Wikipedia talk:Sockpuppet investigations/SPI/Clerks/Archive 4

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 1 Archive 2 Archive 3 Archive 4

Bullseye grant proposal

Good afternoon all, I have submitted a Rapid Grant proposal for bullseye, a tool I have been working on to consolidate detailed information about IP addresses into a single view. It is primarily targeted at checkusers and stewards, but is usable by all editors. At the suggestion of one of the grant coordinators, I am informing potentially interested communities about this proposal. I welcome any and all feedback on the proposal. Best, GeneralNotability (talk) 17:06, 15 October 2021 (UTC)Reply[reply]

Would you consider making this a module of my SPI Tools rather than reinventing the wheel on all the common stuff? -- RoySmith (talk) 18:58, 15 October 2021 (UTC)Reply[reply]
I think the two are fairly different use cases (and I routinely use this outside of SPI anyway). GeneralNotability (talk) 20:48, 15 October 2021 (UTC)Reply[reply]

Local blocks for glocked accounts?

Wikipedia:Sockpuppet investigations/Dereck Camacho lists a bunch of glocked accounts and requests that they be blocked and tagged locally. Is there any value to doing this? Is there any strong reason not to do this? -- RoySmith (talk) 01:22, 29 September 2021 (UTC)Reply[reply]

I guess if they successfully appeal a global lock and the account is unlocked then having a local block in place would be good? Callanecc (talkcontribslogs) 09:24, 29 September 2021 (UTC)Reply[reply]
I agree with Callanecc but for a different reason. Global locks are poorly signalled. In the case of Dereck Camacho you need to scan his CA & look to the block on the Spanish wiki to get a handle on why the lock was applied (sock of Lucifer2000). A block on enwiki for his socks would do no harm & potentially make future socks easier to report. A |crosswiki=yes flag on its {{SPIarchive notice}} wouldn't hurt either. Lucifer2000 is the master. Cabayi (talk) 09:54, 29 September 2021 (UTC)Reply[reply]
Additionally, autoblock might be helpful, though probably only for blocks by non-CUs. ~~~~
User:1234qwer1234qwer4 (talk)
17:56, 4 November 2021 (UTC)Reply[reply]


Two proposals for {{Checkip}} for your consideration, input (and someone with better skills than me Face-smile.svg) appreciated :

Thanks, - Cabayi (talk) 15:11, 11 November 2021 (UTC)Reply[reply]

Jack Frost appointed trainee clerk

I am pleased to announce that Jack Frost has been appointed as a trainee clerk. Jack has been helping out at SPI since February and now wants to make it official. I will be supervising his training. Jack is in UTC+11, a lot of timezones away from me; if there's somebody near his timezone who could provide additional support, that would be appreciated. Welcome to the team, Jack! -- RoySmith (talk) 14:43, 14 November 2021 (UTC)Reply[reply]

  • Welcome to the team, Jack! Quite a freshman class we now have. I'll leave it to Spicy to conduct the various hazing rituals. Just remember that Roy likes his coffee black, and I like my coffee to be chocolate milk. -- Tamzin[cetacean needed] (she/they) 20:55, 14 November 2021 (UTC)Reply[reply]
  • Thankyou both! Looking forward to learning the ropes with you and the rest of the team, and I'll make sure my apron is ready for the coffee run. --Jack Frost (talk) 06:53, 15 November 2021 (UTC)Reply[reply]
    I should point out that my coffee preference is actually not. I only do tea, preferably Earl Grey with lemon. I'm partial to Twinings. -- RoySmith (talk) 01:56, 17 November 2021 (UTC)Reply[reply]
  • Welcome Jack! Happy clerking! Cabayi (talk) 15:00, 15 November 2021 (UTC)Reply[reply]
  • Welcome to the team, Jack! --TheSandDoctor Talk 00:04, 17 November 2021 (UTC)Reply[reply]

TheSandDoctor promoted to full clerk

After discussion, the CheckUser team is pleased to appoint TheSandDoctor as a full SPI clerk. Congratulations, and thank you for your dedication to SPI. As somebody noted on the CU mailing list, "His clerking has been excellent to the point that I [...] thought he was a full clerk for the past year or so." -- RoySmith (talk) 01:50, 17 November 2021 (UTC)Reply[reply]

@RoySmith: Thank you and I extend my thanks to the CheckUser team and mystery commenter Face-smile.svg --TheSandDoctor Talk 18:37, 17 November 2021 (UTC)Reply[reply]
Congratulations TheSandDoctor! Well deserved. I think you get to move the tassel on your Fez to the other side now or somesuch Face-smile.svg - Cabayi (talk) 12:44, 18 November 2021 (UTC)Reply[reply]
Thank you, Cabayi! Moving the tassel over is quite the achievement indeed. No longer needing to wear the bumper sticker on my forehead that says "trainee" is also a nice touch Face-tongue.svg. --TheSandDoctor Talk 17:14, 18 November 2021 (UTC)Reply[reply]
+1. I and many other CUs supported making you a full clerk. Dreamy Jazz talk to me | my contributions 13:23, 18 November 2021 (UTC)Reply[reply]
Thank you, Dreamy Jazz! That means a lot. Face-smile.svg Glad it wasn't a split vote or something hehe Face-tongue.svg --TheSandDoctor Talk 17:15, 18 November 2021 (UTC)Reply[reply]
I can see you've figured out how to work the system. Get promoted and immediately go on vaction. Nice. -- RoySmith (talk) 23:42, 18 November 2021 (UTC)Reply[reply]
@RoySmith: Bahahahaha. That was over by the time I replied, but forgot to remove. Thank you for the reminder! Face-tongue.svg (I never have actually been able to take any time off...throughout that I still logged 41 edits. I need help I think...) --TheSandDoctor Talk 17:51, 19 November 2021 (UTC)Reply[reply]

Discussion of interest regarding the notification checkbox in Twinkle

Wikipedia talk:Twinkle § Request to remove the SPI notification checkbox

-- Tamzin[cetacean needed] (she/they) 17:52, 19 November 2021 (UTC)Reply[reply]

Two useful templates for trainee and/or non-admin clerks

I made these a while ago and realized that I forgot to advertise them to the rest of the trainee and non-admin contingents of the team. (As of TSD's promotion the former is now a subset of the latter.) So, @1997kB, Jack Frost, and Spicy:

Hope y'all find those helpful, and let me know if you have any bugs / feature requests. -- Tamzin[cetacean needed] (she/they) 19:10, 20 November 2021 (UTC)Reply[reply]

Dang, {{caw}} would have saved me a lot of copy-paste when I was in training. -- RoySmith (talk) 19:21, 20 November 2021 (UTC)Reply[reply]
These are great Tamzin; thanks! --Jack Frost (talk) 21:02, 20 November 2021 (UTC)Reply[reply]

Request for input on prospective SPI clerks

In a push to be more efficient and transparent, I would like to request public comments on the current prospective SPI clerks:

Checkusers, clerks and any other editors are welcome to leave endorsements or concerns at the relevant request above. Concerns can be privately raised on the functionaries mailing list if needed ~TheresNoTime (to explain!) 18:34, 27 November 2021 (UTC)Reply[reply]

cuwiki search link

I've added a link to search cuwiki to {{Checkuser/sandbox}}. If nobody objects, I'll move it to {{Checkuser}}. -- RoySmith (talk) 19:19, 28 November 2021 (UTC)Reply[reply]

Yes, good idea. EdJohnston (talk) 22:04, 28 November 2021 (UTC)Reply[reply]
A support from me. Dreamy Jazz talk to me | my contributions 23:06, 28 November 2021 (UTC)Reply[reply]
Done. And undone. It was working fine in the sandbox, but generating bogus displayed content including the "&ns0=1&ns2=1&ns14=1" crud. -- RoySmith (talk) 23:42, 28 November 2021 (UTC)Reply[reply]
And redone. It needed urlencode. -- RoySmith (talk) 23:54, 28 November 2021 (UTC)Reply[reply]
This addition is great, much appreciated!-- Jezebel's Ponyobons mots 21:09, 30 November 2021 (UTC)Reply[reply]

New user attempting to interview SPI clerks

I received an email today from a new user who says they represent the "Institute for Strategic Dialogue (ISD) and the Centre for the Analysis of Social Media (CASM)" (links added) and that they would like to interview me and the rest of the SPI team. I'm guessing I'm not the only one to receive it. My thoughts are:

  1. If anyone does plan to grant this interview, they should be 100% sure that it's real. IMO, the easiest way would probably be to contact ISD directly rather than reply to the email sent via Wikipedia.
  2. Likewise an interviewee should make sure they trust that ISD/CASM is operating in good faith here. I don't mean to impugn either organization's name, but the things they want to ask about are the exact same things that some bad actors would love to ask about.
  3. Finally, one should of course be careful not to compromise any sources and methods, including being wary of the possibility that if one omits certain details of something, but another clerk is also interviewed and omits different details, it may be possible to piece together "classified" information.

