Wikipedia:Arbitration Committee/CheckUser and Oversight/2011 CUOS appointments/CU/AGK

AGK

 * Nomination statement (250 words max.)

I joined the Audit Subcommittee to replace bahamut0013, and in this capacity have made extensive use of the checkuser tool so as to become familiar with its use. Soon, I will be scaling back my use of checkuser - because, as an auditor, I do not intend to routinely use the tools. But from my term on the subcommittee thus far, it is clear to me that we require more checkusers. (There is an especial need for checkusers for routine requests, at SPI, so that the most active checkusers are not shouldering as much of the workload as they presently are.)

My public work with the checkuser tool is reviewable by browsing my contributions to the Wikipedia namespace - the results of most of my checks have been posted to a subpage of Sockpuppet investigations. As somebody with access already, I have the technical aptitude to use the checkuser tool, and, as a current auditor, I have the sensitivity and knowledge of the privacy and checkuser policies to use the tool with the appropriate discretion.

This is a request to retain access to the checkuser tool after I leave the subcommittee, and an offer to devote the required daily time to this important area. There is a word limit on nomination statements, so I have had to speak briefly and generally, but I am happy to answer questions on any specific areas of interest. Thank you for your consideration.

Standard questions for all candidates
Please describe any relevant on-Wiki experience you have for this role.


 * A: I have been a content contributor for over six years now (and an administrator for five), so I am by now a "seasoned contributor"; this is undoubtedly useful as a general matter, especially when evaluating checks and weighing up the merits of a request for checkuser. I already have access to the checkuser function. When I joined the subcommittee, I knew probably had the technical knowledge to understand the tool, but I did not ever think that there would be such a workload for the regular users of the tool; nor did I know that I would be able to use it as competently as I can. (I rather imagined I would probably be able to use it just well enough to scrutinise checks and to not break the whole thing.) Now that I have had access to the tool, I know I can be of use; this experience is probably the most relevant thing I can say here.


 * (Note: The statistics on checkuser use, published to a subpage of WP:AUSC, are updated infrequently, so I am not yet listed there; I imagine I soon will be.) AGK  [&bull; ] 13:21, 26 September 2011 (UTC)

Please outline, without breaching your personal privacy, what off-Wiki experience or technical expertise you have for this role.


 * A: I am qualified to the higher, later levels of secondary school in Computing, and I am more familiar with the concepts that relate to user access data than are most laymen. My history of checks have been solid (with the few re-checks I have asked for being concurrent… *knock wood* that I don't mess up my next check now ;)), and I have always asked for a second opinion when I have been unsure of my results. AGK  [&bull; ] 13:21, 26 September 2011 (UTC)

Do you hold advanced permissions (checkuser, oversight, bureaucrat, steward) on this or other WMF projects? If so, please list them. Also, do you have OTRS permissions? If so, to which queues?
 * A: I have checkuser and oversight on the English Wikipedia, and an OTRS account with access to the oversight-en-l queue. As part of these roles, I am identified to the foundation. AGK  [&bull; ] 13:21, 26 September 2011 (UTC)

Questions for this candidate
Being a checkuser, would you be willing to help with the Checkuser backlog at WP:ACC as there are usually up to 6 requests waiting about 5 days+? -- DQ  (t)   (e)  19:24, 26 September 2011 (UTC)


 * A: Yes, of course. I have actually already offered to do so, at User talk:OverlordQ, but the ACC interface admin has not been online since my message. AGK  [&bull; ] 21:21, 27 September 2011 (UTC)

Would you be proactive in looking at the open cases at SPI to see if they could use a checkuser? -- DQ  (t)   (e)  19:24, 26 September 2011 (UTC)


 * A: Yes, and in fact I already am, because I recognise that the clerks are very busy, and that it can speed up the process if the checkusers themselves are actively looking for requests in which CU data has not been requested but might be useful. AGK  [&bull; ] 21:21, 27 September 2011 (UTC)

You mentioned an SPI backlog, but whenever I look at the page there don't seem to be any cases requiring checkuser attention there. Am I looking in the wrong place? :-) --(ʞɿɐʇ)  ɐuɐʞsǝp 19:41, 26 September 2011 (UTC)


