLM💻MMXX🏰 (TALK🎙 | CONTRIBS) 12:32, 13 February 2022 (UTC)Reply[reply]
First of all, you don't need consensus to do this. Second, consensus and a community willing to do this are two different things. Just because there is consensus, that doesn't mean there's a community willing to maintain this. Imo, this RFC is unnecessary. Just start it yourself then recruit others who have similar interest to maintain it with you. Nguyentrongphu (talk) 21:08, 13 February 2022 (UTC)Reply[reply]
@Nguyentrongphu:, can I just create a page Long-term abus documentation and start importing content? -📜GIFNK📖DLM💻MMXX🏰 (TALK🎙 | CONTRIBS) 17:27, 14 February 2022 (UTC)Reply[reply]
Sure, go ahead. I don't see any policy against doing it. Nguyentrongphu (talk) 03:25, 15 February 2022 (UTC)Reply[reply]
There used to be a project about LTA database proposed by @TheresNoTime: and I think it would be great to hear about her suggestions. Stang 23:53, 4 April 2022 (UTC)Reply[reply]