I don't have much to say on the topic in question, and even if I did, don't trust myself to walk the line perfectly. The last time I was interviewed by an NGO, I listened as the interviewer then called his boss to ask how to work around the fact that I hadn't said what he wanted to hear, so perhaps I'm a bit cynical here. I thus don't intend to grant any interview, but I note all this for anyone who does. -- Tamzin[cetacean needed] (she/they) 22:07, 17 December 2021 (UTC)Reply[reply]

These folks are nothing if not persistent, they've already contacted arbcom and the functionaries, but I don't think any of us went for it either. Beeblebrox (talk) 22:48, 17 December 2021 (UTC)Reply[reply]
Yup, I got the same email, and I'm pretty sure it's legit. They say the interview will be for research only, though I don't intend on accepting as I do not have experience with said topic in question. Sro23 (talk) 01:23, 18 December 2021 (UTC)Reply[reply]
No one's said what the topic is.--Bbb23 (talk) 01:31, 18 December 2021 (UTC)Reply[reply]
Why don't I just forward you the email? Check your inbox. Sro23 (talk) 01:52, 18 December 2021 (UTC)Reply[reply]
Thanks, Sro23.--Bbb23 (talk) 02:55, 18 December 2021 (UTC)Reply[reply]
(edit conflict)yeah, I'm very curious about the topic as well. Maybe it's about "getting rid of a smelly dead body without being noticed by nosey neighbours" or "eating junk food without letting your roommate know" idk. What's common between arbs and SPI clerks? Or maybe they simply want answers without doing hardwork. You know, like somehow getting answer from your homework tasks. "if you are so smart, then tell me the cube root of 2197". post ec after Sro's comment: that's really fishy. I say we simply ignore them. —usernamekiran • sign the guestbook(talk) 02:03, 18 December 2021 (UTC)Reply[reply]
In the normal course of what might be called "legitimate" or "scholarly" research about Wikipedia, the research organization is normally reaching out to the WMF first, posting on a publicly readable mailing list, or alternately is posting publicly on the user talk pages of people they would like to interview. In this case, all of the outreach has been through private emails to either very limited groups, or to individuals. The messages specify that this is for "private" research purposes, and that the research will not be published or publicly available. It isn't specific enough about who their client is, which leads one to think that the client is not someone we as a community would want to provide with private research. It's pretty clear they're trying to find out what processes have been involved in disinformation and anti-harassment processes, again neither of which are really the kind of thing we're going to want to talk about with people who aren't being upfront about who they're working for. I'd discourage anyone from participating. Risker (talk) 05:26, 18 December 2021 (UTC)Reply[reply]

Thoughts on G6

Hi folks, Wikipedia:Sockpuppet investigations/EncyclopediaSupreme and Wikipedia:Sockpuppet investigations/Encyclopedia$upreme were already globally locked. I moved and closed, then was informed it’s a user who’s globally banned. Rather than fixing the redirects and listing for a merge, the pages were G6’d on my nomination as they are actually Wikinger, there were no enwiki blocks or tags, this user is managed 99.9% of the time by stewards as they're globally banned, and there was no real need to keep a ‘chain of evidence’ alive through the SPI page. In hindsight I’m not sure that was the right thing to do, looking for a second opinion if you wouldn't mind. Thanks, --Jack Frost (talk) 09:21, 23 December 2021 (UTC)Reply[reply]

I ask myself what if someone sees activity reminiscent of EncyclopediaSupreme, how will they tie it to Wikinger? The user isn't mentioned in the SPI. Locks are poorly signalled, and their reasoning seldom ever. The user isn't tagged. WP:DENY has a place, but we shouldn't help the user conceal their trail either. That's my 2¢. Cabayi (talk) 10:29, 23 December 2021 (UTC)Reply[reply]
Wikinger mostly impersonates LTA/GRP, to an extent where they are functionally indistinguishable to all but the most committed sock taxonomists – considering that GRP is so well-known and obviously disruptive that he gets (b)locked on sight anyway, I don't think it's worth cataloguing (or tagging) these socks. --Blablubbs (talk) 10:42, 23 December 2021 (UTC)Reply[reply]

Historical stats?

Do we have any historical data on the length of the various SPI queues over time? -- RoySmith (talk) 19:26, 19 December 2021 (UTC)Reply[reply]

I don't think I've seen any. Could be useful to start now? Dreamy Jazz talk to me | my contributions 22:13, 19 December 2021 (UTC)Reply[reply]
Yeah, that sounds like my next project :-) -- RoySmith (talk) 22:52, 19 December 2021 (UTC)Reply[reply]
@RoySmith you could always scan the SPI case table pages (like User:Mz7/SPI case list) for at least historical data. This would provide a way to get the historical stats without having to parse the wikitext, though for bot outages this won't show the full data. Perhaps combining the data from multiple sources (as I know of at least 2 SPI case table bots which have been operating recently) would address that issue. Dreamy Jazz talk to me | my contributions 21:10, 24 December 2021 (UTC)Reply[reply]

Seeking statistics for SPI investigations/confirmations

I would like to know if there are stats available for the number of SPI cases that are filed every year, and preferably also the number of confirmed cases per year.

Searching both Wikipedia and Wikimedia hasn't yielded me any satisfactory results, although I did find this outdated, inactive, non-statistical archive. This surprises me, because it's so easy to find stats about how many edits are made on Wikipedia, how many users there are, how many pages are created every day, etc. But nothing about the number of sockpuppets identified.

To me, this is both a vulnerability for Wikipedia, and a disservice to the extremely talented and dedicated CheckUsers who work day and night to resolve these cases. We have to be aware of the scope of sockpuppetry on Wikipedia, if we are to use the full potential of Wikipedia's administration, because without data or facts the clerks are fighting blindfolded against an invisible enemy.

My search outside Wikipedia has yielded some results. For example, Case Study of Sockpuppet Detection on Wikipedia" (2013) states: "Sockpuppets are a prevalent problem in Wikipedia, there were close to 2,700 unique suspected cases reported in 2012. "

However this figure of ~2,700 unique socks is uncited in this study, and is fairly old. Nevertheless it does indicate that at some point in 2013, someone was aware of the total number SPI socks identified for the previous year (however they found that out).

If anyone can point me in the direction of this kind of statistical reference, I would be greatly appreciative. It wouldn't be the first time I missed something right under my nose on Wikipedia. But if it isn't available, I think it's long overdue to create it, though admittedly I say that as a person who lacks the skill to do anything like that, and I know SPI clerks are already holding up half the world as it is.

Thanks to all for taking the time to read this request, and I wish you a Happy New Year. Hunan201p (talk) 18:09, 24 December 2021 (UTC)Reply[reply]

I do not know of such a list. However see the above section § Historical stats? is along a similar vein. Dreamy Jazz talk to me | my contributions 21:06, 24 December 2021 (UTC)Reply[reply]
  • What Hunan201p wants to do with these statistics makes no sense to me. Above they say ignorance of these statistics means "the clerks are fighting blindfolded against an invisible enemy." Why? How would these stats help anyone fight socking? On my Talk page, Hunan201p said something similar: "I think it's preposterous if Wikipedia doesn't compile data about how many SPI cases are filed yearly. It would be like if the USA didn't compile crime statistics. If there's no data available on the scale and extent of sockpuppetry here, how can we even address the problem?" Again, I don't see that these stats would help anyone fight sock puppetry. I think this is a waste of Hunan201p's time and a waste of ours.--Bbb23 (talk) 21:21, 24 December 2021 (UTC)Reply[reply]
@Bbb23:If we are able to quantify sockpuppetry on Wikipedia in recent years, and compare it to years prior, we may find that sockpuppetry has risen or held steady over time -- while the number of editors as a whole has decreased.
That would be hugely relevant. We would be able to demonstrate that Wikipedia's current strategy against sockpuppetry is not sufficient to reduce sockpuppet abuse.
The impression that I get from the patrollers is that this is a spiritually exhausting up-hill battle against sockpuppets. It's a never-ending game of cat and mouse, where the cat is blind and the mice just won't stop spawning. This is cruel to the patrollers and the CheckUsers who want to improve this website by stopping sockpuppet abuse. A new approach is needed.
But before we can even talk about addressing sockpuppet abuse, we have to have data. We have to know what is going on, or we can't address this problem, because all proposals require evidence. And besides, Wikipedia is supposed to be transparent. As the world's largest encyclopedia, Wikipedia has a responsibility to let the public know the full extent of its sockpuppet problem and to ensure that it is achieving results. From the responses I've gotten so far, it looks like Wikipedia is not actively monitoring and documenting the rate of sockpuppetry in the long term, meaning there's no credible analysis of what is happening. Hunan201p (talk) 21:59, 24 December 2021 (UTC)Reply[reply]
I think analysis of the number of cases for most editors would be a purely interesting stat. Your point on quantifying the amount of socks may be useful, but the activity at SPI is based on the number of editors finding/reporting socks and not on the number of socks present. Although these two do effect each other, they are not equal and thus it may be that in the past less socks were found but now more are found though the number of socks has stayed relatively constant / equal to the increase/decrease in editor numbers. Dreamy Jazz talk to me | my contributions 22:13, 24 December 2021 (UTC)Reply[reply]
@Hunan201p and Dreamy Jazz: for a very, very approximate set of stats for the past four years, see Quarry 60878. It's just a count of how many userpages have been created in any given year by someone other than the userpage's owner. Obviously there will be loads of false positives and false negatives, but I speculate the rate of them will be relatively consistent over this time period, so I think the trendline is meaningful. (The page creation log came into existence on 2018-06-27, which is why 2018 is so light. A query going back further than that would require parsing revision, a very very large table, so I'm not sure it can be done within Quarry's time constraints.) -- Tamzin[cetacean needed] (she/they) 22:24, 24 December 2021 (UTC)Reply[reply]
Wait, those numbers look way too high... re-assessing my SQL. -- Tamzin[cetacean needed] (she/they) 22:31, 24 December 2021 (UTC)Reply[reply]
I suspect a better way would be to process the SQL XML dumps. The disadvantage compared to Quarry is that you can't just write some SQL and fire it off. The advantage is that the data is presented in a way that makes it possible to parallelize. There is a moderately high activation energy barrier, i.e. you need to gear up to build a VPS environment to access the data, but I've done that already. -- RoySmith (talk) 22:36, 24 December 2021 (UTC)Reply[reply]
@Dreamy Jazz: Point taken. Indeed, SPI cases themselves do not necessarily constitute factual sockpuppetry -- only CheckUser activity. However it could be just as easy to take a sample of known sockmasters and observe the rate of sock confirmation over time, no? As always, I appreciate input from people who actually know the system as you do (I do not).
@Tamzin: Such a fascinating idea and helpful reply, thank you for it.
@RoySmith: My faith in Santa is restored! Hunan201p (talk) 22:48, 24 December 2021 (UTC)Reply[reply]
There's still a piece of pie waiting for you at my talk page, but at this point, you better hurry or it's going to be all gone before you get a piece. -- RoySmith (talk) 22:54, 24 December 2021 (UTC)Reply[reply]
My grandfather likes to say: "I've only been wrong once. You see, I thought that I was wrong about something... But it turned out that I was wrong about that." On that note, looking at this snapshot of results for last New Year's, it seems my SQL was true, and my assumption correct that the vast majority of such creations are sock-taggings. 53 results, and I only spot two false positives. Times 358, that gets us 18,974, so if you factor in seasonal variation in editing, 17,437 seems plausible for our year-end total. If anything, I'd guess that the number of intentional no-tags exceeds the number of non-SPI-related userpage creations by other users, putting us maybe in the vicinity of... 20,000 sockblocks this year? And definitely with an upward trend. -- Tamzin[cetacean needed] (she/they) 23:12, 24 December 2021 (UTC)Reply[reply]
I'm sure it doesn't need pointing out to regulars here (including Tamzin) that these numbers are going to be way way off. I would instead suggest collating the block logs for "abusing multiple accounts", "block evasion", "long term abuse", and anything mentioning either "checkuser" or "sock". These are fairly consistent long-standing block reasons and may be enough to provide a general trend - still nowhere near a comprehensive and meaningful number. How does one then quantify IP socking? I dunno. -- zzuuzz (talk) 23:36, 24 December 2021 (UTC)Reply[reply]
Just a bit of shameless promotion of - I have ~145 days of per-day SPI data, which isn't what y'all looking for, but may be useful? -- TNT (talk • she/they) 00:06, 25 December 2021 (UTC)Reply[reply]

