Wikidata

Wikidata:Requests for permissions/CheckUser/Jasper Deng

< Wikidata:Requests for permissions‎ | CheckUser

The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section. A summary of the conclusions reached follows.

Closed as successful --Lymantria (talk) 08:23, 26 April 2020 (UTC)[reply]

Jasper Deng edit

Vote
RfP scheduled to end at 26 April 2020 01:02 (UTC)

Hello everyone. With the amount of sockpuppetry cases we have been having lately, I think we could benefit from having some local CheckUsers. I myself have responded to many of these cases, and most of my work on the project is now dedicated to anti-abuse activities. This includes a large proportion of recent blocks for sockpuppetry. Still, I always ensure that I have sufficient evidence whenever I block sockpuppets.

But this only allows me to block accounts or IP addresses publicly reported or which the user admits to. It does not allow me to discover "sleepers" (sockpuppet accounts created in advance to be used later, and which thus have no edits yet; this is especially common for spambots), block underlying IP addresses of repeat offenders, or investigate borderline cases. Also, when making rangeblocks, I currently can only rely on anonymous edits from the range to determine possible collateral damage, which is an incomplete picture since collateral damage also includes logged-in users, especially in the case of hard blocks. Having access to CheckUser would enable me to do all of these. To be clear, any use of CheckUser would be in addition to, and not in place of, what I already do: if I can solve a socking case without CheckUser, I would avoid checking them to respect their privacy

In terms of background, I have an extensive knowledge of computer networking, particularly IP addresses (I spearheaded the effort to prepare the global community for IPv6, for example), and I am very comfortable doing rangeblocks (I wrote the MediaWiki.org documentation for IPv6 rangeblocks). I have many years of anti-vandalism experience and have made requests to the stewards for CheckUser investigations of Wikidata users. I am identified to the Foundation and have been so since 2014. While my edit count is admittedly low for an administrator, I have become much more active in responding to block and protection requests on the administrators' noticeboard (in part because the coronavirus pandemic has resulted in my having to shelter in place) and intend to maintain this level of activity for the foreseeable future. I am in the Pacific Time Zone (UTC-8 or UTC-7), which means I can cover the European night, during which most stewards are unavailable. Along with Sotiale, who is in UTC+9, I believe we can provide coverage at least equal to if not better than currently provided by the stewards.

I would be happy to answer any questions you might have. --Jasper Deng (talk) 01:02, 12 April 2020 (UTC)[reply]

