[00:17:51] Leslie is cheating! [00:18:04] hehe [00:18:17] * petan wants to review his own changes [00:18:25] :) [00:30:26] i think the downside of reviewing your own changes is that you're on the hook at 3am when the site breaks to fix it :) [01:07:02] RECOVERY Disk Space is now: OK on nova-production1 nova-production1 output: DISK OK [01:20:02] found the credentials missing problem [01:20:22] as usual a bug in mediawiki :) [01:24:40] I'm now also adding memcache support to ldap auth, since I made a change recently with really shitty performance impacts [10:30:12] PROBLEM Disk Space is now: CRITICAL on mobile-feeds mobile-feeds output: DISK CRITICAL - free space: / 0 MB (0% inode=54%): [13:07:14] New patchset: Dzahn; "remove Service[apache2] - duplicate definition, already in generic class" [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1702 [13:07:45] New review: Dzahn; "(no comment)" [operations/puppet] (test); V: 1 C: 2; - https://gerrit.wikimedia.org/r/1702 [13:07:46] Change merged: Dzahn; [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1702 [13:20:06] New patchset: Dzahn; "load Apache rewrite module" [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1703 [13:20:20] New review: gerrit2; "Lint check passed." [operations/puppet] (test); V: 1 - https://gerrit.wikimedia.org/r/1703 [13:20:37] New review: Dzahn; "(no comment)" [operations/puppet] (test); V: 1 C: 2; - https://gerrit.wikimedia.org/r/1703 [13:20:38] Change merged: Dzahn; [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1703 [13:29:42] New patchset: Dzahn; "disable apache_site 000-default" [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1704 [13:29:55] New review: gerrit2; "Lint check passed." [operations/puppet] (test); V: 1 - https://gerrit.wikimedia.org/r/1704 [13:30:12] New review: Dzahn; "(no comment)" [operations/puppet] (test); V: 1 C: 2; - https://gerrit.wikimedia.org/r/1704 [13:30:12] Change merged: Dzahn; [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1704 [13:45:09] New patchset: Dzahn; "add a default Debian/Ubuntu Apache ports.conf, but without NameVirtualHost *:80" [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1705 [13:45:22] New review: gerrit2; "Lint check passed." [operations/puppet] (test); V: 1 - https://gerrit.wikimedia.org/r/1705 [13:49:46] New review: Dzahn; "otherwise if you use apache_confd { namevirtualhost: and the default ports.conf you get Apache warni..." [operations/puppet] (test); V: 1 C: 2; - https://gerrit.wikimedia.org/r/1705 [13:50:27] New review: Dzahn; "(no comment)" [operations/puppet] (test); V: 1 C: 2; - https://gerrit.wikimedia.org/r/1705 [13:50:27] Change merged: Dzahn; [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1705 [13:51:30] New patchset: Dzahn; "now use that ports.conf on wikistats instance" [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1706 [13:51:44] New review: gerrit2; "Lint check passed." [operations/puppet] (test); V: 1 - https://gerrit.wikimedia.org/r/1706 [13:52:25] New review: Dzahn; "(no comment)" [operations/puppet] (test); V: 1 C: 2; - https://gerrit.wikimedia.org/r/1706 [13:52:26] Change merged: Dzahn; [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1706 [13:56:34] New patchset: Dzahn; "fix file path / syntax error, this is not a template" [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1707 [13:56:47] New review: gerrit2; "Lint check passed." [operations/puppet] (test); V: 1 - https://gerrit.wikimedia.org/r/1707 [13:56:55] New review: Dzahn; "(no comment)" [operations/puppet] (test); V: 1 C: 2; - https://gerrit.wikimedia.org/r/1707 [13:56:56] Change merged: Dzahn; [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1707 [13:57:53] oh, hi, somebody else working on labs right now? [13:58:21] duplicate definition related to ssh.pp all of a sudden.. hmm [13:59:02] and problem disappeared again..ok [14:00:31] New review: Dzahn; "yep, that gave me an Apache reloading without any warnings now:)" [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1706 [14:31:21] New patchset: Dzahn; "add an initial index.php for wikistats, require dir, indent one more level,.." [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1708 [14:31:34] New review: gerrit2; "Lint check passed." [operations/puppet] (test); V: 1 - https://gerrit.wikimedia.org/r/1708 [14:32:12] New review: Dzahn; "(no comment)" [operations/puppet] (test); V: 1 C: 2; - https://gerrit.wikimedia.org/r/1708 [14:32:13] Change merged: Dzahn; [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1708 [14:48:15] New patchset: Dzahn; "style fixes per guide - single quotes, represent file modes as 4-digits and quote them, ensure attribute always first.." [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1709 [14:48:25] New review: gerrit2; "Change did not pass lint check. You will need to send an amended patchset for this (see: https://lab..." [operations/puppet] (test); V: -1 - https://gerrit.wikimedia.org/r/1709 [14:50:34] hmm, if my instance is outta storage space, can there be something done? [14:53:52] hi Max, greetings to .ru, afraid i dont know the answer to your question (yet) though [14:54:26] New patchset: Dzahn; "style fixes per guide - single quotes, represent file modes as 4-digits and quote them, ensure attribute always first.." [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1709 [14:54:38] New review: gerrit2; "Lint check passed." [operations/puppet] (test); V: 1 - https://gerrit.wikimedia.org/r/1709 [14:55:59] New review: Dzahn; "(no comment)" [operations/puppet] (test); V: 1 C: 2; - https://gerrit.wikimedia.org/r/1709 [14:55:59] Change merged: Dzahn; [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1709 [14:56:35] * MaxSem waves to mutante [14:59:29] New patchset: Dzahn; "that way the variable aren't expanded in the template anymore" [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1710 [15:00:17] New patchset: Dzahn; "that way the variable aren't expanded in the template anymore" [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1710 [15:00:29] New review: gerrit2; "Lint check passed." [operations/puppet] (test); V: 1 - https://gerrit.wikimedia.org/r/1710 [15:00:45] New review: Dzahn; "(no comment)" [operations/puppet] (test); V: 1 C: 2; - https://gerrit.wikimedia.org/r/1710 [15:00:45] Change merged: Dzahn; [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1710 [15:04:25] MaxSem: you could always create a new instance with more storage and apply your puppet class to it. expanding an existing one might be harder [15:05:01] ...if I used puppet [15:05:13] heh..ok:) [15:05:39] well, I tried all those shiny checkboxes yesterday [15:06:00] ...then Ryan had to add a warning not to check them [15:06:17] make a new instance, and rsync stuff from the old to the news? [15:06:31] s/news/new one [15:06:51] MaxSem: can I help you [15:07:27] I recommend you to create a nfs storage on another instance [15:07:34] does rsync reqire at least 1 byte of temp storage on the old instance? :P [15:07:40] Ryan wanted to install gluster later that would solve it too [15:08:05] depends [15:08:09] maxsem@bastion1:~$ ssh mobile-feeds.pmtpa.wmflabs [15:08:09] ssh: connect to host mobile-feeds.pmtpa.wmflabs port 22: No route to host [15:08:12] MaxSem: you can probably get a couple bytes back with "apt-get clean" [15:08:29] MaxSem: you should copy some stuff to /mnt [15:08:33] that is your storage [15:08:34] meh, too late [15:08:42] ./mnt has 20 gb / 10 gb [15:08:55] you should be using /mnt for storage [15:09:12] though I can't connect to my new shiny 20 gigs instance [15:09:19] when you pick size of vd it's mounted to /mnt [15:09:34] check console output it usually take a 5 mins to finish it [15:14:53] PROBLEM host: mobile-feeds is DOWN address: mobile-feeds CRITICAL - Host Unreachable (mobile-feeds) [15:18:27] gee [15:18:31] ? [15:18:35] ^^ [15:31:03] RECOVERY host: mobile-feeds is UP address: mobile-feeds PING OK - Packet loss = 0%, RTA = 9.94 ms [15:48:48] New patchset: Dzahn; "add MariaDB for labs" [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1711 [15:49:00] New review: gerrit2; "Lint check passed." [operations/puppet] (test); V: 1 - https://gerrit.wikimedia.org/r/1711 [15:49:46] New patchset: Dzahn; "add MariaDB for labs" [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1711 [15:49:58] New review: gerrit2; "Lint check passed." [operations/puppet] (test); V: 1 - https://gerrit.wikimedia.org/r/1711 [15:54:17] umm, does Oracle oppress MySQL THAT much? :P [15:55:30] guess so :P i saw somebody asking for mysql in labs and Ryan said "yes, if we can use MariaDB" [15:56:03] i'd like a db on localhost first, for testing [15:58:15] oh,, problem with gerrit now or just me [15:59:58] not just you [16:08:59] MaxSem: rebooting it, there was no output on mgmt console [16:09:08] hold on [16:09:19] and svn.wm.o too [16:12:03] MaxSem: try now [16:12:48] yep,. works [16:13:32] we have mysql on bots [16:13:34] no problem [16:14:21] why MariaDB not Drizzle, btw? [16:14:34] dunno Ryan was like Maria db is cool [16:14:43] we got maria on bots-sql3 [16:14:59] it works like mysql but it's slightly better though [16:15:11] it can use ldap for auth [16:15:15] etc [16:15:33] it's really easy to upgrade from mysql to maria, "drop-in replacement" [16:15:47] I don't know if it's so easy to put it back [16:16:03] maria has worse support now [16:16:11] it isn't even in ubuntu repo [16:16:42] it took me few hours to get it working because of some conflicts with original repo [16:16:57] we can use apt_key in puppet to nicely import the key, btw [16:17:18] petan: feel like reviewing ?;) [16:17:21] maria repo is overwriting definition of some packages like mysql, so when you type apt-get mysql-client it install also maria libs [16:17:44] reviewing of what? [16:17:48] :o [16:17:52] !change 1711 | petan [16:17:52] petan: https://gerrit.wikimedia.org/r/1711 [16:18:10] ah [16:19:33] cosmetical: 21 is missing ; at the end of line [16:19:42] it isn't necessary but you have it on other lines :P [16:20:02] thx [16:20:13] also need to rename the file [16:20:41] yes [16:20:44] filename is wrong [16:20:53] missing .list [16:21:00] New patchset: Dzahn; "add MariaDB for labs" [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1711 [16:21:19] arr, that too:) just did 53 > 5.3 [16:22:07] New patchset: Dzahn; "add MariaDB for labs" [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1711 [16:22:21] New review: gerrit2; "Lint check passed." [operations/puppet] (test); V: 1 - https://gerrit.wikimedia.org/r/1711 [16:22:30] duh:) one more [16:23:21] New patchset: Dzahn; "add MariaDB for labs" [operations/puppet] (test) - https://gerrit.wikimedia.org/r/1711 [16:23:34] New review: Petrb; "Cosmetic, line 21 of mariadb.pp is missing semicolon at the end of line, otherwise should be ok :)" [operations/puppet] (test) C: 1; - https://gerrit.wikimedia.org/r/1711 [16:24:15] gerrit is ugly [16:24:34] New review: Petrb; "(no comment)" [operations/puppet] (test) C: 1; - https://gerrit.wikimedia.org/r/1711 [16:26:14] I will try it on sql3 once it's merged and in nova [16:28:05] petan, once Roan spent 2 days trying just to set up working environment in order to maybe fix some of Gerrit's uglinesses [16:28:10] he failed [16:28:17] :) [16:29:20] that bot should be registered heh [16:29:41] hi Sumana :) [16:29:52] hi petan. Happy holidays. [16:30:00] thanks! [16:30:03] you too [16:30:23] sumanah: heya! [16:30:28] sumanah: happy holidays to you too :) [16:30:30] petan: did you know we now have a volunteer named Oren who is interested in improving