@Zzuuzz: Funny, I had the same thought process. Pre-edit-conflict post begins:

Oh right, that only counts ones that started as redlinks, so I should have guessed well above 20,000. I just had a brief moment of panic when a block-log-based query showed 165,652 this year, but thankfully I realized that that was largely due to the string socks4 appearing in many ST47ProxyBot blocks. With that regexed away, we're left with Quarry 60880. This data will still have some false positives (overturned blocks, maybe the occasional string containing "sock" for some other reason) and false negatives (blocks without summary, just containing the master's name, etc.), but I reckon is the best you'll get, Hunan201p, without going through things with a fine-toothed comb.

-- Tamzin[cetacean needed] (she/they) 00:15, 25 December 2021 (UTC)Reply[reply]

@Tamzin:I simply can't express how much gratitude I have for your work. In a matter of a couple of hours you have already shown us data I expected to wait months for. You and RoySmith have offered us a treasure trove of data, and the implications of it are very alarming. Although you have said these numbers may not be completely accurate, the 105% increase in sock-related bans from 2013-2021 speaks for itself. During that same time period the number of minimally active Wikipedians dropped from around 75k to 50k.
The rate of sockpuppetry on Wikipedia must therefore exceed the rate of rape, homicide, or even shoplifting in any country, and any warzone, on the planet. That's a crisis. You folks have a job that is more suited to major law enforcement agencies like the FBI and Interpol, rather than a handful of unpaid volunteers. To me, that is an outrage. The bad guys have been given a license to vandalize Wikipedia, while checkusers try to stop that with limited resources and strict procedural doctrines, and without any concrete action to raise the standards of how people edit high-importance Wikipedia articles. I cannot remain silent about this anymore. I'll wait for the bigger minds to sort through the other sources of data, but I think we are well on our way here to forming a case for changing how Wikipedia is edited. I salute all of you.
@TheresNoTime:@Zzuuzz:These are also fantastic routes to explore and I thank you both for offering your talents. Hunan201p (talk) 00:51, 25 December 2021 (UTC)Reply[reply]

@Hunan201p: Before you get too carried away, it's worth noting that more than half of the blocks this year were of sleeper accounts, ones that have never made edits. That doesn't mean the blocks are insignificant, but it does mean that it's not like 45,479 accounts separately disrupted Wikipedia this year as part of sockpuppetry. It was "only" 18,682. See Quarry 60885 for the year-by-year non-sleeper blocks, Quarry 60883 for monthly sockblocks in 2021, and Quarry 60884 for monthly non-sleeper blocks in 2021. If you want to experiment with broader definitions of "sleeper"—say, counting anyone with less than 10 edits—fork either of the sleeper queries and change that > 0 to > 10 or such. No SQL knowledge required. :)

-- Tamzin[cetacean needed] (she/they) 09:34, 25 December 2021 (UTC)Reply[reply]

Whoops, I wasn't filtering enough of the proxy false positives. Still figuring out how to adjust for that. Probably accounts for ~75% of what I'd been considering sleeper blocks. This mistake doesn't affect the non-sleeper figures, though, since registered accounts don't get proxy-blocked. -- Tamzin[cetacean needed] (she/they) 10:20, 25 December 2021 (UTC)Reply[reply]
Well, that's a huge relief. Thanks for correcting me. The rate of non-sleeper blocks is definitely not as severe as I thought, only a 45% increase from 2013-2021. Still very significant, though. Earlier in the previous decade Wikipedia was hovering in the lower 10000s, whereas later on in the decade it crossed the 20000 threshold.
My previous comments about crime rates were ridiculous, but it's still plain as day that Wikipedia has a problem, including the sleeper blocks, who deny Wikipedia access to legit would-be editors who had to use VPNs to access the site because it's banned in their country.
Thanks again for taking the time to break the numbers down for me and show that I was wildly overexaggerating the problem.Hunan201p (talk) 12:51, 25 December 2021 (UTC)Reply[reply]

Please don't edit the (SPI) archives

I've written an essay WP:Please don't edit the (SPI) archives (shortcut WP:DETA) which explains why we don't want people editing the archives. I suggest linking to it in edit summaries when reverting somebody's failure to heed that suggestion. -- RoySmith (talk) 15:41, 27 December 2021 (UTC)Reply[reply]

Looks useful. Thanks. Dreamy Jazz talk to me | my contributions 17:58, 27 December 2021 (UTC)Reply[reply]

I'm pretty sure Category:Suspected Wikipedia sockpuppets of confirmed is a mistake, can somebody sort that out? Pahunkat (talk) 21:01, 31 December 2021 (UTC)Reply[reply]

Thanks Pahunkat, I've fixed them and nuked the category. GeneralNotability (talk) 21:03, 31 December 2021 (UTC)Reply[reply]

Bots vs the archives?

How do we feel about bots editing the archives? See Special:Diff/1043690177. This particular edit doesn't seem to have done any harm, but I'm dubious about the general practice. -- RoySmith (talk) 14:32, 5 January 2022 (UTC)Reply[reply]

I'm not really sure why there's such a strong desire against editing the archives in any way. Reporting new socks, discussing decisions and generally editing or adding comments shouldn't be happening in the archives. Edits by bots or technical editors doing some minor piece of cleanup such as fixing lint errors or replacing templates to be deleted I see nothing wrong with. Those edits will never change the substance of the discussion so I don't see the issue. --Trialpears (talk) 14:39, 5 January 2022 (UTC)Reply[reply]
Likewise I don't see the issue. The point of limiting access to the archives is to avoid disruption, and an approved bot run will (hopefully) not bring disruption, and may have a benefit to SPI team members, to avoid situations where the parser changes how it handles some form of improperly-closed tag and suddenly the whole page is colored the deeppink of my signature. -- Tamzin[cetacean needed] (she/they) 22:50, 5 January 2022 (UTC)Reply[reply]
(edit conflict) I think we should outsource the washing/drying or our dirty laundry.--Bbb23 (talk) 14:41, 5 January 2022 (UTC)Reply[reply]

SPI Zero?

@Mz7: the software engineer in me is always on the lookout for interesting edge cases to verify via tests. So, I gotta ask, have you ever tested the code which generates User:Mz7/SPI case list when there's no input? Because I think that's going to happen soon. A good problem to have, right? -- RoySmith (talk) 14:12, 6 January 2022 (UTC)Reply[reply]

