Wikidata:Requests for permissions/RfRemoval/2014

From Wikidata
Jump to navigation Jump to search

February 2014[edit]

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.

KrattBot[edit]

KrattBot (talk • contribs • new items • new lexemes • SUL • Block log • User rights log • User rights • xtools)

The bot's operator doesn't want to fix the bot, and create another bot account KrattBot1. It's an evasion of block. I suggest to remove KrattBot's flag. @Izno: GZWDer (talk) 12:47, 9 February 2014 (UTC)[reply]

✓ Done This is a really obvious misuse of the right. Also the account is blocked indefinitely. If the operator wants to get the flag, a new RFP is necessary. There he must show that he fixed the bot and understands our guidelines. Additionally he must ensure that evasions of blocks cannot be tolerated under no circumstances. -- Bene* talk 13:26, 9 February 2014 (UTC)[reply]
The above discussion is preserved as an archive. Please do not modify it. Subsequent comments should be made in a new section.

May 2014[edit]

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.

RobotGMwikt[edit]

RobotGMwikt (talk • contribs • new items • new lexemes • SUL • Block log • User rights log • User rights • xtools)

Many mistakes, repeating the same mistake many times, and running tasks which are not per Wikidata:Requests for permissions/Bot/RobotGMwikt. The bot is using categories[1], whereas the bot request was to use infobox parameters. See User_talk:GerardM#Human_groups for the latest repeated mess; example here. This is exactly the same type of problem reported at WD:BN last month; see example John Vandenberg (talk) 14:45, 12 May 2014 (UTC)[reply]

Many mistakes? sure.. in numbers. Is that all there is here ?
At the Hackathon in Zurich it was mentioned that 2 million edits have been done with oAuth authentication in Wikidata. Thanks to the experience with the work done we have learned a lot. This resulted in an RFC on quality. We have learned that many of the infoboxes can not be safely imported and the results will not be enough to create a situation where Wikidata can start hosting that data for a template. After writing the RFC and not getting negative responses, we have aggressively added new items and started to add humans to Wikidata items. There are now 27% more humans known. We have added human for people who died in specific years. This has been an iterative process. We are now at a point where we can manage new humans on a daily basis.
I paused using my bot because there were some improvements in the code being developed. Now we have a revamped AutoList that allows edits to be performed as a bot. The new AutoList prevents disruption of the recent changes by me and, it provides a batch mode that is really welcome. We are experimenting with including AutoList as a generator in pywikipedia. The objective is to create a process that will automate the harvesting of selected templates under pre-defined conditions.. The current notion is that we can safely import date of birth/death and gender for "instance of" "human".
We are looking in how to better support for elections, creating links that can be copied based on Wikidata query results. Subjects like locations and "administrative districts", maps and query results. At the Zurich hackathon work was done on a maps name space included are ideas on Wikidata providing information on several layers of a map and for the selection of maps.
When this proposal results in the loss of a bot flag, the research will suffer and I will not be able to experiment with ways of automating the addition of data and understand the effect of it. As you know I blog early and often about the work that I do.
Thanks, RobotGMwikt (talk) 06:56, 13 May 2014 (UTC)[reply]
You're supposed to test using the test wiki, right?--Jasper Deng (talk) 07:04, 13 May 2014 (UTC)[reply]
What I am doing is add "instance of" "human" based on the existence in a category that indicates death of birth.. There is not much to test realistically. GerardM (talk) 10:40, 13 May 2014 (UTC)[reply]
You have not addressed why you are using categories (with incorrect assumptions about their membershp) instead of infoboxes (which are already structured data).
Even here, you are breaking your bot permission, which does not including writing as a person. Please use your own account to comment in community forum. John Vandenberg (talk) 07:17, 13 May 2014 (UTC)[reply]
Categories indicate for instance that someone died. I can query for the existence in a category and thereby find new articles on humans. This enables to daily run a bot that will find the new articles so that we can harvest the templates. It is the only way in which we get to the point where we have always are near up to date with a Wikipedia.
The argument that I "break my bot permission" by writing as a person is overly legalistic. I use the bot flag not to overflow the recent changes. Also the bot policies do not consider the tools like AutoList of ToolScript. Most of the edits I do are not with a bot script. RobotGMwikt (talk) 10:34, 13 May 2014 (UTC)[reply]
As a side note, please do not use your bot account in order to participate in discussions in the future. As this hides your edits to this page from RC as well, this isn't appropriate usage of your bot flag. Vogone (talk) 15:53, 13 May 2014 (UTC)[reply]

No change in behaviour. Just now Gerard has restored Q31->Q5 (human) on NKVD prisoner massacres (Q2655890) for the third time [2] after reverts by user:YMS and I. John Vandenberg (talk) 16:29, 13 May 2014 (UTC)[reply]