Votes edit

  1.  Strong support as someone who has spoken with you directly about your possibly requesting the right, I believe you would be perhaps the best qualified to use it (certainly much more so than I, who had at one point expressed interest in requesting the right). Mahir256 (talk) 01:09, 12 April 2020 (UTC)[reply]
  2.  Strong support per Mahir256 --Kostas20142 (talk) 01:14, 12 April 2020 (UTC)[reply]
  3.  Support There's a need and on trust side Jasper is serving this project as admin from a long time and I think they can be trusted with this perm too. Additionally I see them active in that area we call finding "Sockpuppets" and for me that is also a plus point. ‐‐1997kB (talk) 01:44, 12 April 2020 (UTC)[reply]
  4.  Support Jasper has the technical knowledge as well as the policy knowledge of when it is appropriate to request CU. I support this request. --Rschen7754 04:34, 12 April 2020 (UTC)[reply]
  5.  Oppose untrusted user. Hardly active on Wikidata. Looks like hat collecting. --- Jura 05:46, 12 April 2020 (UTC)[reply]
  6.  Support - Nikki (talk) 05:48, 12 April 2020 (UTC)[reply]
  7.  Oppose The contributor only totals a little more than 5000 modifications, the majority of his modifications were made more than 6 years ago. Today, his presence is only a few dozen changes per month, or even less than ten. He does not have the necessary experience and presence at all for the project to seek this access. In addition, the contributor is only interested in a tiny part of the project. —Eihel (talk) 06:07, 12 April 2020 (UTC)[reply]
  8. Support, trusted. Stryn (talk) 06:31, 12 April 2020 (UTC)[reply]
  9.  Support - trusted user who is very active in the areas where the tools are needed. Has all the necessary experience, no reason to oppose.-BRP ever 06:54, 12 April 2020 (UTC)[reply]
  10.  Strong support Jasper has demonstrated a long term commitment to this project as an admin since month 2 of Wikidata's existence. His technical knowledge, and familiarity with Mediawiki, make him an optimal candidate when paired with an experienced and mentoring CU like Sotiale. As someone already active in the area of fighting LTAs and sockpuppets, Jasper will have no issues transitioning into the CU position. ~riley (talk) 07:18, 12 April 2020 (UTC)[reply]
  11.  Support Should be fine per endorsements of ~riley, Stryn, BRPever, 1997kB, Kostas20142 and Rschen7754, all wikimedians that I respected a lot. Camouflaged Mirage (talk) 08:51, 12 April 2020 (UTC)[reply]
  12.  Support Quoting word for word Rschen7754. --Sannita - not just another it.wiki sysop 11:38, 12 April 2020 (UTC)[reply]
  13.  Support --Epìdosis 12:37, 12 April 2020 (UTC)[reply]
  14. Why not. — regards, Revi 12:39, 12 April 2020 (UTC)[reply]
  15.  Support Must trusted. Thank you for your work and best wish for CheckUser holding on Wikidata. Regards, ZI Jony (Talk) 15:02, 12 April 2020 (UTC)[reply]
  16.  Oppose; apart from not seeing need for local checkusers (neutral), I am not fully convinced by the candidate's admin performance (oppose), and I do not like the fact that he is admin-only here, with apparently no connection to the content of this project (oppose). —MisterSynergy (talk) 16:39, 12 April 2020 (UTC)[reply]
  17.  Oppose Not really active on Wikidata for years (see Wikiscan). — Envlh (talk) 08:08, 13 April 2020 (UTC)[reply]
  18.  Support I trust all three of you. --Wolverène (talk) 10:07, 13 April 2020 (UTC)[reply]
  19.  Support Jianhui67 talk★contribs 16:17, 13 April 2020 (UTC)[reply]
  20.  Support I agree that this is needed. ArthurPSmith (talk) 17:45, 13 April 2020 (UTC)[reply]
  21.  Oppose Definitely possesses experience, but he is barely active on Wikidata. I prefer users committed to the project to hold advanced flags. --Esteban16 (talk) 19:51, 13 April 2020 (UTC)[reply]
  22.  Support Jasper Deng is an experienced and trusted user. I understand the inactivity argument, but there is very little chance of abuse and definitely the possibility for Jasper Deng to do a lot of good work with the tool. Vermont (talk) 02:42, 14 April 2020 (UTC)[reply]
  23. Support, trustworthy user. Kind regards, — Tulsi Bhagat [ contribs | talk ] 13:37, 14 April 2020 (UTC)[reply]
  24.  Strong support User can be trusted with such a great tool. ミラP 14:51, 14 April 2020 (UTC)[reply]
  25.  Support Trusted enough with 9 years of editing experience under his belt. As per Vermont, inactivity should not be the reason to oppose someone for requesting more rights to aid the community. Minorax (talk) 08:04, 17 April 2020 (UTC)[reply]
  26.  Support, per ~riley, Vermont and Rschen7754. —Hasley 22:39, 20 April 2020 (UTC)[reply]
  27.  Weak oppose per Esteban16. -- CptViraj (📧) 09:41, 21 April 2020 (UTC)[reply]
  28.  Support Good experience, and trusted --Alaa :)..! 07:33, 23 April 2020 (UTC)[reply]
  29.  Support per above --DannyS712 (talk) 07:58, 23 April 2020 (UTC)[reply]
  30.  Support --MichelBakni (talk) 08:05, 23 April 2020 (UTC)[reply]
  31.  Support --Pasleim (talk) 08:27, 23 April 2020 (UTC)[reply]
  32.  Support --Sotiale (talk) 09:21, 23 April 2020 (UTC)[reply]
  33.  Support--باسم (talk) 11:37, 23 April 2020 (UTC)[reply]
  34.  Support. To stop (for example) spammers, local CUs will be a huge help in my opinion. I understand the activity concern, but I appreciate the user's experience. Thanks for volunteering! Bencemac (talk) 17:42, 23 April 2020 (UTC)[reply]
  35.  Support -- Ajraddatz (talk) 03:33, 24 April 2020 (UTC)[reply]
  36.  Support comfortable that the user has knowledge of wikis and earnt trust of community, and the community does need checkusers  — billinghurst sDrewth 11:59, 24 April 2020 (UTC)[reply]
  37.  Support Mz7 (talk) 06:16, 25 April 2020 (UTC)[reply]
  38. Support--*Youngjin (talk) 07:37, 25 April 2020 (UTC)[reply]
  39.  Strong oppose. Not well grounded in this international community. In my opinion tries to enforce enWp rules to WD as my own case shows. --Succu (talk) 21:34, 25 April 2020 (UTC)[reply]