A software engineer walks into a bar. Orders a beer. Orders 2 beers. Orders 2^64+1 beers. Orders -1 beers. Orders beers. Orders 0 beers. Orders -0 beers, -0.0 beers, and -0.0j0e0 beers. Orders "beer" beers. Orders NaN beers, None beers, and Undefined beers. Leaves the bar. A regular person comes in and asks for the bathroom. The bar burns down.
Seriously though, this is exciting! Glad I haven't had time yet to do my next big time-sink filing. (But just you wait, someone will drop some massive UPE filing or find the next Icewhiz sock a minute before PaullyMatthews is closed... kinehore.) -- Tamzin[cetacean needed] (she/they) 14:38, 6 January 2022 (UTC)Reply[reply]
... 9 accounts with a combined 2,500 edits. 2 minutes after I said that. I'm sorry. :( -- Tamzin[cetacean needed] (she/they) 14:47, 6 January 2022 (UTC)Reply[reply]
See photo for an example of somebody who didn't do enough testing. I found this a few years ago in a gym. I tried doing NULL, but I could never manage to get in 12 good reps. -- RoySmith (talk) 14:49, 6 January 2022 (UTC)Reply[reply]
@RoySmith: Heh. I hadn't ever tested it directly before, so I went ahead and did it just now. It works! Specifically, at this line, where I initialize a category object from Category:Open SPI cases, I replaced the category with the currently empty Category:Requests for checkuser to see how the code would run against an empty category, and the result is an appropriately empty table. In my ~4 years of being an active SPI admin, I don't think I have ever seen "SPI Zero". The closest we ever got was this point with 3 entries in the case table on October 30, 2019 Mz7 (talk) 01:42, 7 January 2022 (UTC) correction: the next revision on the same day has only 2 entries in the table. Mz7 (talk) 04:01, 7 January 2022 (UTC)Reply[reply]
Part of me now wants to insert some kind of special easter egg into the bot code that inserts some fun image to the page if it detects no cases. Mz7 (talk) 01:53, 7 January 2022 (UTC)Reply[reply]
Support per nom. -- RoySmith (talk) 02:00, 7 January 2022 (UTC)Reply[reply]
Yes please. Dreamy Jazz talk to me | my contributions 14:15, 7 January 2022 (UTC)Reply[reply]
Perhaps an image of socks, all matched and orderly tucked away in a drawer?-- Jezebel's Ponyobons mots 15:10, 7 January 2022 (UTC)Reply[reply]
Went to find us an image and made the mistake of clicking on c:Category:Women's socks thinking hey, let's promote some diversity in sock type. Jesus, it's... At least only two of them are actual porn. Are Commons contributors aware that women wear socks in just, like, normal ways? I'd upload some pix of just regular women's socks but I really don't want those to be my first Commons uploads in years. Anyways uh... I actually don't see a pic of a tidy sock drawer anywhere in c:Category:Socks. More important images to categorize, I guess. -- Tamzin[cetacean needed] (she/they) 15:24, 7 January 2022 (UTC)Reply[reply]
There's always socks. -- RoySmith (talk) 16:40, 7 January 2022 (UTC)Reply[reply]
I have made a modification to the bot such that if it detects no cases in Category:Open SPI cases, it will transclude User:Mz7/SPI 0. I went with a picture of an empty drawer with the text, "The sock drawer is empty." Maybe I should've left this as a surprise, but honestly, feel free to be bold and change that page if you have a more humorous idea. Mz7 (talk) 17:21, 7 January 2022 (UTC)Reply[reply]
@Mz7: Can we get some sort of "SPI Basically Zero" message for when the only cases left are on hold or moreinfo'd? Since sometimes there's really nothing we can do to clear out that queue, other than wait days to weeks for people to not respond. -- Tamzin[cetacean needed] (she/they) 00:31, 10 January 2022 (UTC)Reply[reply]
That's kind of like saying you want to be in the record books so badly, you don't mind if you have an asterisk next to your name :-) -- RoySmith (talk) 01:40, 10 January 2022 (UTC)Reply[reply]
Well I just gave the OK to closing my on-hold, Roy, so that just leaves yours. :P But like... What if someone comes along right as we're about to hit zero, and files a half-assed request that I can't quite decline but also can't action, and then lets it sit as moreinfo for a week or two? What then? And I mean, last I checked Houston still gets to have the trophy and the pennant. ... Wait dammit I was about to hit enter and there you closed Kurdiyate. I take it all back. True SPI Zero is the only SPI Zero. Dodgers won in '17. Someone close INTSF and Huyu! -- Tamzin[cetacean needed] (she/they) 01:55, 10 January 2022 (UTC)Reply[reply]
And there goes Huyu too. -- Tamzin[cetacean needed] (she/they) 01:57, 10 January 2022 (UTC)Reply[reply]
Perhaps we should just get you to WP:RFA this week. I have a feeling that would solve the frustration you're facing. ;) Mz7 (talk) 02:04, 10 January 2022 (UTC)Reply[reply] pressure to the people Girth Summit emailed, but could they hurry it up and get back to him so that we can close and archive ineedtostopforgetting? :) Sro23 (talk) 02:10, 10 January 2022 (UTC)Reply[reply]

Small request

If pecople could leave 10ish cases open this weekend (15-16 Jan) I would really appreciate it. I'm running a CU training for new arbs and new CUs that morning, and it is much easier to do if I have a few options to go through. Without going into too much detail, different types of cases are easier to give group trainings on than others, so I typically prefer to have a few options. Normally this isn't a problem because I just but 2-3 on hold a few days before, but... doesn't look like that's the case here. TonyBallioni (talk) 02:19, 10 January 2022 (UTC)Reply[reply]

I mean I guess I can spin up a few more sockpuppets... Any requests? Nationalists? Caste warriors? Scatological vandals? Railfans? /j -- Tamzin[cetacean needed] (she/they) 02:25, 10 January 2022 (UTC)Reply[reply]
Just pick a few members of WPTC at random and CU them, I'm sure you'll find some socks. GeneralNotability (talk) 02:31, 10 January 2022 (UTC)Reply[reply]
Okay. I will leave CU cases until there are around 10 active open CU cases. TonyBallioni, do you think it is a good idea to put these 10ish cases on hold while you get a chance to review them and decide which you want to select for training? Just that everyone may not see this message, so the CU case may be declined by clerk/CU or the check ran before the weekend Dreamy Jazz talk to me | my contributions 10:12, 10 January 2022 (UTC)Reply[reply]

We did it

Screenshot for posterity. Mz7 (talk) 07:20, 10 January 2022 (UTC)Reply[reply]
  • Nice work, folks! Mz7 (talk) 07:20, 10 January 2022 (UTC)Reply[reply]
    • Congratulations everyone. We beat the game, now all we need is for no one to sock ever again... :) Spicy (talk) 07:36, 10 January 2022 (UTC)Reply[reply]
      Well I guess Wikipedia is over then, right? It was a fun encyclopedia-themed pattern-recognition MMORPG while it lasted. -- Tamzin[cetacean needed] (she/they) 08:12, 10 January 2022 (UTC)Reply[reply]

Twinkle request: vote this up, please.

There's a feature request open at which would eliminate the number of duplicate SPI reports we get. Please take a look and <canvas style="shameless" location="github">vote it up!</canvas>. -- RoySmith (talk) 17:13, 13 January 2022 (UTC)Reply[reply]

Continued disruption by DarkMatterMan4500

I'm happy to escalate this to ANI if needed, but given that the issue at hand is a user ignoring a clear directive from two CUs to stay out of SPI clerk-like work (or SPI at all), I bring it here for review first. -- Tamzin[cetacean needed] (she/they) 15:07, 23 January 2022 (UTC)Reply[reply]

Blocked two weeks. (And also handled the G6) -- ferret (talk) 15:11, 23 January 2022 (UTC)Reply[reply]
Good block. Dreamy Jazz talk to me | my contributions 16:34, 23 January 2022 (UTC)Reply[reply]
Fine with me, thanks for the ping. Enterprisey (talk!) 03:52, 24 January 2022 (UTC)Reply[reply]

Could we perhaps downgrade the protection on Wikipedia:Sockpuppet investigations/SPI/Blank report template header (edit | talk | history | links | watch | logs) to TPROT? I'm chasing down the various bugs we've been having causing excess whitespace (one done, one PR pending, one someone beat me to it), and the fourth (and I hope final) lies in that projectspace quasi-template. Yes I could just FPER this, but I don't think there's any particular reason that it's not TPROT'd, other than that non-templatespace pages acting as templates weren't picked up in the mass TPROT downgrade of 2014. Plus, asking this way lets me procrastinate figuring out what exactly I need to change. :D -- Tamzin[cetacean needed] (she/they) 14:52, 3 February 2022 (UTC)Reply[reply]

Done. -- RoySmith (talk) 15:27, 3 February 2022 (UTC)Reply[reply]
Thanks. :) I'll try not to blow anything up too hard. -- Tamzin[cetacean needed] (she/they) 15:29, 3 February 2022 (UTC)Reply[reply]
This has gotten me onto a kick of SPI template improvements, and I may have my eyes on Wikipedia:Sockpuppet investigations/SPI/Inputbox blank report for ordinary use (edit | talk | history | links | watch | logs) soon. Would it be possible to downgrade that as well, Roy? (Meanwhile Wikipedia:Sockpuppet investigations/SPI/Inputbox blank report for ordinary use for IPs (edit | talk | history | links | watch | logs) may need an upgrade.) -- Tamzin[cetacean needed] (she/they) 21:00, 3 February 2022 (UTC)Reply[reply]
@Tamzin: Done. Mz7 (talk) 04:09, 4 February 2022 (UTC)Reply[reply]
I've been wondering. Recently WP:RFPP got a bit of a makeover, moving away from these template-style forms and towards this kind of JavaScript form. While we are discussing improvements to this flow, I wonder if something like this would be desirable for SPI reports? Mz7 (talk) 04:11, 4 February 2022 (UTC)Reply[reply]

Proposed change to the sock list in Template:SPI report

As teased in the above section, please see Template talk:SPI report § Proposed change to sock list. -- Tamzin[cetacean needed] (she/they) 04:47, 5 February 2022 (UTC)Reply[reply]

"Behavioural evidence needs evaluation" status for SPI case status

Could we add a new "behavioural evidence needs evaluation" status to Template:SPI case status? Perhaps something like the "clerk assistance required" status but with "Behavioural evaluation is requested in the case below." and its own category? -- TNT (talk • she/her) 11:47, 10 February 2022 (UTC)Reply[reply]