search on Wikimedia sites? [16:30:42] and season's greetings back at ya, yuvipanda :) [16:30:46] that's cool [16:30:52] sumanah: :) [16:30:58] petan: we use Lucene, which is in Java [16:31:06] I didn't know that [16:31:06] aha [16:31:14] I don't know much about java [16:31:27] yuvipanda: I just overheard another name for the sort of Dec 24-Jan 2 period [16:31:29] "winterval" [16:31:35] lawl [16:31:43] i'm stealing that [16:31:46] maybe you noticed the recent discussion on wikitech where people were blaming me for using c# :P [16:31:48] yuvipanda: totally! [16:31:55] i'm mostly working through that, except for a few days, I guess [16:31:58] not sure [16:32:05] petan: yeah, I saw. [16:32:15] petan, glad everything got sorted out. [16:32:20] so I am not really java coder :) [16:32:29] and this place is *cold* [16:32:34] but I don't have a problem with java apart of that [16:32:40] petan: no, but you are a Labs expert, and I think Oren could use a Labs instance [16:32:49] I would be happy to help him [16:33:02] java works ok on labs ^^ [16:33:02] petan: https://www.mediawiki.org/wiki/User:OrenBochman [16:33:08] yuvipanda: where are ya? [16:33:15] sumanah: bangalore [16:33:26] sumanah: it's around 17C-20C and I can't stand *this* [16:34:18] yuvipanda: could be worse, but, my sympathies nonetheless. Perhaps you can redirect the hot air from your laptop to blow across your metatarsals [16:34:40] sumanah: but my macbook air is incredibly cool [16:34:46] noooooo [16:34:47] the aluminuma ctually is colder than my feet [16:34:54] *aluminum [16:35:11] have you considered that you might have accidentally wandered into the realm of a Maxwell's Demon? [16:35:43] sumanah: possibly [16:35:53] a laptop that isn't hot [16:35:56] a wonder indeed [16:37:50] yuvipanda: http://www.jonathancoulton.com/wiki/A_Laptop_Like_You_%28Demo%29 [16:38:06] yuvipanda: you can listen to it on http://www.jonathancoulton.com/store/downloads/ [16:38:18] under "other experiments" [16:38:38] :D [16:38:52] sumanah: I initally read that as MaxSem's Demon [16:38:52] :D [16:39:28] haha [16:40:05] :D [16:40:32] sumanah: any idea on the KGISL people? [16:40:53] yuvipanda: I wrote back to them a few days ago, no response since [16:41:00] ah, okay [16:41:20] yuvipanda: I wrote to them on the 20th [16:41:36] sumanah: hmm, let me see what's up there. [16:42:01] yuvipanda: I cc'd you, "Re: MediaWiki" in case that helps. [16:42:25] yeah, I remember that [16:42:32] assumed they had hit reply instead of reply-to-all [16:42:35] yuvipanda: I would very much like to actually take next week off, but if it makes the difference between partnering effectively vs not, I can do a phone call next week [16:43:26] sumanah: I doubt that would be a problem, so do take the week off. Am sure I can handle that, or it can wait :) [16:43:40] heh my commited code in c# was reviewed in 10 minutes my code in c++ isn't reviewed till now [16:43:51] and they say people do not use c# in wikimedia svn [16:44:02] thanks yuvipanda, I very much appreciate it. And my spouse will, too. [16:44:37] sumanah: transfer him a o/ for BeautifulSoup :) [16:44:48] yuvipanda: :D [16:44:51] one of the things that hooked me on to Python ~4 years ago [16:44:51] :) [16:44:52] so [16:44:52] petan: heh! [16:45:02] That's wonderful to hear! [16:45:08] yuvipanda: ^ [16:45:14] petan: oooh, you're the one running the new bot! [16:45:16] sumanah: :) [16:45:21] yes [16:45:28] !wm-bot [16:45:28] http://meta.wikimedia.org/wiki/WM-Bot [16:46:08] actually the bot is running on labs so I don't know if I am person who "run it" :) [16:46:11] petan: what channels is it logging? [16:46:21] that's not so easy to tell [16:46:24] @channellist [16:46:24] I am now in following channels: #huggle, #wikimedia-dev, #mediawiki-move, #wikimedia-tech, #wm-bot, #wikimedia-labs, #wikimedia-operations, ##matthewrbowker, ##matthewrbot, #wikipedia-zh-help, #wikimedia-toolserver, ##Alpha_Quadrant, [16:46:24] yuvipanda: I am aiming to be more responsive in general but (especially when I travel) I build up a backlog of people to respond to, requests to fulfill, and so on. I'm so grateful any time I can delegate some to a community member, like petan or varnent or logic or ashley or johnduhart or bawolff. [16:47:01] yuvipanda: most of wikimedia channels are logged [16:47:02] petan: can you log #wikimedia-mobile as well? [16:47:08] we've no bots there, afaik [16:47:20] @add #wikimedia-mobile [16:48:17] petan: thanks [16:48:20] petan: it logs? [16:49:52] yes [16:50:03] http://bots.wmflabs.org/~petrb/logs/%23wikimedia-mobile/ [16:50:38] it's quite usefull because I can watch what is happening in this channel on my phone without irc client :) [16:50:47] I want to create some php interface too [16:50:48] for logs [16:57:18] petan: point me to the C# source for this? [16:57:27] petan: I used to be a 100% C# person before my python migration [16:57:32] still love the language [16:57:32] :) [16:57:37] https://svn.wikimedia.org/viewvc/mediawiki/trunk/tools/wmib/ [17:05:20] hi Krinkle did you find any quick issues which I could fix in my extensions? [17:05:26] * extension [17:05:37] online status ? [17:05:43] because thanks to Brion ajax is already there [17:05:44] yes [17:05:53] I haven't had the time to look at it again. [17:06:03] ah... [17:06:21] ok if you found any problem let me know [17:35:19] c# reminds me of java a lot, which makes it quite easy to write. [18:08:24] Hmm should I document cluebot or play with openstack tonight, the later sounds far more fun. [18:08:41] hehe, documenting is always good, though nobody wants to do it [18:09:20] The code is documentation its self, it will just take you 6months and a few thousand braincells to understand what it does. [18:09:34] If in doubt `history` [18:09:42] hahaha [18:10:08] what if you already have killed those brain cells with alcohol ? [18:11:12] mmm alcohol [18:11:55] I have a litre of gin, half a litre of whiskey and a random bottle of half wine half whiskey sad on my desk. [18:12:17] heh [18:12:34] docs are always nice :) [18:12:50] if you're interested in learning openstack, you can work with me on it in the openstack labs project [18:13:13] I set up an all-in-one instance for openstack, but I'd also like a multi-node install version [18:13:25] hi Ryan_Lane [18:13:30] I have an idea :) [18:13:31] petan: howdy [18:13:33] oh? [18:13:52] I wanted to make a small irc server for bots similar to network we already have, but the applications could send messages there [18:14:00] so that cluebot and huggle could communicate using irc [18:14:16] cluebot does a lot of irc spam, some tools allready sit and parse it's feed [18:14:25] I asked if I could use pmtpa irc but i can't [18:14:29] why communicate via irc? [18:14:33] omg tag-teaming [18:14:35] so I decided we could create an instance on labs [18:14:41] irc server [18:14:49] probably better comminication methods around [18:14:50] because irc is simple :) [18:15:15] most of the tools do not need to use encrypted methods, irc servers are already written and easy [18:15:34] wikipedia bots and tools usually use irc for feeds because it's simple way to do that [18:15:43] * Ryan_Lane nods [18:15:43] what other ways do you know? we could use [18:15:49] a queue [18:15:50] I have no idea [18:15:55] other than making own server application [18:16:06] it's reliable and the messages wouldn't get lost [18:16:17] queue? [18:16:20] huh [18:16:26] how does that work [18:16:31] bots can be stopped and re-started and still pick up messages from the queue [18:16:39] ah [18:16:48] right but how do I implement that? [18:16:50] queues are fairly simply. you say "put x into queue y" [18:17:01] do I need some software on server? [18:17:10] then