The bot is also creating items for brand new Wikipedia pages, in the case of Rashid Rehman (Q16799419) and Rashid Rehman (Q16799424) these are both created by the same bot, about a death that happened on the 9th of this month (May) and is international news. The bot immediately created two items when the Wikipedia pages were created (one for Urdu and one for English). The pywikibot script for creating items does not allow item creation immediately by default - the bot operator must override the defaults. This is a classic example of why the rush to create items is not a good idea. As I keep saying, Gerard did not ask for permission to create new items, so these edits are also contrary to bot approval, in addition to being bad edits. John Vandenberg (talk) 16:45, 13 May 2014 (UTC)[reply]

In the last 50 edits right now, are three more examples of wrong Q31->Q5 to Rasyphus and Ravennus (Q7295348), Victorian, Frumentius and Companions (Q4111032) and Quintian, Lucius and Julian (Q7272430). All 50 edits happened in the same minute, and the bot has been operating at this speed for days now. John Vandenberg (talk) 16:54, 13 May 2014 (UTC)[reply]

John, you do not consider the numbers involved and, you expect the work that I do as being developed as pywikipedia code. This is not the case. I can also testify that you are only howling for blood because you are not talking with me. So you do not know what the F is going on. Also the creation of new items for people who died has been going on for the best part of three months now.
I put arguments up and what I see is that you miss them completely. An example; in a previous exchange it was suggested to me that I exclude every "instance of". This is what I now do. It follows that "NKVD prisoner massacres" need to an instance of something. As it is they should NOT be in a category that is about deaths in a specific year in the first place. Given that you only blame me and do not communicate with me, it is sad to observe that your revert does not have the effect that you want. It is because you do not follow it up with the addition of an appropriate "instance of".
I blogged about another categories of fails and you can check it, most of the issues indicated were wrong in Wikidata.
John, please argue the points and communicate. Our aim is the same higher quality for Wikidata. Thanks, GerardM (talk) 17:45, 13 May 2014 (UTC)[reply]
IMO the point is that you are using categories instead of infoboxes, which is contrary to your bot flag request & approval, and you are very foolishly blaming Wikipedia for having pages in the wrong categories, whereas each Wikipedia has conventions of what is acceptable members of a category, and you must accept and adapt to them. Often that means manually reviewing the edits either before or after a bot run, and fixing any errors, *and* ensuring that your procedures dont repeat the same errors, again and again. You are not doing that. ergo, you should not have a bot flag. Your work is not more important than community policy and conventions. You need to plan your intended work, ask for bot approval for the tasks involved, and execute them with professionalism. John Vandenberg (talk) 18:27, 13 May 2014 (UTC)[reply]
+1. I hope Ivan A. Krestinin has no objection citing him: „We have too many invalid claims in Wikidata already”. --Succu (talk) 18:40, 13 May 2014 (UTC)[reply]

I've checked the last ~270 items at Special:NewPages/RobotGMwikt, and these are the duplicates that I can easily spot:

  1. Q16867371 dup of Anwar Ahmed Khan (Q4778045)
  2. Q16831410 also dup of Anwar Ahmed Khan (Q4778045)
  3. Q16862769 dup of Tokuo Yamashita (Q11465958)
  4. Q16860771 dup of Donald Byrd (Q362764)
  5. Q16860760 dup of Jim Kelly (Q714870)
  6. Q16859122 dup of Guillo Pérez (Q2886163)
  7. Q16830082 dup of Marcello Giannini (Q3845596)
  8. Q16799424 dup of Rashid Rehman (Q16799419)
  9. Q16777437 dup of Francesco De Nittis (Q1441009)
  10. Q16773069 dup of Shirish Chandra Dixit (Q15527065)
  11. Q16762671 dup of Ulrik Berglann (Q13579361)
  12. Q16762389 dup of Hiroki Tanaka (Q15674402)
  13. Q16762241 dup of Darlan Romani (Q9202341)
  14. Q16762179 dup of Franziska Fritz (Q16596782)
  15. Q16762173 dup of Maria Adela Constantin (Q16187723)
  16. Q16761695 dup of Nurbergen Zhumagaziyev (Q15830398)
  17. Q16761681 dup of Jithin Thomas (Q16645910)
  18. Q16761680 dup of Takanoiwa Yoshimori (Q11634736)
  19. Q16761664 dup of Viola Kleiser (Q16596824)
  20. Q16755989 dup of Jacky Vincent (Q10304768)
  21. Q16751968 dup of Elena Baltacha (Q232680)
  22. Q16755914 dup of Thorsten Margis (Q7000359)
  23. Q16755887 dup of Tamaryn Payne (Q6138441)
  24. Q16755864 dup of Zhao Jun (Q9348106)
  25. Q16753277 dup of Nadezhda Sergeeva (Q16596911)
  26. Q16753251 dup of Sally Mayara da Silva (Q16187938)
  27. Q16752968 dup of ST1M (Q3920015)

i.e. roughly 10% are duplicates. There are probably more that have been deleted before I could see them.

