[00:54:59] MySQL on z-dat-s4-a is CRITICAL: Cant connect to MySQL server on z-dat-s4-a (146) [01:54:58] MySQL on z-dat-s4-a is CRITICAL: Cant connect to MySQL server on z-dat-s4-a (146) [02:08:11] EXPLAIN doesn't work anymore :-( [02:10:01] What happens when you try? [02:10:09] It never really worked well with views before. [02:54:59] MySQL on z-dat-s4-a is CRITICAL: Cant connect to MySQL server on z-dat-s4-a (146) [02:57:49] question: I emailed ts-admins a day or two ago about resetting my password/pubkey. I haven't heard back yet, is there anyone around who can verify that it actually went through? [03:54:59] MySQL on z-dat-s4-a is CRITICAL: Cant connect to MySQL server on z-dat-s4-a (146) [04:54:59] MySQL on z-dat-s4-a is CRITICAL: Cant connect to MySQL server on z-dat-s4-a (146) [05:01:47] @replag all [05:01:47] Matthew_: s1-rr-a: 1s [-0.00 s/s]; s1-rr-a-wd: 24s [+0.00 s/s]; s1-user: 1s [-0.00 s/s]; s1-user-c: 8s [-0.01 s/s]; s1-user-wd: 24s [+0.00 s/s]; s2-user: 24s [+0.00 s/s]; s2-user-c: 8s [-]; s2-user-wd: 23s [+0.00 s/s] [05:01:48] Matthew_: s3-user: 2m 9s [-0.20 s/s]; s4-rr-a: 8s [-]; s4-user: 8s [-]; s4-user-wd: 24s [+0.00 s/s]; s5-rr-a: 8s [-0.00 s/s]; s5-user: 8s [-0.00 s/s]; s5-user-c: 8s [-]; s5-user-wd: 23s [+0.00 s/s] [05:01:49] Matthew_: s6-user: 15s [-]; s6-user-wd: 18h 54m 21s [+1.00 s/s]; s7-user: 11s [-0.11 s/s]; s7-user-wd: 23s [-0.26 s/s] [05:54:58] MySQL on z-dat-s4-a is CRITICAL: Cant connect to MySQL server on z-dat-s4-a (146) [06:54:59] MySQL on z-dat-s4-a is CRITICAL: Cant connect to MySQL server on z-dat-s4-a (146) [07:54:59] MySQL on z-dat-s4-a is CRITICAL: Cant connect to MySQL server on z-dat-s4-a (146) [08:54:58] MySQL on z-dat-s4-a is CRITICAL: Cant connect to MySQL server on z-dat-s4-a (146) [09:54:59] MySQL on z-dat-s4-a is CRITICAL: Cant connect to MySQL server on z-dat-s4-a (146) [10:54:58] MySQL on z-dat-s4-a is CRITICAL: Cant connect to MySQL server on z-dat-s4-a (146) [11:54:59] MySQL on z-dat-s4-a is CRITICAL: Cant connect to MySQL server on z-dat-s4-a (146) [12:54:59] MySQL on z-dat-s4-a is CRITICAL: Cant connect to MySQL server on z-dat-s4-a (146) [13:54:59] MySQL on z-dat-s4-a is CRITICAL: Cant connect to MySQL server on z-dat-s4-a (146) [14:50:41] hi all. TS admin here? :-) [14:52:20] Luxo: what do you need? [14:54:59] MySQL on z-dat-s4-a is CRITICAL: Cant connect to MySQL server on z-dat-s4-a (146) [15:20:00] join wikipedia-de [15:20:18] Oops [15:54:59] MySQL on z-dat-s4-a is CRITICAL: Cant connect to MySQL server on z-dat-s4-a (146) [16:54:58] MySQL on z-dat-s4-a is CRITICAL: Cant connect to MySQL server on z-dat-s4-a (146) [17:54:58] MySQL on z-dat-s4-a is CRITICAL: Cant connect to MySQL server on z-dat-s4-a (146) [18:54:59] MySQL on z-dat-s4-a is CRITICAL: Cant connect to MySQL server on z-dat-s4-a (146) [19:44:08] hello all [19:44:36] Dispenser: Which database and which wiki? [19:44:47] sql -u enwiki_p [19:46:03] I'm guessing its related to the MySQL upgrade as its affecting other servers as well [19:46:53] mm, yes seems so [19:54:59] MySQL on z-dat-s4-a is CRITICAL: Cant connect to MySQL server on z-dat-s4-a (146) [19:55:03] Dispenser: I will speak to Nosy about a patch. It may take some days [19:55:15] Ok, thanks [20:00:13] DaB. * [Toolserver-announce] MySQL: explain and show are not working [20:02:01] hello [20:02:07] maintenance tonight [20:03:13] DaB. * Re: [Toolserver-l] JIRA session loss [20:22:56] Its almost certainly better to resurrect JIRA, but might not happen considering the Toolserver death sentence [20:27:13] Matthew Bowker * Re: [Toolserver-l] JIRA session loss [20:29:13] Hersfold * Re: [Toolserver-l] JIRA session loss [20:53:33] Dispenser: toolserver should stay for really some time as i heared [20:54:26] have to reboot hawthorn [20:54:59] MySQL on z-dat-s4-a is CRITICAL: Cant connect to MySQL server on z-dat-s4-a (146) [20:55:16] will reboot now [20:55:27] willow is going to be rebooted after that [20:55:36] o_o [20:55:48] willow too? [20:55:51] when? [20:55:57] when hawthorn is back [20:56:05] i better do one machine after the other [20:56:18] ok, let me know before so i can close and finish my stuff there pls [20:56:54] i think you have at least 5-10 minutes [20:56:58] ;) [21:02:16] ok hawthorn is back [21:02:52] then i will reboot willow in some minutes [21:14:31] False Alarm: The system willow will not be brought down. [21:14:32] nosy: are you making fun? [21:14:32] o_O [21:14:42] You aren't going to learn IA64 assembler because it has no