the other bot says "pick up items from queue y" [18:17:13] openstack looks ideal for a dev/prod env, though our production env is a bit bespoke with iscsi mounts + usual storage and a lot of vlans + static ip assignments (with ebtable restrictions in place) so getting my head around openstack with a view to deployment is going to take a while :( [18:17:13] ok but these queues are shared? [18:17:17] yes [18:17:44] so I can create like a queue for each wiki, and send data there? [18:17:51] ok, is there some tutorial how to do that? [18:18:04] we would need instance even for that, btw :) [18:18:22] yeah [18:18:30] I'd say do the irc server for now, I guess [18:18:42] but think about queues in the future [18:19:11] Queues would be fun [18:19:15] hm... ok [18:19:28] For bots like cluebotng it would mean it wouldn't miss edits when it was down [18:19:34] queues are great for simple message passing :) [18:19:40] right I created an instance but I wonder how do I redirect 5556 from public ip of bots to that? [18:19:43] Ryan_Lane: ^ [18:19:45] Which I suppose it could currently parse the edit rss thing but meh that's cruddy [18:19:50] oh sorry [18:19:51] 6667 [18:19:54] petan: what do you mean? [18:20:04] I need to redirect traffic to that port to irc instance [18:20:08] ah [18:20:20] * Damianz throws petan a haproxy box [18:20:21] the bots are internal, why the need for a public ip? [18:20:38] I wanted to use it for huggle too [18:20:51] ah [18:21:03] we had an idea to integrate communication between tools so when someone revert vandalism, other people who use it know that [18:21:11] and do not check it again [18:21:45] It sounds like what you're trying to do is create a standard which would be awesome but would need more than an irc server. [18:21:46] hm. it'll need another public IP for this [18:21:49] also we wanted to integrate huggle and cluebot so that it could communicate between, but I would need to sort that out with Damianz [18:22:01] ah, really? [18:22:05] We where looking at integrating with huggle a long time ago [18:22:07] it definitely sounds like irc is the wrong tool for this, though :) [18:22:10] We integrate with something else instead [18:22:13] heh [18:22:31] you want reliable message passing, and irc definitely isn't that [18:22:39] Stiki have a bot in our irc feed that contains pretty much the change feed + cbng info. [18:22:51] you know.... [18:22:52] They pipe that into their review queues along with other sources. [18:22:52] aha, and your irc feed is where? [18:23:00] #cluebotng-spam on irc.cluenet.org [18:23:01] there's an openstack project that might actually be good for this [18:23:10] there is wikipedia-van too that just has reverted stuf [18:23:13] what the hell was the name of it again? [18:23:20] Damianz: it would be cool to have this on one place [18:23:28] Indeed. [18:23:29] Damianz: using some standards so that all tools could communicate [18:23:44] But I think irc will cause issues for this purpose [18:23:53] https://launchpad.net/burrow [18:24:06] Would be better to have a central list where each tool spools a local list if the central server is down. [18:24:10] Ryan_Lane: that was actually my idea, to set up communication whatever we could use for bots and tools (I wanted to make irc, but if queus are better, no problem) [18:24:16] seems it isn't a full openstack project yet [18:24:20] That would make for some awesome analytics too [18:24:41] Damianz: maybe even that [18:25:01] burrow is a queue that works via http [18:25:16] I assume it's the same idea as RabbitMQ? [18:25:16] * johnduhart looks at burrow [18:25:18] so, it can be proxied [18:25:21] <3 RabbitMQ [18:25:22] Damianz: simila [18:25:24] *similar [18:25:44] clients long-poll with this [18:25:50] would be nice if it supported web sockets [18:25:50] Damianz: but my idea was: cluebot know