In the last few days, here are the majority of the items which Gerard has used this bot to mis-apply P31->Q5 to

  1. St. Gerard Majella Annual Novena (Q7588241) [3]
  2. Robert Burns' diamond point engravings (Q7342515) [4]
  3. Draper & Folsom (Q5305847) [5]
  4. Milsom and Fowler (Q6860749) [6]
  5. assassination of Thomas Ashton (Q16199212) [7]
  6. Women and Birds (Q16856984) [8]
  7. Walls of the Sun and the Moon (Q16782394) [9]
  8. The World Trade Center Tapestry (Q16726206) [10]
  9. Theatrical Productions of André Antoine (Q16066541) [11]
  10. United Nations Memorial Cemetery (Q15465499) [12]
  11. Stratton Brothers case (Q7622258) [13]
  12. Seidenbeutel brothers (Q7446632) [14]
  13. Painting (Blue Star) (Q16856712) [15]
  14. list of works by Eugène Guillaume (Q16840477) [16]
  15. Q16155148 [17]
  16. Q16155154 [18]
  17. Q16155167 [19]
  18. Q16155162 [20]
  19. Q16155182 [21]
  20. Q16155175 [22]
  21. Locations of executions conducted by Albert Pierrepoint (Q16238492) [23]
  22. murder of John and Betty Stam (Q6937921) [24]
  23. Spanish martyrs of the 20th century (Q2409388) [25]
  24. Martyrs of Turon (Q2735762) [26]
  25. murder of Anita Cobby (Q6937792) [27]
  26. Jesse Dirkhising (Q4162115) [28]
  27. martyrs of the UCA (Q6938093) [29]
  28. Purge of the Red Army in 1941 (Q4120996) [30]
  29. Peruvian prison riots neutralization (Q580861) [31]
  30. East Tyrone (Q5329578) [32]
  31. J.B. Van Sciver Co. (Q15996105) [33]
  32. Executions during the Irish Civil War (Q5419787) [34]
  33. Armenian victims of the Great Purge (Q16153040) [35]
  34. David O. Selznick filmography (Q16003910) [36]
  35. Deportation of Armenian notables in 1915 (Q5260701) [37]
  36. 498 Spanish Martyrs (Q752300) [38]
  37. 522 Spanish Martyrs (Q15066740) [39]
  38. Catholic resistance to Nazi Germany (Q15041049) [40]
  39. Alfreda & Bolesław Pietraszek (Q4723688) [41]
  40. 233 Spanish Martyrs (Q4631492) [42]
  41. 1993 Ramada Hotel drownings (Q4588746) [43]
  42. Death rates in the 20th century (Q5247627) [44]
  43. list of works by François Rude (Q16066270) [45]
  44. Clarke brothers (Q5127513) [46]
  45. Allan Cameron and William Withers (Q4730511) [47]
  46. W & T Fordyce (Q7958457) [48]
  47. list of works by Henri Chapu (Q16062445) [49]
  48. early life of Pedro II of Brazil (Q5326865) [50]
  49. Consolidation of Pedro II of Brazil (Q5163320) [51]
  50. list of works by Edmund Kirby (Q6645617) [52]
  51. list of victories of Rudolf Berthold (Q16043574) [53]
  52. House of David (Q16030612) [54]
  53. Listing of the work of Jean Antoine Injalbert-French sculptor (Q16029201) [55]
  54. Victims of the Night of the Long Knives (Q386986) [56]
  55. Tutt Brothers (Q15998170) [57]
  56. Babes in the Wood murders (Q4837796) [58]
  57. wedding of Nora Robinson and Alexander Kirkman Finlay (Q15634446) [59]

(Note that I started fixing them as I reviewed the edits, but gave up when the bot started reverting me. Also note that in none of these case was the error spotted and fixed by Gerard after the bot edit.)

I havent looked at the contribs for this problem back further than that, however I did run into a P31->Q5 added by the bot at week earlier at lynching of Ed Johnson (Q6708492) [60].

As has been covered already, often his bot essentially reverts humans. e.g. [61] after user:Haplology had removed the same claim. See new user user:HHill's report of a similar issue, after they having fixed and reported the same problem by a different bot[62]. In addition to NKVD prisoner massacres (Q2655890), Night of the Murdered Poets (Q469733) & New Martyr (Q2389950) are other cases where he has not checked his watchlist and avoided repeating the same problematic edits. His bot also adds P31->Q5 when the appropriate instance of has already been added by another bot.[63]

More evidence he is not using the bot for the task approved, on Q10049364 the bot added instance of (P31)->Wikimedia category (Q4167836) [64]. Note that this is an empty category: new:पुचः:बु॰सं॰ १२३५य् पलिस्था जुगु बसोबास. John Vandenberg (talk) 02:28, 14 May 2014 (UTC)[reply]

