[02:05:23] Can someone give me +bureaucrat on the new prototype wiki? [02:05:25] Please. [02:05:47] Gah, no SUL? [02:34:31] which new prototype? [02:35:03] you realize this is on a completely different cluster, right? [04:06:39] Ryan_Lane: flaggedrevs [04:06:58] Ryan_Lane: If you want to encourage testers, I don't think they'll care where it's located. They'll just want it to be simple. [04:21:04] it isn't technically possible to include them [04:35:53] Ashlee: it isn't technically possible to include them [04:36:08] I saw your comment earlier. [04:36:17] I don't believe in technical impossibilities. [04:36:25] lol [04:36:37] Anything is possible with enough time and resources. [04:36:51] we should add it as a feature for the future Labs platform: integration with SUL [04:37:01] I heard the labs platform was dying. [04:37:31] it is running on a network that is isolated from the cluster. it was done this way on purpose [04:37:32] I said the future one, i.e. the one that's supposed to replace scattered prototypes at some point. [04:37:47] and I'm the one working on that [04:37:55] and guess what, it's network isolated too [04:38:10] That makes it impossible for it to use SUL? [04:38:19] we can do SUL for all of the prototypes, but not with the production cluster [04:38:26] yes [04:38:39] I think you're using "impossible" too loosely. [04:38:43] the current SUL system requires access to the databases [04:38:56] if we replace the SUL system, it's doable [04:39:06] It might be impractical or annoying, but I think you have access to the SUL databases. [04:39:10] which, is also something I'd like to do at some point in time [04:39:10] So I don't think it's impossible. [04:39:20] Ashlee: no. we do *not* have access to them [04:39:25] on purpose [04:39:29] for very good reason [04:39:35] right, Ryan_Lane , Gerard said you'd do OpenID this year! [04:39:39] lol [04:39:44] don't start with that ;) [04:39:52] Gerard said scientists don't believe bumblebees can fly. [04:40:00] I'm still trying to figure out what the hell he was talking about. [04:40:08] Gerard says a lot of stupid things [04:40:12] but I'm not supposed to say that [04:40:14] Ryan_Lane: Again, I don't think you're speaking of terms of possibility. [04:40:28] If you've intentionally locked yourself out of the SUL database, that's a different matter. [04:40:33] the wingspan of a bumblebee is such that for a long time scientists didn't understand how they could fly [04:40:40] you've never heard that joke? [04:40:46] That's a joke? [04:40:57] there's a joke that goes along with the bumblebee thing [04:41:33] Well, we'll leave it to Gerard to keep the laughs rolling. [04:41:48] paraphrased: Scientists say based on the wingspan of the bumblebee, it should impossible for it to fly, yet they keep on flying anyway, despite the scientists [04:41:58] Or not. [04:42:14] anyway, we purposely have the prototypes network isolated [04:42:25] because it allows us to do something else [04:42:30] What's that? [04:42:39] the ability to give root to anyone we want [04:42:50] including random volunteers [04:42:55] or even just shell access [04:43:06] I'm not sure I understand the point. [04:43:07] we are *way* more liberal with access on those systems [04:43:30] If you've isolated these wikis from the cluster, what's the advantage of Wikimedia hosting them? [04:44:01] it is cheaper [04:44:06] Than? [04:44:08] and we have more control [04:44:18] then using ec2 or rackspace cloud, or other things [04:44:32] and our ops folks manage them (well, I do anyway) [04:44:43] It seems that if you have developers doing development work on a wiki completely isolated from the Wikimedia cluster, they could just use localhost. [04:44:53] Where they could have all the root they wanted. [04:45:25] I guess that would make showcases harder. [04:45:28] then how would we prototype stuff? [04:45:34] yeah [04:45:42] these are also used for working in teams [04:45:54] Who's your boss? [04:46:00] Danese, who