I'd support this. Dreamy Jazz talk to me | my contributions 13:55, 10 February 2022 (UTC)Reply[reply]
This is basically just "checked" status, checkusers should generally only use it when they've checked a case and need a patroller to evaluate behaviour (unless we set it to something more specific like "clerk" or "hold", or just close it ourselves). Maybe we just need to replace "checked" with this new status flag. Ivanvector (Talk/Edits) 14:23, 10 February 2022 (UTC)Reply[reply]
Under what circumstances would this new status be used?--Bbb23 (talk) 14:31, 10 February 2022 (UTC)Reply[reply]
@Bbb23: Namely in situations where a checkuser declines to block based on the evidence (a  Possible for example) and then uses the  Behavioural evidence needs evaluation template to request a patrolling clerk/admin to evaluate the behaviour. Noting this in the status table just "makes sense" to me Face-smile.svg -- TNT (talk • she/her) 14:39, 10 February 2022 (UTC)Reply[reply]
Related to Ivanvector's comment, would there be times when you would use "checked" and other times when you would use this new status? Would some findings trigger one and other findings trigger the other?--Bbb23 (talk) 14:44, 10 February 2022 (UTC)Reply[reply]
I'm not violently opposed, but as others have noted, I don't see how this is significantly different from "checked" to need a new section. Anyway, we've run out of colors. -- RoySmith (talk) 15:11, 10 February 2022 (UTC)Reply[reply]
Well I'm certainly more concerned about the lack of colours 😋 -- TNT (talk • she/her) 15:23, 10 February 2022 (UTC)Reply[reply]
Yeah, I also don't think this is necessary. If an SPI status is set to "checked", it already means a CU was completed and the next step is behavioral evaluation—the {{behav}} template is usually a redundant call-out. If anything, there's an argument to be made that the "checked" and "declined" categories could be done away with and merged with the standard "open" category. Mz7 (talk) 20:44, 10 February 2022 (UTC)Reply[reply]
I find that the "open", "declined", and "completed" categories all somewhat understate or misstate what's actually needed, and, especially in the latter two cases, potentially at the risk of people ignoring the SPI. So, mostly per Mz7, what would make sense to me would be to have it (all with the same color currently used for "open"), a single category of "behavioral evaluation", with subcategories "behavioral evaluation (CU completed)", "behavioral evaluation (CU declined)", and "behavioral evaluation (CU declined by CU)". This could have the added benefit of making it more obvious to people that they haven't requested CU, since, as much as unexplained CU requests are a problem, we also get cases that are clearly meant to be CU requests but aren't marked as such. -- Tamzin[cetacean needed] (she/they) 21:55, 10 February 2022 (UTC)Reply[reply]
I think our longstanding way of doing things has worked sufficiently enough that I don't think we urgently need to carry out such a significant reform like this at the moment. Most of SPIs problems probably trace back not to faulty procedures but a lack of interested administrators, checkusers, and clerks. In the past six months, however, we've done a decent job at appointing excellent new checkusers and clerks, and this is reflected in the optimal condition of the current backlog. (If you don't remember, this is what the backlog looked like in early August 2021.) As a middle ground, we could perhaps reword the text produced by Template:SPI case status to be clearer about the next step in the case:
  • A checkuser has completed a check on relevant users in this case, and it is now awaiting administration and close.A checkuser has completed a check on relevant users in this case, and it is now awaiting a behavioural investigation.
  • This SPI case is open.This SPI case is open and awaiting a behavioural investigation.
Mz7 (talk) 23:54, 10 February 2022 (UTC)Reply[reply]
Regardless of the addition / removal of statuses, I think this wording update is useful. Dreamy Jazz talk to me | my contributions 00:47, 11 February 2022 (UTC)Reply[reply]

SPI-Tools breakage

As some of you may have heard, there was a botnet attack against wikipedia yesterday. One of the emergency fixes put in to place to ward off the attack broke the way spi-tools handles logins via oauth. It will continue to run in anonymous mode (with reduced functionality), and if you're already logged in, you should be fine as long as you don't logout and try to login again. My understanding is that this will get resolved sometime today. -- RoySmith (talk) 14:58, 18 February 2022 (UTC)Reply[reply]

SPI undeleted

I've taken to doing temporary undeletes of pages referred to in SPI reports so non-admin clerks can compare them, as is commonly done at WP:DRV with {{tempundelete}}. To make that easier, I've created {{SPI undeleted}}. People with more template skills than I have should feel free to improve it. -- RoySmith (talk) 16:12, 19 February 2022 (UTC)Reply[reply]

Trialpears appointed trainee clerk

I am pleased to announce that Trialpears (talk · contribs) has been appointed as a trainee clerk. Blablubbs and I will be supervising their training. Welcome to the team, Trialpears! We're looking forward to working with you. Best, KevinL (aka L235 · t · c) 12:10, 8 March 2022 (UTC)Reply[reply]

Glad to have you on board, thanks for volunteering. -- RoySmith (talk) 17:15, 8 March 2022 (UTC)Reply[reply]

Non-cases moved to different case archives

For background, Wikipedia:Sockpuppet investigations/Lillyput4455 is a page where many reports have been filed over the years, but have never resulted in a finding against the named sockmaster. For several years the reports were tangled up with Pakistanpedia but more recently they've been named in a report which turned into Wikipedia:Sockpuppet investigations/Ak131001. So now there is history of investigations for Lillyput in the archives of both of those cases, but {{SPI archive notice}} can only refer to one or the other. What's the best way to handle this? I'm leaning towards deleting the Lillyput4455 case page, since there has never been a finding and they're still active. Ivanvector (Talk/Edits) 15:10, 11 March 2022 (UTC)Reply[reply]

I'm not sure that deleting it would be the best thing. There's links to it from various cases already. As confusing as things are now (and they're plenty), if you delete it, those links will be orphaned, which will be even more confusing. -- RoySmith (talk) 15:30, 11 March 2022 (UTC)Reply[reply]

Pahunkat appointed trainee clerk

I am pleased to announce that Pahunkat (talk · contribs) has been appointed as a trainee clerk. RoySmith and I will be supervising their training. Welcome to the team, Pahunkat! I'm looking forward to taking my first trainee through the process - we'll both be learning as we go - but we have Roy for backup, so I'm sure we'll be fine. Girth Summit (blether) 09:28, 10 March 2022 (UTC)Reply[reply]

Welcome to the team. :) --Blablubbs (talk) 10:53, 10 March 2022 (UTC)Reply[reply]
Welcome, Pahunkat! KevinL (aka L235 · t · c) 21:49, 10 March 2022 (UTC)Reply[reply]
Thanks! I'm looking forward to working with everyone :-) Pahunkat (talk) 22:29, 10 March 2022 (UTC)Reply[reply]

Thanks Spicy and TheSandDoctor

A big shout-out to Spicy and TheSandDoctor who both put in a crazy amount of work the past few days clearing out the huge pile of cases that needed archiving. It's not the most interesting job to do, but it needs doing. So thank you for taking it on! -- RoySmith (talk) 16:31, 11 March 2022 (UTC)Reply[reply]

+1. I noticed the many archives in my watchlist today after I got back, and it must have taken a while. Dreamy Jazz talk to me | my contributions 19:38, 12 March 2022 (UTC)Reply[reply]
@RoySmith and Dreamy Jazz: Thank you! Archiving is just what I like to do haha...I guess that makes me a wikignome of wikignomes Face-tongue.svg? Pinging Spicy so they see this thread. --TheSandDoctor Talk 21:54, 14 March 2022 (UTC)Reply[reply]

Spi-helper problems

There's an issue going on with spi-helper.js which causes the script to be unable to parse the date headers for the various sections, so you only get "All Sections" in the pop-up menu. Tamzin and I have been looking at it. We're pretty sure this is caused by a change to the MediaWiki parsing API which was rolled out yesterday. Any SPI page which hasn't been edited since before the most recent MediaWiki deployment will not exhibit the problem because the unbroken API result is cached on the server side, but as soon as you edit a page, the cache will get flushed and the spi-helper will be broken for that page. At least that's the way it looks. I'm going to bed now. Apparently Tamzin is continuing to look at code. The best I can tell you at this point is Spicy's request that, "no one sock while SPIhelper is down, ok?" -- RoySmith (talk) 05:54, 18 March 2022 (UTC)Reply[reply]

I've put in this emergency PR for spihelper for now. In touch with a developer as to whether this behavior (probably a result of this) should be considered a bug or intended, but off to bed, so will have to wait till the morning to follow up. -- Tamzin[cetacean needed] (she/they) 06:35, 18 March 2022 (UTC)Reply[reply]
Oh, and for anyone who can't wait for the fix to be merged, it's available on-wiki at User:Tamzin/spihelper.js. (Pssst Girth Summit, re this ES.) -- Tamzin[cetacean needed] (she/they) 07:33, 18 March 2022 (UTC)Reply[reply]
OK, thanks for the update - I left that edit summary for GeneralNotability, who I e-mailed about the problem this morning before editing that case. Reassuring to know it's not just me being a doofus. Girth Summit (blether) 09:49, 18 March 2022 (UTC)Reply[reply]
@Tamzin thanks for the hotfix! Saved me mangling a case this morning... firefly ( t · c ) 10:12, 18 March 2022 (UTC)Reply[reply]
GN has merged the change; if anyone's still haveing this issue (including after refreshing your cache), please say so. For now, marking as resolved on spihelper's end. Will follow up with a Phab ticket presently to inquire as to whether this is intended behavior. -- Tamzin[cetacean needed] (she/they) 15:32, 18 March 2022 (UTC)Reply[reply]
Just to clarify, my part in this was noticing that something was broken and asking for help. Tamzin is the one who figured out the root cause, which is the hard part. -- RoySmith (talk) 20:26, 18 March 2022 (UTC)Reply[reply]


Horope (talk · contribs) is a sockpuppet of LastBreath64 (talk · contribs). Can someone merge Wikipedia:Sockpuppet investigations/Horope with Wikipedia:Sockpuppet investigations/LastBreath64? Renat 19:43, 6 April 2022 (UTC)Reply[reply]

@RenatUK The best thing would be for you to file a SPI on one or another of those cases and lay out your evidence there. Thanks. -- RoySmith (talk) 20:24, 6 April 2022 (UTC)Reply[reply]
@RoySmith: Special:Diff/1073710858 should be enough for a merge. --Blablubbs (talk) 20:26, 6 April 2022 (UTC)Reply[reply]
Hmmm. Let me see if I can flag down a friendly admin clerk who could do the merge :-) -- RoySmith (talk) 20:28, 6 April 2022 (UTC)Reply[reply]

tag-check.js vs cuStaleness.js

Those of you who run both of these scripts may have noticed that cuStaleness sometimes claims an account doesn't exist when it actually does, and a reload of the SPI page fixes the problem. Well, I finally figured out what's going on there, and I'll take the blame for breaking how cuStaleness parses the username. I'm working on a fix. -- RoySmith (talk) 16:55, 13 April 2022 (UTC)Reply[reply]

Fixed in Special:Diff/1082567281. -- RoySmith (talk) 22:13, 13 April 2022 (UTC)Reply[reply]
There are three hard problems in computer science: naming things, race conditions, and cache invalidation. Wait, let's run that again... There are three hard problems in computer science: naming things, cache invalidation, and race conditions. -- Tamzin[cetacean needed] (she/they) 22:22, 13 April 2022 (UTC)Reply[reply]

Bullseye down

I'm assuming it's related to the ongoing toolsdb maintenance -- RoySmith (talk) 15:15, 19 April 2022 (UTC)Reply[reply]

Does anyone use the UserCompare reports?

