User talk:Jimbo Wales

Archiving

Could someone please manually archive this page and could other people please not revert it? If there's something still to be said on the old threads, let's start a new one. I am hopeful that the archiving bot will start working again once we reduce the size of the page. Alternatively, I wonder if there is a 200 page limit or something. I don't have time to properly investigate!--Jimbo Wales (talk) 14:12, 25 January 2016 (UTC)

Done; I put the archivebot number at 202. --Francis Schonken (talk) 14:28, 25 January 2016 (UTC)

(edit conflict) Alot of the archive pages are pretty small. Given they are low traffic, they could always be bundled together or something to make fewer pages overall. Mine are about 4 times as big so this archives could be condensed into 50 pages or so. Cas Liber (talk · contribs) 14:28, 25 January 2016 (UTC)

Had a look at the archivebot parameters:

• removed "|index=no", which according to the User:ClueBot III documentation appears to be doing nothing in this case;
• set "|minkeepthreads=" to 1 instead of 2
• set "|maxarchsize=" to 350000 instead of 250000 per Cas Liber's suggestion above (which is not "4 times as big", but about the size of the archiving that I performed earlier to get the apparent stumbling block out of the way. --Francis Schonken (talk) 14:53, 25 January 2016 (UTC)

Suggestion by User:Casliber is typical of what shouldn't be done. Computer problems have to be solved by computer science, not by undocumented adhoc patches. If the archiving bots are in trouble, this is a general problem, that could mess a lot of pages, the administrative pages among them (ANI and the like) and could happen as well on other projects (like meta).
Changing maxarchsize= to 350000 will not result in 350000-sized archives (this is documented).
About the previous problem, I found that at https://en.wikipedia.org/w/index.php?title=User_talk:Jimbo_Wales&diff=next&oldid=482365272, archiving bot was changed from MiszaBot/Lowercase sigmabot III to ClueBot III. This was done for a not documented reason, at 16:59, 17 March 2012. Part of the changes were done using a pair of <nowiki>...</nowiki> to comment out the previous configuration. But this was wrong, since a talk page is a mix of code and data. The pair <nowiki>...</nowiki> disallows the inner {{User:MiszaBot/konfig}} to be recognized as a transclusion. No more, no less. This doesn't avoid this inner {{User:MiszaBot/konfig}} to be recognized as a configuration file by some Bots (at least by OneClickArchiver). And from there, the two counters lived independently.
It could be useful to undertake a full review of the archiving bots, especially a review of how they articulate with each other (auto, oneclick, indexing, etc.). Perhaps a skilled, professional, paid, accountable programmer (put in charge following the adequate commanding chain) could help ? Pldx1 (talk) 15:20, 25 January 2016 (UTC)

Additionally I put the "|age=" parameter to 24 now (which equals Miszabot's prior "old(1d)" algo parameter I suppose.

I completely removed the commented-out Miszabot archivebot configuration. If that's undesirable for the OneClickArchiver functionality, I'd revert to the plain Miszabot configuration (without commenting out, without double configuration of a second archive bot: maybe that's where the auto-archiving went belly-up). --Francis Schonken (talk) 15:39, 25 January 2016 (UTC)

The solution from the skilled, professional, paid, accountable programmers (well, how accountable they are is a matter for debate) was a product called "Flow". How that handles archiving, I don't know.

I'm skeptical that the size of the page, in and of itself, is the problem. The longer we wait for the two sections below to get archived, the more apparent it becomes that the problem is on the back end. The bot was "sleeping" from 07:46, 24 October 2015‎ to 23:33, 9 November 2015‎, a gap of over 600 edits, while a dozen one-click archives were done. The page was briefly below 100,000 bytes during that timeframe, yet even that wasn't small enough to wake up the bot. It eventually resumed archiving when the page was just over 120,000 bytes in size. I'm guessing that we have a similar, as yet unresolved issue, this time. Wbm1058 (talk) 16:56, 25 January 2016 (UTC)

Hmm, "The core got started on a broken Labs instance..." whatever that means. Wbm1058 (talk) 17:05, 25 January 2016 (UTC)

• Flow... I was more prudent in my requirements. Better have a chief of project who understands what will be on input, and what should be on output. Auditing a set of code is something else... and a good starting point for mutual confidence. By the way, {{User:ClueBot III/running}}
  The current status of ClueBot III is: Running
  (was Not running - last edit was Wikipedia:Biographies of living persons/Noticeboard 41032s ago when posting) Pldx1 (talk) 18:39, 25 January 2016 (UTC)

