[00:00:58] I [00:01:00] I'm in [00:01:39] Roan-meeting: Nope. No JS errors on load, and nothing when I click them. [00:01:58] I'll file it in BZ (if not already) [00:15:45] ok, so invalid mix of collations [00:16:38] Yeah [00:16:49] this usually happens when the character type of literal strings is different to the character set of the fields in question [00:17:03] So what I've got now locally is basically s/varchar(N) binary/varbinary(N)/g [00:17:09] you can easily change the character set of all literal strings, but our code to do that is a bit awkward [00:17:11] uhuh [00:17:26] I don't think you need to change any field types [00:17:39] varchar(N) binary is correct [00:17:42] OK [00:17:49] We just need to make sure the installer sets stuff up correctly? [00:17:53] yes [00:18:16] priyanka said the database fields were latin1, if that's the case then that's a different bug [00:18:27] (BTW can you 1.17-tag any revs you committed today that should be in the deployment+release? At least the collation stuff and the revert sounded like that) [00:18:33] Well not quite that [00:18:46] The fields inherit the table's collation which inherits the database's [00:18:49] unless overridden [00:19:07] And we don't have collation overrides anywhere in the schema I think, so once the DB is created as latin1 all fields become latin1 too [00:19:45] <^demon|away> latin1 sounds like $wgDBoptions or whatever it's called not being appended and it falling back to defaults. [00:20:15] we have collation overrides on all the tables [00:20:25] $wgDBoptions, like Chad says [00:20:29] Right [00:20:33] Well this was on page_title [00:20:46] And the CREATE TABLE page statement does have $wgDBoptions in 1.17 [00:20:57] *$wgDBTableOptions [00:21:27] Also: [00:21:29] $wgDBTableOptions = 'ENGINE=InnoDB'; [00:21:32] do you have the problematic database there? [00:21:53] you're saying it's wrong? [00:21:54] Not right now [00:22:02] But she can create a new one right now [00:22:10] (This happened on pdhanda 's box, or on prototype, I think) [00:22:26] I'm not sure it's wrong [00:22:38] I'm just saying that's what it says in DefaultSettings.php and it doesn't seem to force utf-8 or binary or whatever [00:23:12] the installer should do it [00:23:36] Oh look [00:23:40] LocalSettings (on my box): [00:23:42] $wgDBTableOptions = "ENGINE=InnoDB, DEFAULT CHARSET=binary"; [00:24:34] that's how it should work, yes [00:24:58] where's pdhanda? [00:25:48] 20 ft away from me [00:26:14] not on IRC? [00:26:24] i'm here [00:26:27] was distracted [00:26:43] give me one sec I'll give you the db [00:26:56] Also, would $wgDBTableOptions even be applied prior to LocalSettings creation? [00:27:29] if not, that's a bug [00:27:56] but someone probably would have noticed before now [00:29:19] public function getTableOptions() { [00:29:19] return array( 'engine' => $this->getVar( '_MysqlEngine' ), [00:29:20] 'default charset' => $this->getVar( '_MysqlCharset' ) ); [00:29:20] } [00:30:00] if it didn't apply the options, the engine wouldn't be selected correctly either [00:30:02] <^demon|away> Oh, well this does seem wrong. [00:30:13] <^demon|away> I just noticed one of mine is InnoDB, latin1 [00:30:23] InnoDB too [00:32:42] I tested this when I first wrote it, something must have broken along the way [00:33:24] test_install on prototype [00:33:42] RoanKattouw, TimStarling ^ [00:33:49] just created that using the latest installer [00:33:58] thanks pdhanda [00:34:04] np [00:34:21] ) ENGINE=InnoDB AUTO_INCREMENT=2 DEFAULT CHARSET=latin1 | [00:34:28] (from SHOW CREATE TABLE page; ) [00:34:44] also test_install | CREATE DATABASE `test_install` /*!40100 DEFAULT CHARACTER SET latin1 */ [00:34:45] Also: `page_title` varchar(255) character set latin1 collate latin1_bin NOT NULL, [00:38:24] hmm, maybe I didn't actually do this... [00:39:25] maybe in my version it just said "mockup" [00:40:33] <^demon|away> Yes it did :) [00:41:52] do you want me to fix it? [00:44:48] I'm fixing it [00:50:04] <^demon|away> Sorry, I'm distracted with some other stuff right now. [00:53:32] TimStarling: Can I bug you about LiquidThreads comment text storage? [00:53:44] yes [00:54:14] I want to move away from using wiki pages as the backend [00:54:43] why? [00:55:08] I've put my full reasoning here: http://office.wikimedia.org/wiki/LiquidThreads/Backend_updates#Use_of_wiki_pages_to_store_comment_text [00:56:00] but basically because it's fragile, subject to interference, annoying to users, subject to interference, and requires some niceties to be observed like giving each comment a title (which isn't necessarily helpful) [00:56:33] The original impetus was revision control, but that doesn't apply to the metadata, which also needs to be versioned. [00:56:35] ok [00:57:00] so what's your question? [00:57:01] So I'm wondering if the comment text should be moved to a database field, into a separate part of external storage, or somewhere else. [00:57:30] I'm aware that storing thousands of 1-5kB comments in the database isn't necessarily a great idea, because the table won't fit in RAM [00:58:11] well, that's my interpretation, and I'm asking you because I'm not 100% clear on the technical issues. [00:58:27] ...but delta-based compression also isn't gonna help for storing comments [00:58:30] it'll be the same as AbuseFilter [00:58:44] Yes, but AbuseFilter has different object sizes and different access patterns. [00:58:53] you should store a pointer in the text table, which gives you an old_id to refer to in other code [00:58:57] AbuseFilter variable dumps are more towards tens of kilobytes. [00:59:05] and are not accessed in batches [00:59:14] so does it still make sense to use ES for comment text? [00:59:26] yes, I think it makes sense [00:59:35] I believe you can do batch retrieval of objects in ES, yes? [01:00:08] there are optimisations which speed up the retrieval of objects in ES, since that's what the XML backups have to do [01:00:29] I don't think there's a batch interface as such [01:00:29] Can ES be dissuaded of using delta-based compression, or does it automatically notice when the diff is larger than the text itself, or.. ? [01:00:35] *from [01:00:40] if one was added, then it would help backup as well [01:00:59] RoanKattouw: delta compression is done retrospectively [01:01:00] RoanKattouw: compression is handled in Revision, not ES [01:01:07] Oh, right [01:01:23] ES just stores unstructured binary objects [01:02:57] mind you, it's probably better if all callers, including LQT, do use the appropriate functions to uncompress text rows [01:03:00] Where in Revision does this happen? Searches for diff, delta, compression don't turn up related code [01:03:38] Revision::getRevisionText(), it's static [01:03:42] TimStarling: those appropriate functions being? [01:03:49] (answers both) [01:03:54] right. [01:04:07] No, that's what does the decompression. What handles the delta stuff? [01:04:11] (Going in, also) [01:04:26] HistoryBlob.php [01:04:31] Aaah of course [01:04:58] Thanks TimStarling, I'll update my document accordingly [01:06:18] DiffHistoryBlob::isHappy() , nice naming :D [01:14:13] (BTW can you 1.17-tag any revs you committed today that should be in the deployment+release? At least the collation stuff and the revert sounded like that) [01:14:25] that revert was the most incredible case of collective stupidity [01:14:30] I almost fell for it myself [01:14:38] Which revert? [01:14:40] Of the diff code? [01:14:44] r75769 [01:14:51] !r 75769 [01:14:51] --elephant-- http://www.mediawiki.org/wiki/Special:Code/MediaWiki/75769 [01:15:06] Yeah that's the one I meant [01:15:15] an author and three reviewers, and nobody noticed that the patch doesn't do anything because the bug doesn't affect the code being patched [01:15:50] When I looked at it there was a previous comment about wikidiff2, and I guess everyone just assumed it was a bug that affected both [01:16:36] But yeah it's incredible how a group of people can have tunnel vision like that [01:16:53] (I went through your commits and tagged 1.17 as appropriate just now) [01:17:04] thanks [01:17:15] And reviewed most of them [01:18:07] I backed the change out of HISTORY, when you backport it'll be in RELEASE-NOTES instead [01:20:03] Yeah [01:20:16] *jorm yawns [01:32:12] nn [01:34:56] ^demon|away: Are you just gonna deploy trunk of FR or should I merge all that FR stuff that's tagged 1.17 into REL1_17? (Or is it for release?) [01:37:02] <^demon|away> Merge it to 1.17, we're gonna pull FR from that for deployment [01:37:09] <^demon|away> Most of that stuff is already in 1.16wmf4 anyway [01:37:31] Oh you *are* gonna pull from 1.17 for deployment? [01:37:40] Do you want that to coincide with the general 1.17 deployment? [01:37:57] <^demon|away> Yeah, it'll just be part of the normal deployment [01:38:09] <^demon|away> Like I said, a good part of that is already live [01:38:22] <^demon|away> We had a deployment on Dec 22nd, after the merge. [01:38:41] OK cool [01:38:48] Then I can remove the exception for FlaggedRevs from default.conf [01:39:37] <^demon|away> Yeah, should be able to [01:40:02] Committed [01:43:03] <^demon|away> The fix for extensions/FlaggedRevs/maintenance/wikimedia-periodic-update.sh probably needs tweaking though, since that's nowhere in trunk [01:48:40] Oh [01:48:50] Yeah that needs to go into $preservedItems I guess, where it's commented out now [01:49:26] *jorm $jorm [01:49:47] *jorm realizes he didn't hit his watchlist at *all* today. [01:50:04] I never hit mine, I've got zero on-wiki shit going on [01:50:08] The joys of being a developer [01:50:16] OTOH I'm 1900 e-mails behind on mediawiki-cvs [01:50:22] And a couple hundred on wikibugs-l [01:52:08] werdna: ping? [01:52:34] jorm: He just left [01:52:43] grah [02:02:08] <^demon|away> Watchlist? [02:12:34] i have an enwiki watchlist that i troll for stuff. make sure that people didn't go adding article feedback to everything, etc. [02:15:37] robla: I don't think we ever discussed the virtue of a volunteer dev coordinator (now that you all are in the hiring process...). [02:17:12] *jorm blinks. [02:18:14] Hi jorm. [02:18:20] hi Shirley [02:18:37] Is the blinking surprise? [02:19:16] more like "i thought this had been discussed [02:19:27] also that i thought you'd be into the idea of the position. [02:20:04] Last time I brought up, robla deferred the conversation, I think. [02:21:23] hrn. well, it's dinner time for him, probably. [02:21:44] yup :) I was just about to say that I just called in a dinner order that I need to go pick up [02:22:11] i have people coming over in half an hour. [02:22:27] and i don't know jack about the position or i'd answer questions. [02:23:26] The bugmeister position was created to have someone bug paperwork, not actual development. [02:23:28] *jorm starts a "to do" list for LQ screens. [02:23:34] yup. [02:23:35] I guess it's kind of unclear to me whether that's still the case. [02:23:44] afaik that's still the case. [02:23:55] Well, Mark doesn't fit most of the outlined requirements. [02:23:58] i think mark is perfect for it, tbh. [02:24:07] The idea (as explained to me) was to pick someone who specifically wasn't a developer. [02:24:11] he has a mental model of the bug base. [02:24:15] mark isn't a developer. [02:24:21] Yes hei s [02:24:23] You should tell him. [02:24:27] wait, yes he is. [02:24:37] I was under the impression he was training to be a second Tim. [02:24:39] sorry, i blanked that. for some reason i thought he was doing code review. [02:24:48] <^demon> he's also doing that ;-) [02:24:50] well, ask him what he's doing, then. [02:25:02] either way, the job needs doing, and