future, the WMF has declared the same of the Toolserver [21:14:58] not me [21:15:19] afaik [21:15:31] so is there reboot or not? [21:16:31] yes just now [21:17:01] You don't think you could've set a timer!? [21:17:18] there was a timer [21:19:53] Dispenser: sure but the replacement technology has to be there [21:20:03] this still takes some time [21:20:27] and not only technology but migration process too [21:20:58] willow rebooted [21:22:18] Like 10% of Toolserver uses is to put back what WMF removed [21:23:51] Dispenser: sorry i dont understand the sentence due to my lack of english could you please try to express in another way please? [21:24:53] need to reboot wolfsbane [21:26:16] how about nightshade? [21:27:07] 10% of what people use and develop on the Toolserver is for running software that the Wikimedia Foundation has disabled or refuse to run in a production enviroment [21:27:14] nightshade seems to be fine and is a linux so wont touch it tonight [21:28:29] Dispenser: so this means there is a need to a development/community system do you think? [21:29:34] i think it might be fine to communicate this and keep this in mind. still a production system for production ready stuff that is maintained by more people then on the ts would be good [21:29:44] WMF policy is it needs to written by someone who they can trust. They only trust people who they pay. [21:29:47] And their WMF labs MMP ain't going to work. It is the same reason I don't slog though other people's illogically and uncommented code. [21:29:59] imho you need a staging system for development anyway [21:30:09] this has to be discussed as i understnad [21:30:28] but it will take a lot of load off the ts *hope* [21:32:32] BTW, what's the best run a query (SELECT DISTINCT rev_sha1) for 200 hours? [21:33:40] Dispenser: go ahead you pain-in-the-ass user :D (still thats fine) [21:34:12] wolfsbane is back [21:34:34] Dispenser: 200 hours?! can't you get that information from the XML dumps? [21:35:24] Maybe, it needs to be in a database to check if the edit was a revert [21:36:25] need to reboot turnera [21:36:42] I figured out a new better metric for user engagement. [21:39:27] [2012-12-16 21:29:44] WMF policy is it needs to written by someone who they can trust. They only trust people who they pay. [21:39:34] Wikimedia has never deployed volunteer code. [21:39:39] That would be insane! [21:40:23] lol...fucking open source...all written by students... [21:40:35] and this open encyclopedia... [21:40:38] sorry [21:40:41] i stop [21:40:45] :-) [21:41:31] Dispenser: The Wikimedia Foundation wants to kill the Toolserver, but I'm not sure they have the resources currently to support DB replication. [21:41:47] And until that exists (and is stable), the Toolserver will continue to remain popular. [21:42:11] Though the Toolserver is no beacon of stability and we ought not wear rose-colored glasses regarding it. [21:43:00] Wikimedia has never deployed volunteer code. [21:43:06] ... I hope this is sarcasm... [21:44:13] I was responding to Dispenser's silliness. [21:44:33] Though volunteer code does take much longer to get deployed. [21:44:51] Usually longer by months or years. [21:45:34] Volunteer code gets deployed on the same schedule as employee code [21:46:01] Though in some cases it might take longer to get into repos [21:46:36] Unless we're talking about new extensions, that is. [21:46:45] Unless, unless, unless. ;-) [21:47:42] GeoHack extension running on the Toolserver that I've been "maintaining" is 7 years old [21:47:59] wow [21:48:04] Isn't there a MediaWiki extension replacement now? [21:48:07] GeoData? [21:48:10] is the toolserver running since then too? [21:48:18] The Toolserver started in 2004ish, I think. [21:48:22] There's a history page somewhere. [21:48:34] https://wiki.toolserver.org/view/History [21:48:50] I suppose 2005 would be more accurate. [21:49:27] GeoData, they never asked me, 'cause the person that's been working with community and is aware of all the edge cases is easily replaced, right? [21:50:08] You can file bugs now. ;-) [21:50:31] Is that how you submit resumes now? [21:50:37] Generally we want highly used tools/scripts to be moved in MediaWiki extensions. [21:50:43] into * [21:50:51] The Toolserver is nice for development. [21:51:04] Krinkle * Re: [Toolserver-l] JIRA session loss [21:51:11] But when a tool needs stability and reliability and code sanity, it should be in core or in a MediaWiki extension. [21:51:21] I don't think