This thing. The thing that's updated hourly that seems to very often not be updated hourly, and which seems to not know to remove the 1= from template parameters.

I was thinking of adding links to spi-tools to the tool links bar in {{sock list}}, and while I'm at it was wondering if we still need that UserCompare link. -- Tamzin[cetacean needed] (she/they) 21:26, 24 April 2022 (UTC)Reply[reply]

I think I once clicked on UserCompare, and quickly figured out it wasn't of any value to me. I'd put it in the category of "Historical tools for which the author is to be commended, but better things have come along since then". And I think putting spi-tools in its place would be a good thing (full disclosure: I wrote spi-tools). Before you implement anything, let's chat off-line about a way to configure whether a user wants the dev or the production instance. -- RoySmith (talk) 14:43, 25 April 2022 (UTC)Reply[reply]
I generally used to run it on every case but it's been useful to me once or twice over the entire workload, so not that helpful clearly. --qedk (t c) 20:45, 27 April 2022 (UTC)Reply[reply]

Tamzin and Spicy promoted to be full clerks

The CheckUser team is pleased to announce that Tamzin (talk · contribs) and Spicy (talk · contribs) are promoted to be full SPI clerks. Congratulations, and thank you both for your dedication to SPI. Best, KevinL (aka L235 · t · c) 22:07, 7 May 2022 (UTC)Reply[reply]

Congratulations! --Blablubbs (talk) 22:09, 7 May 2022 (UTC)Reply[reply]
Let me also take this opportunity to thank and congratulate Blablubbs, who has done the vast majority of the training for both Tamzin and Spicy. Blablubbs is a terrific trainer and mentor, and has the process down pat. We're very lucky to have him on the clerk team. KevinL (aka L235 · t · c) 22:15, 7 May 2022 (UTC)Reply[reply]
Thank you for the kind words, they are deeply appreciated. I couldn't have asked for a better co-mentor, or for better trainees. Face-smile.svg --Blablubbs (talk) 22:23, 7 May 2022 (UTC)Reply[reply]
Whoa, cool! Do I get my own parking space now? -- Tamzin[cetacean needed] (she/they) 22:09, 7 May 2022 (UTC)Reply[reply]
sorry those are reserved by seniority KevinL (aka L235 · t · c) 22:25, 7 May 2022 (UTC)Reply[reply]
Isn't there some rule that says you're only allowed one promotion per account per calendar week? -- RoySmith (talk) 23:12, 7 May 2022 (UTC)Reply[reply]
If you can battle with the resident LTAs for one, then perhaps. Dreamy Jazz talk to me | my contributions 23:25, 7 May 2022 (UTC)Reply[reply]
Congrats! Dreamy Jazz talk to me | my contributions 23:26, 7 May 2022 (UTC)Reply[reply]
Congrats both! Well deserved. :) firefly ( t · c ) 09:48, 8 May 2022 (UTC)Reply[reply]
Thanks all. Spicy (talk) 15:17, 8 May 2022 (UTC)Reply[reply]
Tamzin, Spicy - Congratulations to you both! :-) ~Oshwah~(talk) (contribs) 03:17, 9 May 2022 (UTC)Reply[reply]
Many congratulations to you, young'uns! --qedk (t c) 15:17, 9 May 2022 (UTC)Reply[reply]

How do I become a clerk?

I am really sorry to use this board for asking this, but I didn't find a better place to ask this. I am really interested in SPI cases, and want to help out as a SPI clerk. There is lots to do in this field. So many cases need archiving, and a lot can be done too. I know it is not possible to be a clerk so easily, but atleast I can start with the process. There are many experienced users who watch this page, and that is why I have posted it here to ask for a suggestion. Please, am I suitable enough to be one of the trainee clerks at the least? What experience should I gain to be one of the clerks here? How many days at the least is needed to gain the experience? I would be really happy if these were cleared to me. And please forgive me, if I have posted this thing at the wrong place. Yours sincerely, ItcouldbepossibleTalk 06:01, 14 May 2022 (UTC)Reply[reply]

Hi Itcouldbepossible: I'm not a clerk myself, but before considering applying, I'd suggest getting familiar with SPI for a (long) while. A good starting point is User:Blablubbs/How to file a good SPI as well as everything linked from User:Blablubbs/How to file a good SPI#See also. MarioGom (talk) 10:36, 14 May 2022 (UTC)Reply[reply]
@MarioGom Thanks for the essays that you pointed me at. I see that you have requested for being an SPI clerk, which is I am quite sure, likely to be accepted. Now tell me something. You asked for that role, not to make a fool of yourself, that is in other words, knowing that there is a high chance that your request can get accepted. So how did you prepare yourself in a manner that your request gets accepted? How did you start out with the process? Maybe these questions might be a little foolish from my part, asking about "how he can prepare himself to be an SPI clerk with 8 months of experience", but still, please try and answer these questions for I am always a little 'inquisitive' about how I can improve my position here. ItcouldbepossibleTalk 13:10, 14 May 2022 (UTC)Reply[reply]
The best way to get accepted as a clerk trainee is to be active at SPI, adding value to the process. Mario already pointed you to Blablubbs's excellent essay, but I'll add that when I'm looking at a case, what I'm really looking for is diffs that make the socking obvious. So what you could do is read through the oldest cases and start slogging through the contribution histories of the various accounts looking for good example. Then, present them in the case in a clear way:
  • Here's account1 making a change: <link to diff>
  • Here's account2 making exactly the same change after the first one was reverted: <link to diff>
It's a lot of work to go through and find these diffs. By doing the legwork to lay this kind of evidence out clearly, you make the job of the clerks and CUs a lot easier. Do that enough and the CUs will start having conversations like, "Hey, I've seen ICBP around SPI a lot and they're doing good work, we should invite them to be a clerk trainee". -- RoySmith (talk) 14:07, 14 May 2022 (UTC)Reply[reply]
Here's a more specific example of what I was talking about. See Wikipedia:Sockpuppet investigations/백돌#14 May 2022. It's full of vague (and thus, useless) descriptions like similar vandalism to various articles. There's a bunch of links, but they're just links to history pages, not to specific diff pairs. If somebody wanted to make the case easier to process (i.e. get noticed as a potential trainee candidate), they could add something like this to the report: Namuwiki made this edit. After that was reverted, Naziwki made exactly the same edit. Also note the edit comment in the later edit. That would let a clerk or CU immediately zero in on the most significant evidence. This is what Show, not tell is all about. In this case, the usernames really make the connection obvious, but I just picked up this case and it made a perfect example. -- RoySmith (talk) 14:56, 14 May 2022 (UTC)Reply[reply]
I'll also plug Wikipedia:Advice for prospective SPI clerks as worthwhile reading. GeneralNotability (talk) 15:03, 14 May 2022 (UTC)Reply[reply]
Thanks a lot RoySmith to be patient enough to answer this beginner's question. Your detailed explanation explains me why you are an administrator and not me. Explanations detailed as these are really difficult to frame, but yet you presented it to me in such a simple and easy way. I will keep in mind about what you said. I am quoting a part of your answer here: So what you could do is read through the oldest cases and start slogging through the contribution histories of the various accounts looking for good example. Could you please explain me this? What does this specific phrase actually mean "start slogging through the contribution histories of the various accounts"? Which accounts should I see the contribution histories of? How do I find the account that should be examined? Could you please expand on these points? ItcouldbepossibleTalk 15:09, 14 May 2022 (UTC)Reply[reply]
Look at the big multi-colored chart on WP:SPI#Cases currently listed at SPI, and find the section of "Open" cases. These are the cases that people have already filed and are waiting to get worked on. Some of the cases look easy; Muhammeddhalifjr for example lays out the evidence in an easy to understand way with a nice table of easy-to-click-on diffs. On the other hand, Samirdu75 talks about filter logs, but doesn't actually link to them, so I need to go on a scavenger hunt. If you could add links directly to the filter logs, i.e [1], that would make the case that much easier for somebody to understand. -- RoySmith (talk) 17:08, 14 May 2022 (UTC)Reply[reply]
Hi RoySmith, I have tried helping in the Samirdu75 SPI case. Could I provide any help in the case? I understand that I am a beginner in commenting in SPI cases that are not in my area of interest. My area of interest primarily is Bengali film and TV shows and actors and actresses working in the area. I can comment on the behaviour of users who edit in these areas in a really well fashioned manner. For example, recently I opened a case against YAM, who doesn't exactly work in this area, but focuses on Indian TV, and thus made some edits to articles on Bengali TV shows, which attracted my attention. But in the Samirdu75 case, I tried my best to help as much as possible, and I will always be trying to do the same from now on. I will try and help in open SPI cases as much as possible. In this way, I will be familiarized with sockpuppets, and also get a chance to help the CUs and clerks to decide the result doing the least hard work. Sincerely, ItcouldbepossibleTalk 14:43, 15 May 2022 (UTC)Reply[reply]

SPI case list moved to project space

Cross-posting this here for visibility: Wikipedia talk:Sockpuppet investigations/Archives/Archive24#SPI case list moved to project space. Mz7 (talk) 00:22, 29 March 2022 (UTC)Reply[reply]