 * A: As often as not, there are no cases awaiting checkuser data (they would be listed under "Awaiting checkuser"). But very often, there are requests awaiting clerk approval, or even cases that have not included a request for checkuser data, but that would require (or at least could find use for) the input of somebody with CU access. As well as the SPI workload, there is also a lot of use for checkuser data at ACC and at CAT:UNBLOCK. However, our current checkusers work very hard, so there is often not a huge backlog (although there is a sizeable workload); as I am sure you are aware, what is more important is that there are more highly-active checkusers, in order to allow those who run the majority of the checks to not shoulder as much of the burden. AGK  [&bull; ] 21:21, 27 September 2011 (UTC)

As a CheckUser, you will likely, from time to time, coordinate and communicate with the Stewards. What cross-wiki experience can you bring that can help out not only the Stewards, but editors, administrators, and CheckUsers on other wikis? –MuZemike 21:27, 26 September 2011 (UTC)


 * A: From my experience with the tool, most Steward co-ordination takes place on the mailing list, Checkuser-l. Users with CU access on Enwiki can assist the stewards on other wikis by sharing information when needed. If there is an especial need for co-ordination at other venues, I don't think I am aware of it. AGK  [&bull; ] 21:21, 27 September 2011 (UTC)

In your own words, what are the main differences between the WMF's CheckUser policy and the privacy policy? –MuZemike 21:27, 26 September 2011 (UTC)


 * A: The checkuser policy primarily governs the use of the checkuser function, whereas the privacy policy governs the access and dissemination of restricted information (such as IP addresses). There is much overlap between the policies, but essentially the former dictates when the CU tool may be used, whereas the latter dictates when we may access, and in what cases we may release, personally-identifiable data. (Additionally, potential violations of these policies are investigated separately: the CU policy is enforced by the Audit Subcommittee, and the Privacy Policy by the Wikimedia Foundation's Ombudsman Commission.) AGK  [&bull; ] 21:21, 27 September 2011 (UTC)

Under what circumstances do the above policies give on the release of CheckUser data? –MuZemike 21:27, 26 September 2011 (UTC)


 * A: The privacy policy allows the release of checkuser data under the following circumstances: 1) by judicial order (or subpoena); 2) with the consent of the subject of the data; 3) to investigate suspicions of sock-puppetry; 4) where the subject of the data is a rogue web spider; 5) to allow an abuse report to be filed with the ISP of the subject of the data; 6) or in order to "protect the rights, property or safety of the Wikimedia Foundation, its users or the public" (which is simply an emergency provision). The checkuser policy is more restrictive, and recommends as a general matter that checkuser data is not released. It is as a result of the checkuser policy that users with checkuser access will usually answer "like a magic 8-ball", and that results are generally given only in non-specific terms (eg. "Link between these accounts is confirmed/likely/unlikely", "Users are editing from same region and ISP, so link is likely", and so on). AGK  [&bull; ] 21:21, 27 September 2011 (UTC)

Give some examples on when CheckUser requests of a sensitive nature or discovered CheckUser results of interest that would not be posted on-wiki. –MuZemike 21:27, 26 September 2011 (UTC)


 * A: Off the top of my head, I would imagine that results should not be posted publicly (but instead referred to ArbCom) where they suggest socking by a seasoned contributor and/or a user with restricted access (such as an administrator or functionary). AGK  [&bull; ] 21:21, 27 September 2011 (UTC)

(relatively minor question, given the circumstances) You were appointed to ArbCom's Audit Subcommittee on September 3, to take over for the now late User:Bahamut0013. Understanding that you have only been in this subcommittee for about 3 weeks, what experience, if any, do you bring from there? I'm not looking for specifics obviously, but general experiences. –MuZemike 21:27, 26 September 2011 (UTC)