that this 60% vandalism so it doesn't revert, instead of that it would report to channel where other tools would grab it / when you see diff is ok, you flag it so, and other tools remove it from queue [18:26:25] Damianz: so there is interactive feedback, when someone would revert something cluebot didn't know it could also insert it to own db [18:26:34] Well with cbng our idea was for Stiki that they pick up that chunk of vandalism there is just under the threashold for cbng but is VERY likely to be vandalism. [18:26:51] Damianz: that was also our idea :) [18:26:53] We wanted to integrate their GUI to pipe data into our review system too so it got back into the ANN [18:27:10] Sadly their db uses shared mysql details and is open to corruption, efforts to write a secure system never got anywhere [18:27:13] Damianz: I think we should make some standard then [18:27:30] Damianz: I think that huggle, stiki and cb could co operate [18:27:43] people are mad because cluebot is faster etc. this would prevent it [18:27:48] lol [18:28:07] THere is a lot for people to do when it's up [18:28:15] http://burrow.openstack.org/examples.html#multi-cast-event-notifications [18:28:16] THey just mostly see what it reverts though [18:28:20] :) [18:28:32] !queue is http://burrow.openstack.org/examples.html#multi-cast-event-notifications [18:28:32] Key was added! [18:28:54] messages are kept in queue, all workers can read message for the length of the ttl [18:29:13] If we could take data from everyone reviewing stuff and pipe it into the bot it would make the bot far better. The bot getting better doesn't mean less reviewers because the data changes constantly and we need people in the system but some people seem to think they are killing themselves by helping. [18:29:25] !queue del [18:29:25] Successfully removed queue [18:29:28] !queue is http://burrow.openstack.org/ [18:29:28] Key was added! [18:30:04] burrow looks useful for one of my projects actually :D [18:30:25] burrow will hopefully be fairly useful in general :) [18:30:26] Damianz: I think we should start email conference about it [18:30:32] I'm not sure how stable it is considered [18:30:33] with other devs of stiki etc. [18:30:34] Well currently it just polls every x seconds and was going to be replaced by websockets but that is more awesome. [18:30:54] yeah, long-polling [18:31:16] petan: What we should aim for is a RFC + that to become a standard accross wp stuff but that invovles a LOT of people and a lot of burocracy. [18:31:24] definitely [18:31:26] I hate bureaucracy. [18:31:35] well, start with a proof-of-concept [18:31:42] then work your way into the BS [18:31:49] heh [18:31:55] lol [18:32:09] Damianz: it's important if you want it to work for all tools [18:32:10] I'm happy to look at cbng, I want to re-write the bot engine tbh. [18:32:12] give examples of how it can be used, and make docs [18:32:27] It's was Cobi's baby but he hasn't touched it in a long time as he has really busy and it's kinda fallen into my lap [18:32:28] if it's good, people will adopt it, and it'll eventually be a de facto standard anyway [18:32:48] or they will argue that their idea is better and we should switch to it [18:33:09] and if your idea actually has code, they'll have to show a working solution as an alternative :) [18:33:10] I hate to rewrite working stuff just because someone has "better" idea, we should first set up a standard and then work on it [18:33:17] inertia is awesome [18:33:34] petan: Oh cb wise is because it falls over everytime wp has an edit fest :P [18:33:46] Can anyone remember that website that had like a few thousand reviewed edits on en.wp? [18:33:52] That was an awesome idea =/ [18:34:12] by reviewed do you mean something else than 'patrolled'? [18:34:14] btw I am busy now, Damianz we will talk about this again :) [18:34:16] yes [18:34:33]