This was the other major event of the year requiring disclosure. I have given my word to avoid or minimize commenting on the case or the user, and will do so as best I can, but a few evidence links are needed to substantiate the statements below. The chronological summary includes events discovered later:
- 1. December 2007
During and after the Arbcom 2007 election, I was subjected to a campaign of defamation and harassment by a user (who was community banned in 2009, in part for still engaging in the same attack campaign 1.5 years later).
The attacking user had made a blog post during the election that referenced a Wikipedia edit and attempted to use it indirectly as a way of making a defamatory comment related to the author of that revision ("The author of this revision..." etc). The blog was taken down but had already been spidered and was "in the wild". David Gerard, an oversighter, oversighted the revision as potentially defamatory on the basis that the edit made the blog defamatory (by linking the blob claims to the author of the edit); with the edit removed, the link would fail and the defamation would not exist when the blog circulated at blog aggregators. Moments later it seems he reversed himself and realized it should have been deleted, which would have been sufficient. Gerard emailed Jimmy Wales and stated openly he had incorrectly used the tool and was in the wrong. Jimmy replied a day or so later that he would ask the developers about getting the edit reinstated again. For whatever reason this never happened.
This took place simultaneous to the 2007 election. My election Q&A was 370 K long - I was answering thoroughly in one of our community's toughest elections, and had almost no spare energy or attention for anything on-wiki except writing answers and skimming emails for whether or not action was required, a factor that most ACE candidates will know well and which turned out to be crucial later on.
The oversighting did not affect the election. The user had already had his concerns roundly dismissed by most of the community long before that happened. The purpose of the blog post was to incite off-site, not to educate Wikipedians, with users told they could "expect to hear MUCH more" . The oversighting took place on 7 December 2007 22:02 UTC. By that time about 80% of all votes and personal votes had been cast with the edit being in full view, and even so, the election looked like this (final results). The mood was towards specific rejection of the matter and support for willingness to edit difficult areas .
From election onwards, I studiously ignored the user (who was banned at this point) and all references to him, in order to not be incited to feed the matter. I occasionally heard defamation was continuing elsewhere and a couple of times was witness or party to discussion of this by other users. I did not track this or read it. I did not visit WR. Perhaps I should have, in which case I would have understood comments alluded to by others.
- 2. March to July 2008
Between March and July there were a couple of mentions of "oversighting" but a check of records and logs shows that none of these made clear there was a reference of substantive claim of an actual real oversighting. Many wild and spurious claims had been made, and the tenor of the few mentions was vague ("an oversighting" or "the oversight") or seemed like a future concern - "what if he claims oversight was used".
A few emails were sent by an Arbitrator to the arbcom list. But these were sent to the private Arbcom list, where the only people who knew of the oversighted edits (Jimmy and David Gerard) were not members, and where I automatically filtered and archived them due to being related to the banned user. The older oversight log was inaccessible at the time for technical reasons (too long/no paging) so nobody could view the oversight log > 30 days ago either. The net result was that 1/ those who knew of an oversighting never saw the emails about it, 2/ I auto-archived all emails related to the user so I did not read them, and 3/ the old oversight log was uncheckable in any event.
In July a user asked me about the edits. This was the first time I took note of them, since an on-wiki inquiry is noteworthy.
Q: FT2, thanks for the elaborate explanations! While you are here can you either confirm or deny that a few of your edits presented by [user] were oversighted? Can you recollect the rationale for the actions?
A: I'm not sure how I could tell, this being the first mention of any such to me. If this was in the last 30 days - the duration of the oversight log - I can check for myself though. Be aware there is no ability to search the oversight logs by 'name of editor of oversighted revision' though [...]
This was 4 July. OrangeMarlin exploded around 27 June while away. My computer power supply failed a day or two after return. Someone very close to me was dying 2-3 July. I was using minimal technology and trying to keep up, taking care of my family, and lacking sleep. I had not registered the oversighted edits as an "issue" and had not had them raised to me as a real matter before; tiredness and lack of tracking the affairs of the banned user caused me to not recall the few brief mentions. I believed my answer was 100% accurate, honest, and complete, that this was the first mention of them to me (in effect it almost was, previous mentions had been vague in the extreme or not seen by me), and I was trying to help. I checked the oversight log (of course) but having little idea what edits to look for, the first thing I noticed was that the oversight log was not working. I said as much on-wiki, said what I believed, and promised to check it up when the log was back. I thought no more of the exchange.
I also gather that some assumptions made by others may have been in error, 1/ it was assumed the emails had gone to a list where I or someone would have recognized their significance, and 2/ there was a separate oversighting of material by the user (100% legitimate) discussed on May 3 which may have been conflated with the previous ones.
- 3. Sept - Dec 2008
The user was rebanned in September. The Oversight log was re-enabled. I found the edits and spoke to David Gerard who confirmed the matter and copied me his disclosure/apology to Jimbo. I was sent copies of various old (2007 and early 2008) emails between others that had not been sent to me before related to the user's actions.
The user appealed his ban and the entire matter went to RFAR. On December 9, in an unrelated search related to the RFAR case, I found an old email from Jimbo to myself. It was written during the election a year before in 2007 and stated that edits had been oversighted by mistake but this was going to be reversed.
This was a problem. Unknowingly I found from an email in December 2007, that I had in fact misled the community in July 2008. Not that it would have made a practical difference; I still hadn't any involvement or done wrong, and the email had arrived at an immensely busy time, but I'd been sent an email about it and not remembered.
In judging honesty, consider my next action. This was an email that could only put me in a very negative light. Nobody else knew about this email. It was a year old and Jimbo had surely forgotten its existence if (as it seemed) he couldn't remember the entire matter much himself. Not one other person on the project knew. I could have ignored it and nobody would ever have known. Instead I emailed Jimmy and David Gerard the same day, disclosed the discovery, and asked their advice. The replies were from Gerard - to disclose. The problem was that the oversight had been blown up into a question of the proper use of oversight, I think partly by people targeting David Gerard (not 100% sure). A full disclosure would also lead to followup questions focusing on real-world information leaked from Arbcom's list, and that was also a problem.
(Briefly it had been asked whether had he "pulled strings". The election was all but final at the time of oversighting so it had no effect whatsoever. Gerard had indeed met me (unexpectedly) at an "open source" discussion and on learning who I was, said "welcome to hell, you'll need a shovel" or the like, as the election was visibly likely by then. But to discuss or confirm this would confirm people's real-world information and the accuracy of a leak from Arbcom's mailing list. Any comment on this point - whether acknowledge, confirm or deny - had to be weighed against the potential impact on the privacy of individuals and the list.)
I consulted widely. I asked advice from Arbitrators individually and the committee as a whole, from people I trusted and people unconnected. The advice was conflicting. Nobody could suggest a concrete disclosure that would deal with the original matter but not lead to likely breach of the privacy leak issue.
I emailed Thatcher and Jimbo at some point: "But mainly [...] it's already my long standing intention to be open and seek full disclosure [...] in the next few days, and several have tried to talk me out of it. I've declined citing that I want full accountability to the community. If afterwards you are not satisfied, I will have said all that there is to say, and the community and Jimbo may assess me as they will."
- 4. I stand down.