 * A: At this point, I am probably fully competent in its use (and certainly, I have ran all my checks without issue). The only area in which I am not yet fully competent is the knowledge of which ISPs distribute IP addresses dynamically and which give semi-static or static addresses (although I have picked up the worst offenders); and indeed, I imagine that it would take some months and many checks to fully acquire this knowledge. Therefore, the primary experience that I bring is a full, pre-existing knowledge of how to actually use the tool. Also, coming from a background as an Auditor, I am perhaps more acutely aware than most new CUs of the demands of the privacy policy, and the need to be perpetually evaluating the merits of requests for checkuser at SPI and related processes. The present team is, of course, not to be faulted on this note, and I do not for a moment mean to suggest that any of the candidates would violate the privacy or CU policy if granted access; but I do think that, as a candidate with a background in the AUSC, it is an advantage that I have, since my first day of using the tool, always placed much weight on these policies. AGK  [&bull; ] 21:21, 27 September 2011 (UTC)

Checkuers are often relied on to determine whether someone is using anonymising proxies to perform their sockpuppetry. Please describe your general experience in this area. Please also describe, preferably with an example, how you (would have) suspected, identified, confirmed, and blocked a socking open proxy on Wikipedia. -- zzuuzz (talk) 17:16, 28 September 2011 (UTC)


 * A: A number of tools use honeyproxies to maintain an index of anonymising proxies; these are the easiest option for identifying an open proxy. One could also run a manual check, by opening a manual terminal connection to the address of said suspected proxy, then trying to connect to a web page. If the connection works, then one must ascertain whether the proxy is passing IP information. I would probably use my Chrome extension that allows one to connect through a proxy, and then I would use something like whatismyproxy [dot] com to determine whether the proxy is passing data or not; if it is not, then it is anonymous. I would block with Openproxy for an appropriate period (proxies can be anonymous for years, can be online for a few days before "closing", or anything in between). One could manually check the proxy using nmap or something similar, in lieu of checking it against an index (which might be useful if the socker is very inventive, and is actively maintaining an open proxy somewhere that is used only for editing WP—and therefore would probably not appear on a honeypot site). I will not pretend that there are not various, well-written documentation pages for those who are new to, or need a refresher on, OP checking.WikiProject on open proxies/Guide to checking open proxies explains the process rather much better than I can; open proxy checking for me is a little like the checkuser tool—a lot of puzzling and mulling and digging. It is, however, not something that I find challenging (as I said, I'm more technically adept than the layman), and I am comfortable with the important role it has in the work of somebody with CU access. AGK  [&bull; ] 22:32, 28 September 2011 (UTC)

Comments

 * Comments may also be submitted to the Arbitration Committee privately by emailing 


 * I wasn't really soliciting any particular kind of answer with my question and as such there was no right or wrong answer, but as it turns out the answer you gave was good. :-) --(ʞɿɐʇ) ɐuɐʞsǝp 20:22, 29 September 2011 (UTC)
 * Thank you :). I forgot to mention Unblock-en-l; having somebody with the CU tool read e-mails has proven useful, since I re-enabled my subscription a while back. Perhaps… I can tempt you with a free, no-obligation trial subscription…? You can also choose from your choice of free Parker pen or generic cuddly toy! AGK  [&bull; ] 22:27, 29 September 2011 (UTC)


 * In my experiences with Anthony, he comes across as a very wise administrator and has a good head on his shoulders. My 5c probably doesn't mean much, but I see no reason why AGK would do poorly with the CU tools. Steven Zhang  The clock is ticking....  10:59, 3 October 2011 (UTC)
 * I love AGK's commitment level to reduce backlogs abroad as an AUSC member, and would like him to continue this with the normal use of tools. -- DQ  (t)   (e)  17:46, 3 October 2011 (UTC)


 * Thank you, Steve and Delta :). AGK  [&bull; ] 21:31, 3 October 2011 (UTC)


 * AGK is among the most trustworthy administrators I know, combined with his experience with checkuser and the audit subcommittee, I highly recommend for him to get the checkuser tools. Secret account 19:18, 4 October 2011 (UTC)
 * This is really a no-brainer. T. Canens (talk) 04:12, 5 October 2011 (UTC)
 * If we were forced to elect a dictator-for-life who would permanently oversee and have overriding control of the project, not only would AGK be at the top of my list... he would be the only one on my list.  Trusilver   06:20, 6 October 2011 (UTC)