We have added something like 497,672 humans since April 20. This has raised the total number of humans in Wikidata by 37%. This is based on their existence in categories like "death in 2013". I have written an RFC that is about quality and I have identified that the lack of data on most of the items in Wikidata is what hampers quality. A paper on duplicate items between Portuguese and English Wikipedia suggests that there are massive numbers of duplicates that have not been identified in interlanguage links. With Wikidata some of that problem has been improved thanks to Wikidata. Not all of it. The scale of this issue is compounded by the large number of projects.
John compounds two practices. The second has nothing to do with bot activities. Thanks to EMJR I changed the practice of identifying likely false positives by looking for claim[31] ie the existence of "instance of". As I showed in a blog post, many false positives indicate wrong information in Wikidata. Given the large numbers involved, there are errors. You can imagine that some runs were in the tenths of thousands. They are however within the expected error rate of not more than 8%. One redeeming quality is that exactly because of there now being statements, it is possible to query for them. This is also indicated in the blog post.
John does refer to the work I do on identifying people who died in 2014. Thanks to this work we know about 4148 humans who died from many Wikipedias. In order to register them, items are created for them. As articles are written in many languages, we do find that many articles are linked by the authors in Wikidata. This is not true for many of them as John has found out. In this work the deaths are registered one at a time and manually. This makes that it is not part of the use of the bot flag.
You cannot make an omelet without breaking eggs. You need items in order to make statements about the subjects involved. There is no helping that. :In another example John made it has it that I identify a category that is empty ... ??? Is that an issue ??? This was authenticated by Widar it was not a bot job. The issue is very much that Widar enabled tools can flood the recent changes however the tools are executed interactively. They are not scripts run in a batch mode. When I harvest data from templates, I use a bot. Thanks, GerardM (talk) 04:22, 14 May 2014 (UTC)[reply]
@GerardM: You are making invalid statements. Duplicates are invalid entities. They skew aggregate data. You do not know about 4148 humans. You know about 4148 Wikipedia pages, many of which are humans but some are not, and it appears that 10% are duplicates. Which is to say, you know precious little from the mess you have created.
Of course a large percentage of articles created in the last few weeks are not linked to the other articles in Wikidata about the same topic. That is news to you? Big sigh.
user:RobotGMwikt is a bot. Read the page User:RobotGMwikt; that is what is says. It links to the policy WD:Bots, which you need to abide by when you use that account. You are not, and you continue to misuse your account with the bot flag.
You can use your 'GerardM' account with Widar when you are doing tasks that you do not (yet) have permission to undertake with RobotGMwikt. John Vandenberg (talk) 05:38, 14 May 2014 (UTC)[reply]

GerardM has created another 9 items with the Widar bot, with one David Malet Armstrong (Q16872877) duplicate of David Malet Armstrong (Q1173590). Even if 8% was an acceptable error rate, the creation of items is consistently higher than 8% error rate. John Vandenberg (talk) 05:07, 14 May 2014 (UTC)[reply]

I've finished another 50, from May 5, so 10 days have passed since these were created. Of the 50 new items, 12 are dups, or almost 25%.

  1. Q16234963 dup of Vasco Silva Dantas Rocha (Q10389345)
  2. Q16752963 dup of Josip Barišić (Q11728516)
  3. Q16752959 dup of Dan Smith (Q15906708)
  4. Q16752951 dup of Rufo Sánchez (Q16626844)
  5. Q16752942 dup of Antonin Rouzier (Q2857142)
  6. Q16752933 dup of Aleksei Pushkarev (Q16689706)
  7. Q16752924 dup of Aleksandr Kozyrev (Q15732151)
  8. Q16752046 dup of Dritan Smajlaj (Q3715377)
  9. Q16752026 dup of Nadezhda Paleeva (Q16596875)
  10. Q16751951 dup of Liudmila Udobkina (Q16596942)
  11. Q16751921 dup of Yū Kamiya (Q8977221)
  12. Q16751769 dup of Fabiana Santos (Q16596904)

Gerard, I notice you are doing other things rather than fix the above problems. John Vandenberg (talk) 06:32, 14 May 2014 (UTC)[reply]

Thank you John for finding the duplicates. Creating an item for an article is never an error; it is part and parcel of our aspiration to be the source for information in info-boxes. This is impossible if we do not allow for the creation of an item for each article.
The percentage of 20% for items that can be merged is roughly in line with the paper on Portuguese English missing inter-language links. It is therefore an endemic problem. As to what I have done this morning, I have removed statements I identified in my "earl of Mayo" blog post and I have added date of death/birth human information on people who are identified as having died in 2014. By doing this for any and all language, I raise the quality of Wikidata and, I make a foundation for tools that identify duplicates based on such statements, the ones I or you miss. When I recognise duplicates, I do merge them as well. This often happens on statements I added earlier.
So I am fixing issues as I find them, problems however become apparent when you touch on them. The problems are however not mine to solve alone. We are all seeking resolution of the issues we face. My approach is different but valid. It is valid because the RFC has not been contested. I know of issues with bots that produce problematic results but they are better than not having data. I do not attack the bot operators because I know issues will be addressed when we are ready for them. Thanks, GerardM (talk) 09:51, 14 May 2014 (UTC)[reply]
@GerardM:, Which RFC are you referring to? John Vandenberg (talk) 10:30, 14 May 2014 (UTC)[reply]
I think he referring to this one: „Quality is measurable”. --Succu (talk) 14:15, 14 May 2014 (UTC)[reply]
@GerardM: could you confirm which RFC you are referring to please. John Vandenberg (talk) 05:53, 15 May 2014 (UTC)[reply]
Succu has it right. GerardM (talk) 07:37, 15 May 2014 (UTC)[reply]