Comments edit

  •  Question: Looking at your blocking log, I see several instances where you cite shared use as a reason for blocking, saying things like "accounts may not be shared by multiple people". If you are granted CheckUser here, how can we be confident that you will block only in accordance with local policy? Bovlb (talk) 17:15, 12 April 2020 (UTC)[reply]
    • @Bovlb: 100%. The shared use concern was an oversight on my part (in most cases, it was not the sole reason for the block, but still, I should not have included it as a block reason at all). My sockpuppetry blocks have always been in line with policy. But also, I'm always willing to admit mistakes and I think that is important. While I don't expect to make many mistakes, being able to admit to having done so is important to me.--17:20, 12 April 2020 (UTC)[reply]
      • Thanks. I appreciate the response. As an admin on multiple projects, I also sometimes find it hard to keep track of which policies apply where. When dealing with issues of acceptable account naming and use (as opposed to local behaviour), I find it useful to remember that accounts are global, and Wikidata is in a special central position, so a good starting point is to permit anything that any client project would permit. Bovlb (talk) 17:44, 12 April 2020 (UTC)[reply]
        • @Bovlb: In regards to CheckUser in particular, I actually don't intend to do many "CheckUser blocks", which I intend to reserve as a last resort, since their nature is such that it's hard for others to review those blocks. Thus whenever I do so, it would be with absolute certainty that it is the best way to protect the project, and if I have any doubts I would not hesitate to consult with another CheckUser. My standard for issuing a block has always been: if it were to be disputed, would it have consensus that it was the correct course of action and in line with policy?--Jasper Deng (talk) 17:50, 12 April 2020 (UTC)[reply]
  •  Question while I came to the conclusion that you are not very active on this website, it seems to some that you have no connection at all to the content of this project. This may be possible as you seem to have been granted adminship before the website actually had any content and you didn't really start contributing actively afterwards. While I haven't looked into this in detail, maybe you could outline how you contributed content of this project, i.e. what do you consider your biggest content contributions. --- Jura 00:04, 13 April 2020 (UTC)[reply]
    • CheckUser is mostly a management position, and does not have many to do with the contents.--GZWDer (talk) 01:15, 13 April 2020 (UTC)[reply]
    • The question is about the user's involvement in the project as such. Adminship is closely linked to content creation. --- Jura 11:06, 13 April 2020 (UTC)[reply]
    • @Jura1: While my present involvement is primarily administrative in nature, I began by creating items. I am a very technically-minded person so my work began by making items on things like Remote Desktop Protocol (Q49150). Afterwards, however, most things I wanted to create items on were already imported. I personally do not believe in automatic editing en masse; I don't oppose its use by others, but think that it is important to be transparent, hence why I do not seek a high edit count. I've created two properties here. My biggest contribution is probably one that I have been wanting to do, but which I have not gotten around to yet: overhauling the ontology for our current coverage of tropical cyclone (Q8092); however, a stumbling block is that we will need an efficient way to have not just annotated key-value pairs, but to also store lists of tuples of key-value pairs in order to import a best track of a tropical cyclone. The data would be imported from International Best Track Archive for Climate Stewardship (Q43269466). It might be possible to put it in a tabular file on Commons, but I also am not sure about having the data decentralized, especially as SPARQL queries would need to be significantly more complicated (I am strongly familiar with SQL, which is very similar to SPARQL, and also studied database structures over the course of my academic career). In addition, there are numerous derived quantities in this field, such as maximum sustained wind, Accumulated cyclone energy (Q339687), fastest intensification, etc. that would be tied to the contents of the best tracks; thus I would like to see calculated properties be implemented.
    • Overall, though, I believe my strength is more on the anti-abuse side. I was one of the first to express concerns about Wikidata:Living people and created the first RfC on it, though the community rejected my ideas at the time, and I respect that: I'm not always going to agree with the community on a personal level, but I always will agree to be bound by its consensus. My philosophy is that the less sockpuppet and spambot abuse we see, and the quicker it is dealt with, the more time others will have for useful content contributions, and the less need for maintenance overall. I am also more active than my edit count would suggest at first glance; I am very responsive on email and IRC, and also quickly responsive to vandalism reports. Since I have done so from day one of the project, I am very proficient with identifying