else? [04:46:10] Well, actually I guess it's Mark now [04:46:17] Mark Bergsma [04:46:21] Okay. [04:46:35] There are a lot of "EPM"s or whatever. I don't think it was an unreasonable question. [04:46:49] Well, there is only one ops EPM, which is now mark [04:47:40] anyway, the prototype is only one of the vms we have for test/dev, and it is important for flexibility reasons not to include it in the production cluster [04:47:47] so we can't do SUL, in its current state [04:48:33] The problem isn't that it is network isolated, but that SUL is limited [04:48:58] I like the qualifier. [04:49:01] That's a bit better. :-) [04:49:18] I was more curious about who was pushing forward the idea for more prototypes. [04:49:26] Presumably you're setting them up on someone's orders. [04:49:29] Is that Danese? [04:49:49] well.... [04:50:09] so the original prototype was for the usability initiative [04:50:13] Right. [04:50:18] it was on a linode server [04:50:38] I was hired to do ops for usability, and one of my tasks was to create a selenium cluster [04:51:09] I decided the right solution for it was virtualization, since we needed a bunch of low powered systems [04:51:30] so we did a hardware order, with a piece of hardware that had enough resources to scale [04:51:49] the linode server was becoming increasingly slower and slower, so we needed to move it [04:52:00] since we had a ton of capacity on the virt system, I moved it there [04:52:22] once people figured out we could make vms, it took off. now we have 25 VMs on that system [04:52:29] so it's now the tesla cluster... on one system [04:52:30] :D [04:52:37] I'm replacing it with something more cluster like [04:52:42] 25 VMs? [04:52:50] There's that much development? [04:52:50] yep [04:53:06] 7 of those are for selenium [04:53:37] these are the VMs on it: http://ryandlane.com/wiki/Category:Servers [04:54:05] and, yes I have that documented in wikitech too, but it's easier to do the documentation on my wiki first [04:54:16] Interesting. [04:55:12] ds1, storage1, and storage2 are shared infrasture i'm building to make a real cluster with [04:55:21] so really only like 15 dev vms [04:55:32] http://ryandlane.com/wiki/Server:Mobile.tesla.usability.wikimedia.org heh, support.microsoft.com? [04:56:04] http://mobile.tesla.usability.wikimedia.org/ doesn't seem to load. [04:56:45] There's not a concern about using the wikimedia.org domain? [04:57:35] no, why would there be? [04:57:51] mobile is for testing access to mobile from windows [04:58:02] it isn't running a web server, it is a windows vm [04:58:04] I'd be worried about *.wikimedia.org cookies or something. [04:58:06] Ah. [04:58:33] I don't think the cookies are a concern [05:00:12] the plan is to turn the tesla cluster into a functional replica of the production cluster, but with test and development in mind [05:00:23] also for testing of ops volunteers [05:00:33] Hmm, cool. [05:00:55] also for infrastructure testing [05:02:07] Infrastructure testing? [05:02:30] yeah. like testing automated failover [05:02:39] between continents [05:02:42] Hmm. [05:02:45] or datacenters [07:44:26] http://blog.chromium.org/2010/09/webp-new-image-format-for-web.html interesting [16:58:13] man, i hate my car. [16:59:36] What kind of car is it? [16:59:43] *guillom summons siebrand. [17:00:16] it's a 2006 honda civic si. [17:00:34] I drive a 2000 Honda Civic. It's such a little bitch car. [17:00:39] I want an SUV again. :-( [17:00:56] and like all sports-type vehicles, it hits finicky bits that requires bizzare maintenance. [17:01:15] <^demon|away> I like my car. [17:01:42] *Dispenser lols at Ashlee SUV comment [17:03:25] I had a VW Golf, and I'm glad I got rid of it. A car is so much trouble. [17:54:09] TrevorParscal: Does Eclipse have some spell checker for comments? [17:54:19] he he - not that I know of [17:54:21] I will look [17:55:02] I recently discovered my editor has one [17:55:15] Turned it back off because it doesn't recognize hacker jargon very well [17:56:29] TrevorParscal: http://www.mediawiki.org/wiki/Special:Code/MediaWiki/73170 read my mind; did you do this for the Vector ext. too? [17:56:55] Never mind, it's in the very next rev *facepalm* [17:59:47] RoanKattouw: you are right, that I shouldn't have moved/changed in one rev - I'm trying to get better at this [17:59:52] but in http://www.mediawiki.org/wiki/Special:Code/MediaWiki/73037#c9738 [18:00:15] you are wrong, it's not necessary to prepend window. to os_* [18:00:30] you only have to do that when you are decalring something [18:00:38] you can still access window stuff in the global namespace [18:00:44] Oh of course [18:01:46] Please refute that on CR then [18:01:55] Oh you already did [18:15:19] TrevorParscal, do you mind if I go ahead and rename Vector-the-extension? [18:15:49] to what? [18:16:09] I say we first discuss whether to move it to core (I think we should) [18:16:12] Before renaming it /again/ [18:16:36] I was thinking VectorPlus, but if you prefer another name, it's also good [18:16:46] omg, please no [18:16:48] RoanKattouw, r73030 seems stalled [18:16:58] *TrevorParscal recalls Windows 98 Plus! [18:17:03] ouch [18:17:20] and from the ml I think it will end up as an extension included in the install [18:17:25] Platonides: Stalled? [18:17:43] Oh the discussion there [18:18:11] ah, you thoght the revision itself xD [18:18:25] my bad, I should have been clearer [18:18:59] TrevorParscal: Re r73096: so what about the loading JUI on demand thing? [18:19:25] RoanKattouw: we could do that [18:20:05] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/73150#c9749 [18:20:11] your last comment [18:20:57] The toolbar vs. dialogs loading order thing [18:21:19] *Platonides puts homework for the weekend to Trevor: find out a good name for Vector-the-extension [18:21:33] RoanKattouw: there's no race condition here [18:21:56] *TrevorParscal is find with it as it is, but will give it some thoguht [18:23:23] How is there no race condition? [18:28:07] hmm [18:28:29] ok, I'm looking into it some more - I had figure it out, I thought, but now I'm not so sure [18:29:22] ok, so there is an inclusion order issue, but it didn't crop up until I added the ext. to things [18:29:25] ha ha... [18:29:27] working on fix [18:30:17] I implemented that originally, and I think I put it in the dialogs module at first as well [18:30:31] But then realized fairly quickly that that was never gonna fly and put it in toolbar, forgetting to explain why [18:30:44] dialogs requires toolbar, so toolbar goes first [18:31:07] I just need to make dialogs attach it's stuff to an event that gets executed when toolbar is added [19:39:00] hey can one of y'all run a quick batch delete on cswiktionary for danny_b? [19:39:14] thanks! :D [19:39:26] Ah, sure [19:39:33] What does he need deleted? [19:39:42] ah lemme get him to poke you [19:39:43] RoanKattouw: will describe on pm [19:39:49] \o/ [19:46:47] <^demon|away> Ok, I think that's quite enough code reviewing for today. [19:46:50] <^demon|away> I'm going nuts. [19:48:52] brion: For mass deletions in the past, did you use deleteBatch.php ? Does it do bot flags or do I have to do that in the DB myself? [19:49:45] ^demon|away: i'll help catch up on the weekend :D [19:50:04] RoanKattouw: deleteBatch yeah.... i think it has an auto-bot flag, or you can maybe set the user [19:50:17] You can set the user [19:50:25] <^demon|away> brion: Whee! I spent most of today reviewing changes to the installer and code review :) [19:50:26] But I'm not confident it'll inherit the bot flag from that [19:50:29] Or wait [19:50:31] Actually, yes, it will [20:30:52] ^demon|away: any updates about hack-a-ton? [20:32:17] I was just looking at that page. [20:32:21] Venue is still TBA? [20:32:37] <^demon|away> I have more information. Should be able to update mw.org later this evening. [20:33:26] ^demon|away: particularly about the venue... do we have anything finalized or do you need me to keep bugging people [20:34:05] <^demon|away> /msg :)