user:GerardM has used the bot to create another three items, and one is a blatant dup. H. R. Giger (Q16874228) dup of H. R. Giger (Q312640). This is in spite of user:Magnus Manske kindly updating the Creator tool so that the operator needs to override the default behaviour in order to create dups which are so blatant. John Vandenberg (talk) 11:42, 14 May 2014 (UTC)[reply]

user:GerardM continues to create blatant duplicates with the bot flag, sucking in brand new Wikipedia pages without checking the source wiki or wikidata for potential matches. e.g. José Falcó Sanmartín (Q16872876) and José Falcó Sanmartín (Q16874879) are both dups created since the last update here, of José Falcó Sanmartín (Q3186274). John Vandenberg (talk) 23:16, 14 May 2014 (UTC)[reply]

As mentioned above, GerardM is creating new items for brand new Wikipedia pages. pywikibot's newitem.py prevents this happening, unless the operator overrides the defaults. Widar does not have any sort of check, and its users are not examining these pages before importing them. For examples Tim White (Q16762431) (created by Gerard) and Q16210901 (created by GZWDer's bot). Also potential BLP issues - (all by RobotGMwikt) see Q16761666 (w:Hany Helmy Mendoza), Q16755981 (w:Anish Uppalapati), Q16751999 (w:John M. Becker), Q16751666 (w:Xiao Na), Q16751452, and Q16751294. Articles currently at AFD like James Suttles (Q16751743) (w:James Suttles) and Q16751488 (w:Anthony R. Mills) Also lots of pages created by socks/spammers like Q16751655 (see w:User talk:MER-C#John Graden). GerardM's bot is indiscriminately importing items, yet he claims it is not a bot. John Vandenberg (talk) 00:08, 15 May 2014 (UTC)[reply]

And two more obvious dups freshly minted:

Both are Simple Wikipedia articles, which by the nature of that wiki is highly likely to always be duplicates. John Vandenberg (talk) 03:20, 15 May 2014 (UTC)[reply]

Given that John assumes things, He forgets to take the ass out of it all. I do not use pywikipedia bot to create items. I never have. So all his wonderfull crafted assumptions are not worth the attention it gets. GerardM (talk) 06:44, 15 May 2014 (UTC)[reply]

@GerardM, John Vandenberg: Please tone it down. It would be silly if the community had to go through the trouble of banning interaction between you two and/or blocking you two for the sniping going on here.--Jasper Deng (talk) 06:46, 15 May 2014 (UTC)[reply]

I am on record that I want to communicate with John. I also am on record that I feel I am being harassed/stalked. I am on record explaining what I am doing and how I am doing it. I have written how we can improve quality given the size of our database. What I want is a discussion with the topic "how do we improve the quality of Wikidata and how can we achieve goals like adoption of our content". GerardM (talk) 06:53, 15 May 2014 (UTC)[reply]

I endorse removing the bot flag from this account in case further edits are made in contradiction to WD:Bots or if further disruptive edits are made repeatedly. Though, I disagree with Jasper Deng's interaction ban threat as preventing discussion where it is needed would get us to nowhere. Vogone (talk) 07:16, 15 May 2014 (UTC)[reply]

@Vogone: I don't want "discussion where it is needed" to include multiple ad hominem comments.--Jasper Deng (talk) 07:19, 15 May 2014 (UTC)[reply]
Yes, I agree that ad hominem comments do not help anyone here. But don't interaction bans prevent every kind of discussion between this two users? That way we wouldn't have this not-so-unimportant discussion ongoing here and as far as I am aware of there isn't any disruption ongoing at other places here on Wikidata. Vogone (talk) 07:54, 15 May 2014 (UTC)[reply]
What would be the alternative? I'm glad that it's not needed (an interaction ban, which doesn't exist here yet even).--Jasper Deng (talk) 16:21, 15 May 2014 (UTC)[reply]

@Vogone:, do you mean more edits not approved in broad accordance with Wikidata:Requests for permissions/Bot/RobotGMwikt, such as new (since your message) item creation Cassandra Lynn (Q16878835) dup of Cassandra Lynn (Q15140796)? John Vandenberg (talk) 09:09, 15 May 2014 (UTC)[reply]