For those of you who are getting 500 error in spi-tools, this is why. I'm working on an emergency deployment to fix this. -- RoySmith (talk) 01:06, 29 March 2022 (UTC)Reply[reply]
OK, it's back up. Um, Mz7, next time, could you give people a heads up before changes like that, please? -- RoySmith (talk) 01:21, 29 March 2022 (UTC)Reply[reply]
@RoySmith: Whoops! Sorry about that. Looks like I was too ambitious and assumed too much here – I will definitely give a heads up for future changes like this. Mz7 (talk) 02:37, 29 March 2022 (UTC)Reply[reply]

tag-check.js broken, and fixed

Sometime yesterday, tag-check.js started to fail in mysterious ways, and the problem spread from case to case. To make a long story short, the problem was a change in Parsoid's behavior which turned out to be a legitimate change which my script couldn't handle. Individual user accounts became affected one by one (in a manner not unlike the zombie apocalypse) as various back-end caches timed out and the new parse output was exposed. Thank you to Izno and SSastry (WMF) to quickly figuring out what was going on and suggesting an effective fix. Said fix has now been deployed to User:RoySmith/tag-check.js. -- RoySmith (talk) 02:13, 14 July 2022 (UTC)Reply[reply]

Template:Sockpuppet category, which is used on sock category pages, uses {{user21}}. Would it be possible to add a link to checkuserwiki, similar to what shows up on {{checkuser}}? Best, KevinL (aka L235 · t · c) 19:21, 3 August 2022 (UTC)Reply[reply]

I don't see why not, but when I took a look at the source for {{user21}}, my head started to spin. I think all that has to be added is
[{{urlencode:{{{1|Example}}}}}&ns0=1&ns2=1&ns14=1 cuwiki]
but I'll leave it to somebody who speaks template better than I do to make the change. -- RoySmith (talk) 19:37, 3 August 2022 (UTC)Reply[reply]
@L235 and RoySmith: I've added this to Template:User21/sandbox and hid it for non-CUs. I know that it works but I am not sure if it will need consensus to be added. 0xDeadbeef 04:30, 4 August 2022 (UTC)Reply[reply]
Thanks. It's a shame we can't do this on Special:Contribs directly since that's MediaWiki hardcoded (see [2]) – unless a script could do that? Anyway, I started a brief discussion on the talk page and will implement the sandbox change in a few days if nobody objects. Best, KevinL (aka L235 · t · c) 07:22, 4 August 2022 (UTC)Reply[reply]
@L235: You might be looking for Template:Sp-contributions-footer. 0xDeadbeef 12:38, 4 August 2022 (UTC)Reply[reply]
Thanks. I was more referring to the header which is fixed. Best, KevinL (aka L235 · t · c) 23:20, 4 August 2022 (UTC)Reply[reply]

MarioGom as clerk trainee

I am happy to announce that MarioGom has been accepted as a clerk trainee. Mario has been doing some SPI for for a while, and brings to the table a solid understanding of how proxies are used in the real world by people trying not to be seen, and how to detect them. -- RoySmith (talk) 01:16, 8 September 2022 (UTC)Reply[reply]

Congrats on the clerk traineeship. Dreamy Jazz talk to me | my contributions 08:50, 8 September 2022 (UTC)Reply[reply]
  • Thanks Roy, Dreamy Jazz, Tamzin! "Awkward workplace introduction" goes here. MarioGom (talk) 06:12, 9 September 2022 (UTC)Reply[reply]
Welcome, MarioGom! Congratulations on the traineeship! --Jack Frost (talk) 03:55, 10 September 2022 (UTC)Reply[reply]

UserCompare out, spi-tools in?

There was a previous discussion about removing UserCompare from the {{sock list}} template, and adding spi-tools [3]. Nobody disagreed but participation was fairly low. I think it's a cheap and useful change. Can we make it? MarioGom (talk) 16:49, 9 September 2022 (UTC)Reply[reply]

Works for me :-) -- RoySmith (talk) 18:57, 9 September 2022 (UTC)Reply[reply]
I'm going to be bold and do this now. Dreamy Jazz talk to me | my contributions 21:35, 9 September 2022 (UTC)Reply[reply]
@Dreamy Jazz Actually, now that I think about it, it would make more sense on {{SPIarchive notice}}. It's per-case, not per-account. -- RoySmith (talk) 00:35, 10 September 2022 (UTC)Reply[reply]
But this is fine too. Thanks! -- RoySmith (talk) 02:23, 10 September 2022 (UTC)Reply[reply]
Thanks! MarioGom (talk) 17:48, 11 September 2022 (UTC)Reply[reply]

SPIs by PravinGanechari

I have concerns about PravinGanechari's engagement at SPI. This editor has been very prolific at SPI recently - within their rather short editing history about 9% of their total edits have been made to SPI, which is one of the highest proportions of anyone who is not an admin or clerk. Many of their SPIs are lacking in evidence or are nearly incomprehensible. See e.g. WP:Sockpuppet investigations/Siddhartha sengupta2001 (G6'd), WP:Sockpuppet investigations/Mickeydkop/Archive#19_August_2022, WP:Sockpuppet_investigations/Yash_Prasad9/Archive, WP:Sockpuppet_investigations/Syed_amjad08/Archive#03_September_2022, WP:Sockpuppet_investigations/Aarshshahproduction/Archive, WP:Sockpuppet_investigations/Aditya_rao63877, WP:Sockpuppet_investigations/AtharvVaibhavMinase (G6d). There's also this inappropriate comment. I will grant that some of their SPIs are actionable (perhaps because finding socks in the Indian TV topic area is like shooting fish in a barrel). However, in my opinion, far too many of the SPIs from this editor are inappropriate fishing expeditions or confusing messes that place an undue burden on a process that is already backlogged. It's my understanding that per Wikipedia:Sockpuppet_investigations/SPI/Clerk_and_checkuser_procedures#Patrolling, any full clerk may ask a user to stop contributing to SPI, and in my opinion this would be appropriate. However, since I've been involved with many of these user's filings I would like input from other members of the SPI team before proceeding. I've pinged the editor so that they're aware of this post. Thank you, Spicy (talk) 15:58, 16 September 2022 (UTC)Reply[reply]

I had also asked a question about this in the Tea house . Yes sir i have made some mistakes no problem you can take action. It would be best to block sir. Thank you PravinGanechari (talk) 16:16, 16 September 2022 (UTC)Reply[reply]

Blocking is an unusual suggestion. You will lose access to the Wikipedia Library. You can just stop filing SPIs if there is a consensus for you to do so. 0xDeadbeef 16:25, 16 September 2022 (UTC)Reply[reply]

It would be better to block me because my language is Kannada and Hindi. I can read and speak English perfectly, I find it difficult to write. There is a lot to learn in the SPI case. Yes, you allow me to take the help of a user for the language. So I promise I will not give you a chance to speak in SPI. PravinGanechari (talk) 16:42, 16 September 2022 (UTC)Reply[reply]

If you want, stop me from creating the page, but give me a chance to do the SPI case. This is a sort of mess cleaning job. PravinGanechari (talk) 17:26, 16 September 2022 (UTC)Reply[reply]

I have no idea what you're trying to say here, which is part of the problem. Spicy (talk) 17:34, 16 September 2022 (UTC)Reply[reply]

Wait sir PravinGanechari (talk) 17:43, 16 September 2022 (UTC)Reply[reply]

नमस्कार सर , मैंने SPI के बारे में Teahouse में भी सवाल पूछा था। हाँ सर मैंने कुछ गलतियाँ की हैं. यह गलतियां सिर्फ मेरी भाषा की वजह से हुआ है बाकी कुछ नहीं. मेरी भाषा कन्नड़ और हिंदी है। मैं पूरी तरह से अंग्रेजी पढ़ और बोल सकता हूं, मुझे लिखना मुश्किल लगता है। SPI case मामले में सीखने के लिए बहुत कुछ है। ( मैंने खुद Fact Check के लिए हिंदी और कन्नड़ के लिए 2 साल काम किया है ). आप मुझे भाषा के लिए एक उपयोगकर्ता की मदद लेने की अनुमति देते हैं तो मैं वादा करता हूं कि मैं आपको SPI case में बोलने का मौका नहीं दूंगा. मेरे लिए SPI case करना मतलब गंदगी साफ करने जैसा है. आप चाहे तो पेज क्रिएट करने में आप मुझे रोक दीजिए. नहीं तो आप मेरे ऊपर जो चाहे कार्रवाई कर सकते हैं. ( Please Languages Translate Cc: Ravensfire, usernamekiran ) PravinGanechari (talk) 18:05, 16 September 2022 (UTC)Reply[reply]

@Spicy: I agree with your assessment, and have asked PravinGanechari to stop — TheresNoTime (talk • they/them) 18:28, 16 September 2022 (UTC)Reply[reply]
Sorry, I'm unable to translate. PravinGanechari, I share some of the concerns raised above. In the Akshay Singh Rajput Thakur SPI (very overdue for cleark/cu review, hint hint hint), you added Satyamsinghsengar to the list, saying one edit shared a single word, Rajput. That's the sole connection you found. That's way, way too small of a connection to try and tie to a particular sock. It looks like you're just putting every account you find that's edited the same article as a sock. You need SOLID and SIGNIFICANT evidence. Right now, you've shown you don't understand what that means. Taking a break from SPI and just editing would be helpful. Ravensfire (talk) 18:31, 16 September 2022 (UTC)Reply[reply]

Ok, After today I will not do any kind of SPI case PravinGanechari (talk) 19:22, 16 September 2022 (UTC)Reply[reply]

Thank you moving to other activities. Nobody wants to block you, so we are very happy that you have found something else you enjoy doing on Wikipedia. Cheerio and happy editing! 🙂 Mako001 (C)  (T)  🇺🇦 13:57, 18 September 2022 (UTC)Reply[reply]

Jack Frost promoted to full clerk

I am pleased to announce that Jack Frost has completed his training and has been promoted to full SPI clerk. Folks, please commence the ritual hazing at your earliest convenience. -- RoySmith (talk) 17:07, 19 September 2022 (UTC)Reply[reply]

Merge cases?