vandalism on this project, which has a unique look compared to other projects' vandalism; in particular, mere removal of claims is not vandalism as often as blanking pages on other projects is (for example, if that claim happened to violate the living people policy, and I cannot quickly verify whether it is true, the removal of the claim likely wouldn't be considered vandalism). Although French is the only other language I can read proficiently, I know enough to fairly easily identify vandalism in other Romance languages, and if there is a doubt, I always ask someone who is more experienced in the language first. When it comes to handling requests for CheckUser, in the worst case I can refer requests in a language I cannot understand to stewards, who are also trusted with similar access, and who can then translate the request for me, allowing me to then respond as usual.--Jasper Deng (talk) 05:16, 14 April 2020 (UTC)[reply]
  •  Question You mention hard range-blocks - under what conditions do you think that would be necessary? And it's not entirely clear from your statement, would you promise to otherwise avoid looking at the IP addresses of regular user accounts here, except in the case of suspected sock puppetry or where a hard range-block was necessary? ArthurPSmith (talk) 17:50, 13 April 2020 (UTC)[reply]
    • @ArthurPSmith: Hard rangeblocks are a last resort. In the following, "soft" means anonymous users only, and "hard" means blocking of registered users as well. The following conditions would have to be met:
      • The collateral damage as found by checking is minimal. That is, there is very low probability of good-faith editors being caught in the rangeblock. In particular, for a hard rangeblock, there must be very few (ideally no) good-faith logged-in users. If there are, then it must be considered whether it is practical to assign them IP block exemption; in most cases, though, having good-faith logged-in users use the range would rule out a hard rangeblock.
      • Other measures, like single IP blocks, soft rangeblocks, abuse filters, or even the spam blacklist (in the case of spambots) have repeatedly failed to stem the sockpuppet abuse.
      • In particular for hard rangeblocks, the sockpuppet abuse cannot be mere block evasion using (logged-out) IP's in the range. The sockpuppetry must entail use of accounts. The strongest use case is when the sockpuppeteer makes their accounts in one place, but carries out almost all of their actual abuse from the range in question (the place where the accounts were made would be blocked before I would consider any hard rangeblocks).
      • These conditions are necessary, but not sufficient (necessity and sufficiency (Q875267)). Even in such a situation, it may still be better to just expand previous measures, such as a larger soft rangeblock, a tweaked abuse filter, etc. On a similar note, I cannot know the first condition without checking the range, and the third condition would have been determined by which IP's show up when checking sock accounts. Such rangeblocks also will be made as small and as short as possible; if this means making several small rangeblocks rather than one big one, the extra work is worth it to minimize collateral damage.
    • The policy also permits the use of CheckUser to check for collateral damage, and I expect to be asked "would I cause too much collateral damage by blocking this range?" by others. In such cases, I would check the range and report as vaguely as I can while still answering that question and respecting privacy. Further checks would be warranted only if I also happen to discover actual uninvestigated sockpuppetry in the process (such as an active account whose behavior (edits) is strongly reminiscent of a known sockpuppeteer).
    • Along with verifying a need for someone to have IP block exemption, these exhaust the permitted uses under our policy. Accordingly, I would choose not to conduct checks besides what you mention, and refuse any requests to do otherwise. I have a firm grasp of what is permitted and not (I have not had a request for checking a Wikidata sockpuppetry case refused, at least not for a very long time), but if I ever have doubts, I would refer the case to a colleague. Under no circumstances would I perform a check that I am not absolutely certain is within policy and necessary (in the sense that other measures have proven inadequate) to protect the project.--Jasper Deng (talk) 04:52, 14 April 2020 (UTC)[reply]
  • @Jasper Deng: If I find that odd that in (almost) ten years, you find 1 item to mention as content contribution and fill half a talk page commenting about it and the inadequacy of Wikidata for your content. Maybe this explains why your admin action somehow illustrate a lack of understanding editing related MediaWiki functions. --- Jura 16:55, 20 April 2020 (UTC)[reply]

Conclusion edit

Votes support-oppose are found as 32(82%)-7(18%) hence closed as successful. Adding flag has been requested at meta (by User:Ymblanter). Lymantria (talk) 08:33, 26 April 2020 (UTC)[reply]

The above discussion is preserved as an archive. Please do not modify it. Subsequent comments should be made in a new section.
Retrieved from "https://www.wikidata.org/w/index.php?title=Wikidata:Requests_for_permissions/CheckUser/Jasper_Deng&oldid=1175810223"