Yes, I do. The quality of bot-created items should not be much less than the quality of human-created ones. As it isn't acceptable to flood Wikidata with manually created dups it is even less acceptable to do so with a bot flag. Vogone (talk) 15:12, 15 May 2014 (UTC)[reply]
✓ Flag removed. Since the usage of the bot flag not in accordance with WD:Bots has continued, I have just removed the flag from the account. Feel free to reapply for the flag in case you fixed the problems mentioned on this page. Kind regards, Vogone (talk) 15:16, 15 May 2014 (UTC)[reply]
I am surprised as I logged of yesterday from that account and have not used it since. Today I have done other things, I have been mostly away.
Given that I have been accused of things that are imho factually wrong, it is high time that there is a discussion on how to achieve quality, how to achieve our goals. The way I read the revocation of the bot flag is that it was for formal reasons withdrawn. There has been no arguments used about the issues at hand. There was no discussion at all. 15:44, 15 May 2014 (UTC)  – The preceding unsigned comment was added by GerardM (talk • contribs).
@GerardM:What? The bot has performed several actions that it was not approved to do, and you've used it to comment on this discussion. It's an obvious misuse of the flag. The removal was very appropriate. "I have been accused of things that are imho factually wrong" What things? --AmaryllisGardener talk 16:12, 15 May 2014 (UTC)[reply]
@GerardM: And I have blocked your bot account because you made four edits from it after Vogone removed the bot flag. You are to file a new request for bot approval in order to continue using that account.--Jasper Deng (talk) 16:21, 15 May 2014 (UTC)[reply]
The above discussion is preserved as an archive. Please do not modify it. Subsequent comments should be made in a new section.

September 2014[edit]

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.

The permissions were removed by stewards following TBloemink's own request and the WMF statement. At the time I am closing this, the consensus of this discussion (more than 50%) would be for removal of the permissions. Whereas this may be not the best occasion, I would still like to thank TBloemink for his valuable contribution to Wikidata.--Ymblanter (talk) 16:38, 2 October 2014 (UTC)[reply]

TBloemink[edit]

TBloemink (talk • contribs • new items • new lexemes • SUL • Block log • User rights log • User rights • xtools)

After Requests for comment/Privacy violation by TBloemink and JurgenNL and a little AN discussion, this request for removal of TBloemink's adminship is an opportunity for our community to express their refusal to any kind of off-wiki harassment. Since TBloemink did not start a new admin reconfirmation himself, we will have to go through the for him more favourable process of an RFP/R which requires 50% remove votes instead of just 30%. Regards, Vogone (talk) 17:32, 27 September 2014 (UTC)[reply]