there's much dispute about this. [21:51:35] other than it being in PHP of course [21:51:57] That's not really a problem. [21:52:30] Most MediaWiki-related code is PHP. But if there's a really good reason, you can keep code in C, Python, or whatever else. [21:52:34] Particularly in an extension. [21:52:48] You need a good reason, though. A consistent environment is a Good Thing. [21:52:56] The environment is optimized for PHP, not for other languages. [21:53:35] will reboot clematis [21:55:09] MySQL on z-dat-s4-a is CRITICAL: Cant connect to MySQL server on z-dat-s4-a (146) [21:56:07] Susan: is toolserver to be replaced because it is unstable? [21:56:27] nosy: It was never really clear why the Wikimedia Foundation decided to try to kill it. [21:56:37] I think they only want to support Wikimedia Labs. [21:56:46] So the death of the Toolserver is just a byproduct or a symptom. [21:57:07] There are also reasonable arguments that the Toolserver's setup (per-user model) isn't the best for collaboration, code review, deployment, etc. [21:57:41] the structure is not scaling on all levels too [21:57:44] I don't think stability was cited by the Wikimedia Foundation, but it is something that I think gets overlooked in the discussion about Labs v. Toolserver. [21:57:48] it feels like in the 90s ;) [21:57:59] Sourceforge effectively demonstrates that they'll have to limit projects [21:58:27] Limit projects? [21:58:52] clematis is back [21:59:24] * Susan peeks at "ruptime -a". [21:59:30] If I were to follow the WMF labs model, I'd likely have 100-200 projects by now [21:59:50] Maybe. [21:59:55] Would that be a bad thing? [22:00:11] Thirty-five users on hemlock... is that a user host now? [22:00:22] I always use willow. [22:00:29] Have you seen the code of our fellow Toolserver users? [22:00:41] It's usually pretty bad. [22:01:02] Most Toolserver users don't seem to understand XSS or other Web gotchas. [22:01:19] Susan: i have a screen running on hemlock [22:01:27] and many shells sometime [22:01:29] I liked out tim1357 defined a bunch of lambda instead of actual functions [22:01:31] Ah. [22:01:32] no user server [22:01:35] liked how* [22:01:47] What happened to him? [22:01:51] I think he went AWOL. [22:02:03] Susan: that's because most of users aren't real programers, mostly just some fundaments on some language [22:02:10] Susan: He came back recently [22:02:16] Alchimista: Right. [22:02:20] legoktm: Did you he restart his bots? [22:02:29] I think there are orphaned database reports because of him... [22:02:30] I rewrote his tools without XSS exploits, and now AWOL [22:02:44] real programmers are hard to find anyway as real admins are ;) [22:02:44] Alchimista: And that's fine. It's a dev environment, after all. [22:02:48] Heh. [22:03:03] The illusive good code... [22:03:07] Susan: I don't think he restarted his bots, and I moved one of his webtools into my account since his expired [22:03:11] Written by real programmers.™ [22:03:23] Susan: yes, mostly because that's how big ideas can see the world, and then some expert inprove it [22:03:26] Right, account expiries and volunteer burnout are another issue for the Toolserver. [22:03:45] That also speaks to stability. [22:03:51] So there's host stability and then user/tool stability. [22:03:53] Fun times. [22:04:02] but the labs problem is that non experts will get riden of that *experimental* tool development [22:04:06] Susan: all communities I know ( i know 2 more) are working the same [22:04:39] there are few active people getting burned out after a while and new ones come and then a lot of "hangarounds" [22:04:39] nosy: Not surprising. These are common themes in FOSS communities, I think. [22:04:45] Right. [22:04:49] How long would an MMP take to fix an XSS exploit? I'm guessing never since I'm the only one touching the GeoHack source code. [22:04:50] There are social laws about this as well. [22:05:05] https://en.wikipedia.org/wiki/Pareto_principle [22:05:09] Etc. [22:05:27] Dispenser: Depends who's part of the MMP, what the XSS exploit is, etc. [22:06:14] It was a bit complicated to avoid double escaping shit [22:06:54] Susan: i am not very sure it is different in the professional industry world [22:08:38] Heh. [22:08:49] will reboot ortelius [22:13:41] ok ortelius is back [22:55:08] MySQL on z-dat-s4-a is CRITICAL: Cant connect to MySQL server on z-dat-s4-a (146) [23:44:34] nacht ts [23:55:09] MySQL on z-dat-s4-a is CRITICAL: Cant connect to MySQL server on z-dat-s4-a (146)