I'm not sure if this is the right place to request merging of SPI archives, but anyways:

Shouldn't Wikipedia:Sockpuppet investigations/Tscdrwh be merged into Wikipedia:Sockpuppet investigations/Awolf58? Steel1943 (talk) 18:09, 27 September 2022 (UTC)Reply[reply]

Based on what evidence? -- RoySmith (talk) 18:42, 27 September 2022 (UTC)Reply[reply]
Tscdrwh was declared to be a sockpuppet of Awolf58. Steel1943 (talk) 19:26, 27 September 2022 (UTC)Reply[reply]

All Sir / Mam Please Request

Hi Sir / Mam

I have a little language problem. Because of that I have made many mistakes while reporting on SPI. And I admit I have made a mistake but I need a chance to correct my mistake. All of you are requested to give me one more chance to report on SPI case. And I promise I won't waste your time — Preceding unsigned comment added by PravinGanechari (talkcontribs) 15:30, 7 October 2022 (UTC)Reply[reply]

I don't think it would be good for this appeal to be accepted for you and SPI in general. It was often hard for me to understand your reports and evidence at SPI. In the above appeal there are several problems with your English that made it harder to understand.
Furthermore reading the two reports you have left in October at User talk:DatGuy and the one in October at Wikipedia:Administrators' noticeboard these also have the problem. For example at WP:AN you said Doubt that this user also belongs to this group which reads to me that you don't think the account is a sock, which is obviously not the intended meaning.
If you are using Google translate or similar to make these reports, I would suggest that SPI on the English Wikipedia is not the place to be contributing. This is because the complicated nature of SPI including linking diffs and using words which either don't exist, are not in common use, or are rarely used in specific contexts make using general purpose translation machines (such as Google translate) difficult while also having a good report. Dreamy Jazz talk to me | my contributions 17:22, 7 October 2022 (UTC)Reply[reply]

Yes sir I totally agree my original language is Kannada, Tulu and Hindi language. But don't make false allegations, if you want, block me and leave the access to the talk page for me, I will give information there. I will take the help of a user to report if you have permission. PravinGanechari (talk) 18:03, 7 October 2022 (UTC)Reply[reply]

  • This is bananas. I've indeffed the user without TPA.--Bbb23 (talk) 18:12, 7 October 2022 (UTC)Reply[reply]


I saw their hashtag on Special:Diff/1100563501. This appears to be a legitimate collaboration with WMF to write about African topics. See this WMF tweet. I don't know any more about it, but figured it would be of interest to the SPI crowd. -- RoySmith (talk) 13:41, 9 October 2022 (UTC)Reply[reply]

Thanks Roy - added to the contest & editathon tracking filter just in case. firefly ( t · c ) 14:08, 9 October 2022 (UTC)Reply[reply]
Also meta:AfroCreatives WikiProject -- RoySmith (talk) 14:40, 9 October 2022 (UTC)Reply[reply]


Do any of us use the template {{IPsock}}? I sampled a few transclusions, and it seems like this template was used a lot around 2006–2010, but I suspect that nowadays the idea of sock-tagging IP addresses has fallen completely out of favor. Unlike tagging accounts, the usefulness of tagging IP addresses decays with time: it is very unlikely that the person using a particular IP address today is the same person who used the IP in 2006–2010. In the worst case scenario, we are biting newcomers by incorrectly flagging them as potential socks. I was honestly pretty close to starting another TfD for this template, but wanted to ask if anyone thinks differently from me. I can see the template is still mentioned at Wikipedia:Sockpuppet investigations/SPI/Administrators instructions#Sock puppets (IP addresses) with little guidance on when it might be appropriate. Mz7 (talk) 02:48, 23 October 2022 (UTC)Reply[reply]

I agree with TfDing it. The likelihood of an IP being stable enough to warrant a tag (in addition to a block) is quite a low usecase. There's always the option of adding a switch to {{sock}} so that the language and links automatically switch for IPs. — Preceding unsigned comment added by Callanecc (talkcontribs)
As a checkuser I generally consider it generally too risky for me to use it. However I have used it in the past. It is particularly good for identifying ranges and tying them to sockmasters (which can obviously be useful for a couple of reasons). You don't need stable IPs, just stable ranges. -- zzuuzz (talk) 06:53, 23 October 2022 (UTC)Reply[reply]
I have never used it from my memory and also rarely seen it used correctly. Dreamy Jazz talk to me | my contributions 12:47, 23 October 2022 (UTC)Reply[reply]
Agree that it needs to go. TfD sounds like a good plan. firefly ( t · c ) 13:17, 23 October 2022 (UTC)Reply[reply]
Yeah, nuke it. This is what cuwiki is for. -- RoySmith (talk) 13:41, 23 October 2022 (UTC)Reply[reply]
Only if you're a checkuser. This is the non-CU alternative. -- zzuuzz (talk) 13:48, 23 October 2022 (UTC)Reply[reply]
The non-CU alternative are SPI pages. In my experience, sock-tagged IPs (more than 5,000 pages only counting Sockpuppet tags on IPs) are generally noise. Including random users who have dedicated time to tag most IPs in the same /24 ranges as socks of one of the IPs in that range. MarioGom (talk) 17:33, 27 October 2022 (UTC)Reply[reply]
Personally, I've never used it and don't plan to. Most recent transclusions seem to be by users who are not admins or clerks, and most are unnecessary or unhelpful (e.g. tagging individual IPs in a v6 range [4][5][6], tagging a single IPV6 that made 4 edits months ago [7], tagging an IP that wasn't blocked at the time in violation of WP:HSOCK [8])...) For the purpose of figuring out who's using a range, I've found rangeblockfinder to be much more helpful. Spicy (talk) 14:40, 23 October 2022 (UTC)Reply[reply]
Will this template be useful with the new IP Privacy initiative that's coming? ~ Matthewrb Talk to me · Changes I've made 16:26, 23 October 2022 (UTC)Reply[reply]
As I understand it, historical data will not be affected, and current IP addresses will still be available (for 'IP-lookers' and even more so for checkusers). You will still be able to look at an unregistered vandal and a category of IP addresses and ask, "does this one match?". There's probably going to be a limitation on adding it to any recent masked 'edits' by the IP, and any IP edits won't be visible, but I'm not entirely sure about that. In other words, IP masking will probably gradually reduce the template's usefulness going forward. -- zzuuzz (talk) 19:36, 23 October 2022 (UTC)Reply[reply]
To add to this, I don't know if anyone's thought about how we are going to tag these new masked pseudonyms, if at all, but the idea of using a (probably new or adapted) template to tag them while deleting this template would seem to have an air of inconsistency about it. -- zzuuzz (talk) 20:13, 23 October 2022 (UTC)Reply[reply]
I would be more content to see pages containing this template blanked, after say 10 years, which seems to be a commonly agreed time for IP staleness. -- zzuuzz (talk) 19:36, 23 October 2022 (UTC)Reply[reply]
I would delete it, or at the very least, remove it from all User:<IP> pages. There's little use for this template, and the overwhelming majority sock-tagged IP pages are tagged by random users acting out of policy. And there are some practical caveats, such as the pollution of sock categories, which get a performance impact from this without any benefit justifying it. MarioGom (talk) 17:08, 27 October 2022 (UTC)Reply[reply]

Template:Sockpuppet usage: Unblocked IPs without second parameter

Since Template talk:Sockpuppet § Deprecation of "an editor has expressed a concern" version was started, I have been tracking various types of inaccurate use of {{Sockpuppet}} tags. One of these types are tags without second parameter (deprecated) on unblocked IPs. I think it is uncontroversial that this kind of tag is not useful and even harmful to later users of these IPs, who may be unrelated to whoever was using these IPs years ago. Most of them are also out of policy, since they were placed on user pages by non-admin/CU/clerk users.

There are more than 5,000 user pages in this category (not counting tags in User talk). These sometimes point to other IPs as master (example) and others point to named accounts ([9]).

Would you agree to remove these tags? I can prepare a detailed log of each candidate removal upfront with its rationale if you want to review it. I can also break them into some subcategories (master IP, master account, no master). MarioGom (talk) 11:18, 29 October 2022 (UTC)Reply[reply]

I drafted an initial backlog and did some trial removals of ancient examples logged at User:MarioGom/Sockpuppet tags cleanup. I have not included {{IPsock}} at the moment, just {{Sockpuppet}}. Since zzuuzz mentioned some use case for ranges, we could maintain tags that are less than 10 years old, and added by CheckUsers/admins/clerks? MarioGom (talk) 08:48, 30 October 2022 (UTC)Reply[reply]
In general I don't disagree with this specific proposition/list. We know some LTAs last longer than 10 years and their IPs remain relevant. I see Jason Gastrich is in this list, and they were mentioned at WP:AN only the other day. But anyway, I do think the HarveyCarter (ipsock) tags are an excellent example of usefulness. Some of those ranges tagged in the last 6 years (and longer) are still relevant today, and even ones tagged in 2007 still show usefulness. This type of information is not always captured at SPI for several reasons, and it's rarely displayed as efficiently. But if someone wants to draw a line at 10 years I probably won't object. -- zzuuzz (talk) 09:23, 30 October 2022 (UTC)Reply[reply]
Since I'd like to focus on uncontroversial cases first (and have a more substantive review of other cases later), I could focus initially on unblocked IPs without second parameter and not tagged by CheckUsers/admins/clerks (dropping the 10 years line). After all, my point is focusing on removing noise, not pushing to remove tags that CUs do find useful. MarioGom (talk) 10:11, 30 October 2022 (UTC)Reply[reply]
Fully support this - I’ve long thought that these sort of tags (of long stale, unblocked IPs) are worse than useless. firefly ( t · c ) 10:27, 30 October 2022 (UTC)Reply[reply]

Malformed SPI

Could someone look at fixing: Wikipedia:Sockpuppet investigations/Farzanfa007; also feel free to block anyone bickering for disruption if needed regardless of socking. Tried making sense of the back and forth, but in my sleep-deprived state decided to punt :) TonyBallioni (talk) 23:43, 14 March 2023 (UTC)Reply[reply]