TBloeminks admin rights have been removed following a Wikimedia Foundation request. Prior TBloemink also requested the removal of his admin rights. Could a bureaucrat please close this discussion. - Sincerely, Taketa (talk) 13:28, 2 October 2014 (UTC)[reply]
  • Remove Remove Vogone (talk) 17:32, 27 September 2014 (UTC)[reply]
  •  Keep I don't really have a lack of trust in TBloemink. After discussing the matter with him, yes it is a serious thing and yes boarders on harassment but I still have trust in him regardless of the situation. In addition the fact he resigned from his highest position of trust (steward) in light of the matter without the need to open any discussions regarding removal, show to me he does feel sorry for the events. John F. Lewis (talk) 17:35, 27 September 2014 (UTC)[reply]
  • Remove Remove IRL harassment is a large offense, especially when it is a violation of the privacy policy. Inactivity is also a concern I have. --Jasper Deng (talk) 17:38, 27 September 2014 (UTC)[reply]
  •  Keep, to avoid losing a friend. The message is clear, still trusted person imo. Sjoerd de Bruin (talk) 17:54, 27 September 2014 (UTC)[reply]
  • Remove Remove per vogone and JD--DangSunM (talk) 18:01, 27 September 2014 (UTC)[reply]
  •  Keep The tools have not been abused and I do not feel that he has done anything that impacts his trust here on Wikidata. --Jakob (talk) 18:21, 27 September 2014 (UTC)[reply]
  • Remove Remove IRL harassment is a huge offence. Unfortunately, I and many others have lost trust in TBloemink. Even though he has not used his sysop rights inappropriately, being an administrator indicates a high level of trust in the community, which he no longer has. IMO it's hard to believe that some think he should keep his rights with rationales: "yes [it] boarders on harassment but...", and "Keep, to avoid losing a friend". --AmaryllisGardener talk 18:40, 27 September 2014 (UTC)[reply]
  •  Keep Per Jakob. I don't think TBloemink will abuse this power as everyone makes mistakes in life. --Eurodyne (talk) 19:52, 27 September 2014 (UTC)[reply]
  • Remove Weak remove I'll admit I'm on the fence about this one. On one hand, I hate to add insult to injury, and this isn't a role that handles much private info, if at all. And this isn't en.wikipedia where being an admin is a "big thing", and on the harassment scale, it was fairly low. However, three things tip me towards removing: 1) there was significant room for improvement in how he handled the situation, even once the RFC on Meta had begun, 2) his last 50 edits here go back into 2013 and his last logged actions were in July, and he only has 395 edits here (and wouldn't pass a modern Wikidata RFA on that) so he's not very active here anyway and 3) I don't think that someone who recently was involved in this sort of thing should be representing the Wikidata project. Sorry, TBloemink. --Rschen7754 22:32, 27 September 2014 (UTC)[reply]
    • Upgrading to full remove. After seeing Savh and ElfjeTwaalfje's comments, and having thought about all I have seen (including posts to stewards-l) I just don't think he gets what he did, and what is appropriate from here on. --Rschen7754 13:51, 1 October 2014 (UTC)[reply]
  •  Keep per John and Jakob. Jianhui67 talk★contribs 02:39, 28 September 2014 (UTC)[reply]
  •  Keep The underlying case is definitely a problem, but he has already dropped his stewardship, and admin on wikidata is no big deal. --Krd (talk) 07:10, 28 September 2014 (UTC)[reply]
  •  Keep so far it has nothing to do with admin tools on Wikidata. I have never understood an argument "he is badass, don't give him (take away from him) a flag". Flag is given to person who know what to do with the flag, how to do it, and will not do any bad with this flag. Anything related to privacy is oversight area, it is no related to admin flag. All I see in this discussion sounds like "he did something bad somewhere, he is a bad person, let take the flag away from him". Even the beginning of this vote seems like a person attack ("...for him more favourable process..."). And I can't support it in this way. Did he done something wrong on Wikidata or other project using admin flag? No, as far, as i know (from provided info). Is it likely that he will use his admin flag to make bad things on Wikidata? No. Noone presented arguments that last two points are incorrect. That's all arguments that matters for me. -- Vlsergey (talk) 12:24, 28 September 2014 (UTC)[reply]
    • Sorry, but also admins have access to deleted information. And yes, I am of the opinion that real-life stalkers should not be granted access to such information, under no circumstances. Vogone (talk) 19:15, 28 September 2014 (UTC)[reply]
      • From my point of view, anything really private must be oversighted, otherwise not counted as private (i.e. nonrecoverable). There is special tool for that and special flag (oversight) and admin shall not be required to hold not only admin responsibility but oversight responsibility burden as well. Because of that i can't agree to argument "should not be granted access to...". There shall be no access to this information for any admin in first place. Admins have no access to oversighted information. No access -- nothing to worry about when talking about granting or taking back admin flag. -- Vlsergey (talk) 19:35, 28 September 2014 (UTC)[reply]
        • Well, I will say (oversighter hat on) that though that should be the ideal, that doesn't always happen - sometimes, due to language barriers and a lack of awareness of the OS tool, not everything that should be oversighted is oversighted, and also, admins are usually the ones who find stuff that should be oversighted and report it to us. No, we don't hand our new admin candidates a 20 question application to see if they can handle private data, but there is the expectation that if they do run across something like that, they will handle it appropriately. --Rschen7754 06:04, 29 September 2014 (UTC)[reply]
  • Remove Remove per Vogone -- Bene* talk 14:33, 28 September 2014 (UTC)[reply]
  • Remove Remove to prevent possible future (even unlikely) problems. --Jklamo (talk) 16:39, 28 September 2014 (UTC)[reply]
  •  Keep - if he is inactive, he will be desysoped with normal inactive run. And it is not wikidata issue. — revi^ 16:44, 28 September 2014 (UTC)[reply]
    Remove Remove changed my mind. — revi^ 14:52, 29 September 2014 (UTC)[reply]
  • Remove Remove - There's really no valid reason at all why TBloemink should keep his admin rights on Wikidata or be allowed to. EvilFreD (talk) 17:42, 28 September 2014 (UTC)[reply]
  • I'm not convinced that TBloemink has or will abuse his sysop tools here, and as such he has not lost my trust as a WD admin. Ajraddatz (talk) 18:54, 28 September 2014 (UTC)[reply]