he can do it. [02:25:16] He's also a /temp/ BM until we hire one, right? [02:25:33] that is my current understanding, but don't quote me as authoritative. [02:26:00] <^demon> Last I heard we weren't hiring BM and the responsibility will eventually be rolled into the dev coordinator. [02:27:28] Last I heard we were hiring a VDC first then a BM [02:27:48] I think this means I need to ambush robla when he comes into the office next week [02:28:14] is he in next week? [02:28:37] I think he's in every other week [02:28:43] And he wasn't here at all this week [02:28:59] yeah. [02:29:44] Well, [02:29:51] I'm glad I'm not the only one who's confused. [02:30:01] Dude I'm right there with you [02:30:31] (Of course I have plenty of other things to worry about ATM, but I was there when we pitched the BM idea to Brion in April 2009 so I really want us to hire one already) [02:30:44] My larger point is that these people (while somewhat possibly needed) aren't really going to attacking the other issues... like code review (particularly patches and the like). [02:30:45] iirc, the blog post said that we are making mark the BM temporarily and that we are focusing on the VDC. [02:31:06] Though maybe that's what Mark will be doing. But that doesn't fit in with the Bugmeister position (as least as it was written). [02:31:09] *Shirley shrugs. [02:31:33] There's no reason the BM couldn't deal with patches, is there? [02:31:45] The role was written as a non-technical role. [02:31:51] Almost completely. [02:31:52] I mean if they're not a dev they couldn't review them, but they could at least figure out who should review them and bug them [02:32:05] Not that different from bugs really [02:32:14] They specifically sought people who weren't developers, because they wanted people to work on bugs paperwork, not try to fix them. [02:32:19] There are direct quotes about this somewhere. [02:32:33] <^demon> Yes, that is the point. [02:32:35] Well I do think the BM should maybe be a bit more of a technical role, like not a complete non-tech person [02:32:57] ^demon: Which makes Mark kind of an odd choice... [02:33:02] Even if only temporarily. [02:33:23] And if they can review a simple patch or fix a simple bug, fine, go for it, but their main job would be to triage bugs and push digital paperwork [02:33:31] Yeah Mark's overqualified I guess [02:33:33] Maybe the VDC will be doing code review and patch review, but I doubt it. [02:33:52] Code review should be done by developeres [02:33:57] You can't review well if you don't code [02:34:07] And you can't keep your review chops up if you don't keep coding [02:34:16] You mean volunteer developers or paid? [02:34:24] Either [02:34:34] We talked about this in here, I think. Paid developers keep getting sidetracked. [02:34:37] Just in general, to be able to review stuff well you need to have hands-on experience with the code [02:34:48] And that experience needs to not get outdated [02:34:50] code review is a thankless job. difficult to get volunteers to do it. [02:34:52] Including Tim, for that matter. For years... [02:34:56] Yes, that's a whole 'nother issue [02:35:04] Paid devs not getting enough time for CR [02:35:17] Right. That's my concern with the VDC. [02:35:20] In a nutshell, I guess. [02:35:30] you think they'll get sucked into code review? [02:35:40] Well I had a similar concern when they basically said "we're hiring someone to do CR" [02:35:48] No, I think they'll be the replacement for someone who could be doing CR. [02:35:50] Like you can't just hire someone and have them do CR. A dev matures to a reviewer [02:36:12] you realize, right, that we're NEVER going to hire someone to just do code review? like, ever? [02:36:20] <^demon> You shouldn't dedicate any one resource to CR. You burn out too easily. [02:36:21] Well sure, if currently devs are kept too busy to do CR, then we need more devs to