  Special:Contributions/ClueBot III: last update as of this writing was 19:08, 25 January 2016 to archive Talk:Disruptive innovation. As I said, the bot is running, which makes it more curious why it isn't archiving this particular talk page. The archive of the BLP noticeboard was several edits ago for that bot. So the User:ClueBot III/running is kind of bogus. That is updated by DamianZaremba Scripts, not by CB3 itself. I'm sure that the bot is "not running" most of the time. It's likely scheduled to run at specified intervals. It runs at those intervals, performs whatever archives it deems are required, then sleeps until the next scheduled interval. If enough time passes between CB3 edits, perhaps because there is nothing yet ready to archive, then the DamianZaremba Script seems to arbitrarily decide that it "is not running", even though it is; it just has nothing to do. And what's the point of a script to monitor whether it's up or not, if the volunteer operators are unresponsive when it's down? Wbm1058 (talk) 19:24, 25 January 2016 (UTC)

  4492 transclusions of User:ClueBot III/ArchiveThis. So I suppose it takes a while to walk through them all; perhaps the bot never does take a rest, but just begins another cycle through them as soon as it finishes cycling through all 4492. Would be interesting to run a test to see how long it takes to get back to a page that it's just archived. Wbm1058 (talk) 20:02, 25 January 2016 (UTC)

Auto-archiving on this page broken? Providing a test case

Don't write anything in this section, this is a test... Pldx1 (talk) 19:12, 20 January 2016 (UTC)

This is just another test

If ClueBot III is working as documented, then this section should be archived very quickly! Wbm1058 (talk) 15:49, 25 January 2016 (UTC)

James Heilman

So the James firing is all clear, solved and archived by now. There was no need to condense or conclude. Glad to know. (That is: not a broken archive bot decided so, but someone involved). -DePiep (talk) 20:51, 25 January 2016 (UTC)

It's clear to me. It has been explained to you. If you don't agree, that's fine. We don't all have to agree on everything.--Jimbo Wales (talk) 22:06, 25 January 2016 (UTC)

James Heilman: not trusted by the WMF board, but allowed to edit?

New/remaining question: if editor James Heilman/Doc James is not trusted by the WMF board then why is he still allowed to edit? -DePiep (talk) 20:51, 25 January 2016 (UTC)

DePiep, I'm certainly no fan of the WMF's current management but that's a ludicrous question. Around 99% of the people on Wikipedia are people I'd trust to edit Wikipedia, but not trust with decision-making for a global institution with $80 million in assets. ‑ Iridescent 20:56, 25 January 2016 (UTC)

I found three fallacies in your reply. Prize? -DePiep (talk) 21:57, 25 January 2016 (UTC)

What fallacies do you see in the answer exactly?--Jimbo Wales (talk) 22:06, 25 January 2016 (UTC)

• So . . . was the actual reason for the removal of Doc James ever answered by Wikipedia's founder?! This archiving feels more like a more a "brush off" then an answer to the situation. --MurderByDeletionism^"bang!" 21:39, 25 January 2016 (UTC)

It's been answered clearly. As a quick review - my vote to remove him was because of a pattern of behavior and actions that I viewed as violating the trust and values of the community. One example emerged clearly after he was removed - he made a false claim about why he was removed, and I got a unanimous statement from every board member involved that it was false. The community deserves better than that.

James has made a lot of noise about why he was dismissed which is utter and complete bullshit. He wrote a nice piece for the Signpost about transparency which implied that the board got rid of him for wanting more transparency. Utter fucking bullshit.--Jimbo Wales (talk) 22:06, 25 January 2016 (UTC)

Arnnon Geshuri

The discussion may have been archived away from this page, but the controversy surrounding the ill-advised appointment of Arnnon Geshuri to the WMF Board of Trustees isn't going away anytime soon. Joe Mullin, "Wikipedia Editors Revolt, Vote “No Confidence” in Newest Board Member," Ars Technica, Jan. 25, 2016. Time for somebody to do the right thing. Carrite (talk) 21:05, 25 January 2016 (UTC)

Dude - it's archived. That means that there's nothing requiring further response, right?  pablo 22:03, 25 January 2016 (UTC)