    After hearing that TBloemink has continued to brag about the incident, and various other dramas, I am inclined to vote Remove Remove. Ajraddatz (talk) 16:02, 1 October 2014 (UTC)[reply]
  •  Keep Per all above. --Wiki13 talk 20:14, 28 September 2014 (UTC)[reply]
  • Remove Remove I have completely lost trust in TBloemink's judgement on how to deal with privacy and confidentiality issues. I have given it some time to think this over, but for me this means that I really can't trust TBloemink with administrator rights. I feel insecure after what happened and the views on the events and apologies by TBloemink (and JurgenNL) have not taken that away. Lymantria (talk) 09:50, 29 September 2014 (UTC)[reply]
  • Remove Remove I don't know him, but this behavior is not acceptable. An admin should be trustable. He isn't after this. --Succu (talk) 14:20, 29 September 2014 (UTC)[reply]
  • Remove Remove I'm really sorry about this situation, I don't know exactly how it is serious, on the other hand I believe that he will stay active in Wikidata (and this community). Matěj Suchánek (talk) 16:10, 29 September 2014 (UTC)[reply]
  • Remove Remove Off-wiki privacy violation? It won't get much worse. There is no way anyone can trust TBloemink in his position after the mistake he and JurgenNL made. JurriaanH (talk) 14:03, 30 September 2014 (UTC)[reply]
    • "Off-wiki" doesn't do justice to the case. This was a "real life" privacy violation. As in "actually being a real life threath to a living person and his loved ones and giving them a feeling of not being secure in their own house". EvilFreD (talk) 14:10, 30 September 2014 (UTC)[reply]
  •  Keep JurgenNL (talk) 17:44, 30 September 2014 (UTC)[reply]
  •  Keep per Krd. --Steinsplitter (talk) 18:09, 30 September 2014 (UTC)[reply]
  • Remove Remove per the concerns I mention on the RfC, and the given impression that he has not yet understood the problem, as per IRC. Savh (talk) 18:19, 30 September 2014 (UTC)[reply]
  •  Weak keep IRL harrassment imo is not worse than other abuse like socking (compare Russavia). In addition, this user feels sorry for it and this would pretty much not occur again.--GZWDer (talk) 01:08, 1 October 2014 (UTC)[reply]
    @GZWDer: You're joking, right? Right? Hopefully? IRL harassment, the same level of offence as sock-puppeteering? What? --AmaryllisGardener talk 01:19, 1 October 2014 (UTC)[reply]
    @GZWDer: I hope you're indeed joking, I don't see how this is on the same scale as socking on wiki. And additionally, on IRC he doesn't give the impression to be sorry either: He still considers the only thing he did wrong was to walk on a "public street". I'd love to publish IRC logs, but he is unwilling. Additionally, here he (and someone who shares an IP adress with him as well) opposes MoiraMoira, which in light of the occurred events he simply shouldn't have done. The on-wiki apologies seem to be solely with intention to hold on to rights, considering his activity here was limited to 4 days in which he logged (26 July, 10 June, 2 June and 22 May 2014) in the past 6 months. Savh (talk) 07:29, 1 October 2014 (UTC)[reply]
  • Remove Remove Regretfully. His vote against the victim of his visit in the first round of the sysop confirmation procedure on nlwiki (which started today) just decided this for me. Restraint would have been the proper way; I don't doubt the sincerity of his vote, this just shows for me that a usually competent admin and steward right now temporarily lacks the proper state of mind to have the additional sysop rights here. ElfjeTwaalfje (talk) 02:06, 1 October 2014 (UTC)[reply]
    • Tbloemink expressed his loss of confidence in MoiraMoira on 8 september, after she used the word stalker, in a discussion where a third wikipedian, Behh, complained that Tbloemink was cyberbullying Behh on IRC [65]. No action followed because the channel was outside wikimedia. Sincerely, Taketa (talk) 15:12, 1 October 2014 (UTC)[reply]
      • While this is one of the reasons, she performed some odd behaviour in the past which is why I'd like to see a community vote. TBloemink talk 20:33, 1 October 2014 (UTC)[reply]
        • So stalking is ok with you, User talk:TBloemink? --Succu (talk) 21:13, 1 October 2014 (UTC)[reply]
          • I have never approved of any stalking and that is why I have huge regret of my actions after these have been called stalking. So no, it is not ok with me. TBloemink talk 21:46, 1 October 2014 (UTC)[reply]
            • @Taketa: Regardless of whether Behh was right in identifying TBloemink as a cyberbully (and I was prepared to believe TBloemink until now, now I'm not so sure anymore), his reaction to MoiraMoira was way over the top. I think that MoiraMoira made an excellent balanced reply addressing the very real issue of Behh with compassion and empathy while avoiding to accuse a specific user. I certainly couldn't have done it better. There was no need to lose confidence in MM at all, and after what he and Jurgen did, he should have stayed miles away from MM's sysop confirmation procedure. Yet another piece of evidence for me that confirms my assessment that TBloemink should not keep his sysop rights. ElfjeTwaalfje (talk) 22:57, 1 October 2014 (UTC)[reply]
  • Remove Remove I admit that I had to think some days about what I'm going to answer here. After Savh comment above I have to support removal of TB's adminship. --Stryn (talk) 15:40, 1 October 2014 (UTC)[reply]
  • Remove Remove Some things are just so far beyond the pale of acceptability that they (should) disqualify you from holding any advanced permissions anywhere. Courcelles (talk) 17:30, 1 October 2014 (UTC)[reply]
  • Remove Remove People who abuse their access to private info should no longer be trusted with it, it's as simple as that. IJzeren Jan (talk) 21:44, 1 October 2014 (UTC)[reply]
The above discussion is preserved as an archive. Please do not modify it. Subsequent comments should be made in a new section.

December 2014[edit]

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.

 Invalid request, please ask the stewards for assistance. Vogone (talk) 18:46, 5 December 2014 (UTC)[reply]

Cycn[edit]

Goodday,

Because of the changed Wikidata:Deletion policy after the closure of Wikidata:Requests for comment/Redirect vs. deletion I wish to request the removal of my administrator rights. I am categorically opposed to this scheme and I won't be a part of its implementation, but as it has become policy it would be best to disable me from deleting items and this violating this policy.

I will remain active in other ways, as far as the policies allow me to.

With regards, - cycŋ - (talk • contribs • logs) 10:38, 5 December 2014 (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/RfRemoval/2014&oldid=220712639"