free them up [02:36:24] Yes [02:36:24] jorm: My other concern is that WMF is 0/1 on volunteer coordinators. [02:36:25] <^demon> It's a waste of time. [02:36:29] CR is not a person, it's a group [02:36:32] Shirley: ? [02:36:45] RoanKattouw: Cary's position no longer exists for a resaon... [02:36:48] reason * [02:36:51] And CR is not entry-level, it's done by experienced folks [02:36:56] Oh, meh, they're not hiring a new Cary? [02:36:56] what reason is that? [02:37:06] They're hiring a developer Cary. [02:37:20] Well sure [02:37:35] But I'm not sure that's related to the absence or presence of a new regular Cary [02:37:38] Right. Which is why I keep asking if anyone looked at that position before trying to create a replica. [02:37:54] If it pretty much failed as a standalone position. [02:37:57] Well Cary didn't spend much if any time on tech things, did he? [02:38:04] And how did Cary's position faik? [02:38:05] *fail [02:38:06] jorm: Reasons, I guess. Unclear scope seemed to be the biggest reason. [02:38:15] It didn't work out. [02:38:21] Is there a better word to use than fail? [02:38:28] The position no longer exists. [02:38:32] Did you mean the fact that he left, or something else? [02:38:33] Right [02:38:39] unclear scope or unclear definition? [02:38:43] they are different things. [02:38:53] And I have no clue what the reasons for that are, and I don't think anyone's said anything about it either, so it's all just speculation [02:39:03] *^demon raises hand [02:39:05] <^demon> I have a question [02:39:05] jorm: They seem fairly similar in context. [02:39:12] yeah. i know he left to go study religion full time. [02:39:18] *Shirley calls on ^demon. [02:39:28] <^demon> What is a volunteer dev coordinator supposed to do? [02:39:31] jorm: But the position was then eliminated. [02:39:40] I think it's insane to not read into that. [02:39:50] well, scope implies "the degree to which power can be exercised" and definition means "what powers are available." so it's subtle, but interesting. [02:39:52] He was one of the oldest WMF employees, as I recall. [02:39:58] positions get eliminated for MANY reasons. [02:40:09] Sure. [02:40:17] So before creating a similar position, it makes sense to examine the past. [02:40:24] And figure out if there are lessons there. [02:40:28] ^demon: offhand, i'd say "wrangle volunteers into doing work and provide a bitching point." [02:40:44] ^demon: Well, there's a posted job description. [02:40:46] mediawiki work is FAR different than wikipedia work. [02:40:46] <^demon> You shouldn't need a bitching point. [02:40:58] http://wikimediafoundation.org/wiki/Job_openings/Volunteer_Development_Coordinator [02:41:07] yup. that seems right. [02:41:08] Not that those are generally useful or helpful. [02:41:13] well, "point of contact" [02:41:24] traffic cop for issues, maybe. [02:41:29] *^demon shrugs [02:41:31] anyways, that seems to be what it is. [02:42:18] looking around and finding someone willing to do some performance work on the login system is a different gig than herding wikipedian cats. [02:42:21] Is a traffic cop what's missing from code development? [02:42:31] no, but it could be useful. [02:42:39] So could a lot of things. [02:43:00] Could be useful doesn't scale so well with very finite resources (including and particularly people). [02:43:10] a central point of contact will allow someone to answer questions like "who do i talk to about X?" rather than people just waiting around for an anwer. [02:43:17] <^demon> "Publish and promote list of possible projects for newcomers, ensuring the list is easy-to-find, properly prioritized, and current" [02:43:23] <^demon> ^ That's the single best idea on the list [02:44:07] is "explain why patches are rejected" on there? [02:44:21] No mention of "patch" on there. [02:44:29] Or "bug". [02:45:05] Or "issue," "ticket," "review' ... :-) [02:45:09] well, a big set of the jobs there are about ensuring we're transparent, which i figure you'd like. [02:45:37] I missed http://wikimediafoundation.org/wiki/Job_openings/Software_Developer_(Features) [02:46:06] that looks like a backend developer. [02:46:32] Transparency through obfuscatory job descriptions, heh. [02:46:38] I read these and the words just pass over me. [02:46:44] well. [02:46:54] i pay attention because i like to know what resources are available. [02:46:55] Maybe it's me. [02:46:59] we're starving for front-end. [02:47:09] "The focus of this role is on PHP/MySQL development." [02:47:14] that's a back-end developer. [02:47:17] it's pretty clear. [02:47:34] if it said CSS/Javascript, that would be front end. [02:47:43] Sorry, but how the hell does that match "features software devleoper"? [02:47:50] front end guys don't know from DROP TABLE or TRUNCATE [02:47:52] developer * [02:47:57] features developers are both. [02:48:05] That's a pretty strange definition of features. [02:48:10] it basically means its alolita's team and not robla's. [02:48:17] not at all. [02:48:22] what is your definition of "feature"? [02:48:32] A software enhancement. [02:48:39] <^demon> All developers should have some knowledge of the backend *and* frontend. [02:48:46] sure, but that's meaningless in this context. [02:48:54] Not really. [02:49:01] software's been published. everything we do is supposed to be an enhancement. [02:49:02] Fetaures implies frontend, surely. [02:49:10] Features * [02:49:13] I can't type. :-( [02:49:15] making the database respond faster is a core system. [02:49:39] yes, it implies front end experiences. but "features developer" doesn't imply that the code reaches the front end. [02:49:46] andrew is not really a front-end developer. [02:50:13] <^demon> I can make boxy layouts with borders, that's about it :) [02:50:20] trevor is more front-end than back-end. neil sits on the fence. arthur is back-end. kaldari front-end. [02:50:29] right. back-end. [02:50:45] development is usually done in pairs, a front-end and a back-end. [02:50:52] <^demon> I can put a *functioning* UI on my work. But it sure as hell isn't pretty :p [02:50:57] I wouldn't describe back-end as "features" and hope for anyone to be able to understand what you're talking about. [02:50:58] Sure [02:51:10] It's esoteric. [02:51:11] Yeah, I also think we should be hiring a features dev [02:51:12] you have to understand the feature. [02:51:19] General eng is even more understaffed [02:51:28] i think we need about five people, personally. [02:51:35] The features vs. general eng thing is probably irrelevant for the actual hiring process in practice [02:51:37] Well, you're hiring three. [02:51:55] Actually, five in "technology." [02:52:05] http://wikimediafoundation.org/wiki/Template:Jobbox for reference. [02:52:28] Ah, right, Perf Engineer [02:52:33] That's a general eng thing [02:52:44] Is that like a DBA? [02:52:47] Is that a DBA? [02:52:48] Yes [02:52:50] Probably [02:52:53] yeah. [02:52:53] Well, DBA plus other stuff [02:53:05] domas/tim hybrid. [02:53:06] And yet it doesn't use the word "database." [02:53:17] Like, Domas doesn't just mess with our DBs, he also profiles things in MW and fixes them [02:53:18] I can't be the only one who finds this needlessly "wtf." [02:53:19] because performance engineers know that's implied. [02:53:27] i think you may be. [02:53:35] Okay. [02:53:45] Data Analyst and VDC are hardly technical positions, although they're in the tech dept [02:54:03] <^demon> VDC job description asks for a technical background [02:54:22] I see irony in asking almost every applicant to be able to write clearly and effectively when these job descriptions seem unclear to me. [02:54:33] But it's probably just me. [02:54:49] tech resources, tech budget, roan. [02:54:52] Yeah I've always thought the JDs are weird [02:54:54] Sure [02:54:57] It is in the tech dept [02:55:02] It's just not a technical job [02:55:35] I mean, Danese is in the tech dept (heads it, actually), but she doesn't really do any technical things as part of her work [02:55:40] And that's fine, we need that kinda stuff too [02:56:07] But in a lack of engineering resources discussion they're kinda irrelevant still [02:58:59] <^demon> I'm tired. Goodnight. [02:59:05] Night, ^demon. [03:09:53] jorm: still need me? [03:10:07] no, i got peeps over now anyways. [03:12:41] ahk [03:12:49] I'm going to wander over to the castro pretty soon [03:13:13] due at 8:30, but I was thinking of walking [03:13:27] how advisable is walking down market st at 8pm on a thursday night? [03:13:48] It's quite a way to walk from the apt to the Castro [03:14:17] that's why I'm allowing an hour [03:14:18] You can take Muni, there's a station at Market/Castro/16th (Castro stop) [03:14:20] OK [03:16:43] I know I can take MUNI, but I was wondering if it would be interesting to walk [03:17:47] I don't know. I've walked from an Indian place near the Tenderloin to Noisebridge once [03:18:20] Takes Market part of the way, from ~8th to ~11th I think [03:18:28] Or wherever Valencia branches off [03:39:25] i would not do that walk after 10 pm. [03:39:29] not even after 9. [19:45:35] How is the 1.17 release going? [19:45:58] We're hoping to deploy early next month [19:55:13] Awesome, thanks ;-) [19:57:36] Hi werdna, thanks for the update on bug 19699 [19:58:00] also... [19:58:45] werdna: are you still working on http://www.mediawiki.org/wiki/Extension:LiquidThreads/Q1_2011_re-engineering ? I'd like to merge it with http://www.mediawiki.org/wiki/Extension:LiquidThreads/WMF_project_information [20:49:31] robla, i'd prefer you didn't yet. [20:49:39] because we've got, like, 5 documents. [20:50:04] people only ever read one version of a document, as it were. [20:51:25] How's this for killing your childhood: Cobra Commander is dead now. In the comics. Shot in the head. [21:12:53] jorm: I'm confused....I'm trying to reduce the number of documents [21:13:35] we're writing diff. docs for different audiences. [21:13:59] how are those two for different audiences? [21:15:50] jorm: well, I don't have a problem having a single landing page [21:15:58] those ones aren't; i'm currently trying to get a handle on everything and where it is. [21:16:06] as am I [21:16:08] Since what I wrote is just the "landing page" where people can go to find out what's going on with the project [21:16:16] so I'm fine with those two pages being merged [21:16:26] The WMF project information page is out of date, anyways [21:16:26] then merge them; i haven't had a chance to look at them. [21:16:53] robla: if you'd like a hand, I'd be happy to help out [21:17:10] I do want to make sure that there's documentation to make sure that everybody's on the same page [21:17:21] werdna: thanks! [21:17:54] jorm: I'm mostly sitting in front of a large piece of A3 paper (yay to Dana for helping me find it) with a pencil and ruler [21:17:58] doing schema designs [21:18:03] the backend design document is more or less finalised [21:18:25] *robla looks at the subpages of Q1_2011_re-engineering... [21:18:27] robla: we're going berserk with documentation on this. [21:19:41] robla: you'll find a lot more on office.wikimedia.org/wiki/Special:PrefixIndex/LiquidThreads [21:21:46] *robla starts merge [21:22:00] thanks robla, let me know if you want a hand or two [21:22:46] werdna: sure thing. I'll do the initial rough merge, and then feel free to munge to your hearts content [21:22:55] :D [21:26:56] i'd prefer to not merge my documents yet. [21:29:57] jorm: which documents are you talking about? [21:30:17] they're in my user space; you can see them with my contribs, i suppose. [21:30:31] anyways. if it's in my user space, please don't merge it yet. [21:31:17] that's fine [21:31:44]