[00:07:26] Hi. Do we have a database code for userrights-interwiki for http://prototype.wikimedia.org/flaggedrevs/Special:RecentChanges ? [08:05:31] if anyone is around with a long enough memory... did revision text used to have \r\n for blank lines or something back in the day and at some point we stopped doing that? [08:07:02] basically I can see some very old revisions with \r\n in them and new ones don't, but the xml representation (and therefore the dumps) strips out the \r and I just wonder if we care [18:17:46] ok so what do we use instead of the interwiki table these days? [18:18:00] I see it's empty in newer wikis but dang if I can figure out what replaces it [18:23:02] apergos: a cdb file somewhere, afaik [18:23:19] hmm I wonder how that's generated [18:23:22] *apergos pokes around [18:29:41] Jorm (and anyone else familiar): I have a couple of questions about Athena: a) Do you expect it to change much? b) Is the WMF doing anything differently this time compared to Vector because of some of the negative comments then about communications, etc? [18:29:54] (This is for the Signpost Tech Report, I should add.) [18:34:10] i'm here. [18:34:42] a) yes, I expect it to change much. It's a first run design; we're in a discussion phase. [18:35:34] b) I'm not sure what this question means. [18:35:46] or, what you're asking about. [18:36:52] One sec, I'll make sure I've got my facts right first :) [18:41:23] oh seriously [18:41:39] $this->makeInterwikiSQL( $this->getOption( 'd', '/home/wikipedia/conf/interwiki/sql' ) ); [18:41:43] I mean really? [18:43:23] seems legit. [18:43:49] jorm: Okay, nope, can't find. Anyhow, I seem to remember some complaints along the lines of "I/we never asked for it, I/we don't want it, it's being forced on me/us..." [18:44:43] Given that we're at the beginning of the Athena road, is there a plan to avoid that ending? [18:45:58] that looks like it builds these tables which are supposedly obsolete now though.. grrrr [18:47:18] there will always be people who dislike any and every change. that's unavoidable. that said, none of that has gone into my design methodology so far; deployment issues are far out. [18:48:10] i am an *extremely* transparent designer, so even if a "full deploy" scenario happens, people will feel that they've had ample time to discuss it and voice objections. [18:48:33] but right now, no, there is no "deployment plan" [18:48:39] does that answer your question? [18:49:22] speaking of athena, i'm about to dump maybe 75 athena-style icons up to commons. [18:49:34] Sure, that's a good answer :) [18:49:58] So in essence you're working on getting the design right first, then figuring out about deployment nitty-grity. [18:50:00] *? [18:50:30] heh. more like "I'm working on getting the design right." [18:51:09] there is enthusiastic support for it but it's not even an "official" project yet. [18:52:47] Ha, yes. So would you say that the whole project is a long way off then? Or could it pick up quickly? [18:52:54] i would very much like it if this were NOT to be a "the foundation is doing stuff without community input" article, because that's not true. [18:52:57] no, i wouldn't. [18:53:05] there are multiple aspects. [18:53:05] t [18:53:06] her [18:53:07] e' [18:53:08] s [18:53:11] my broken irc client. [18:53:18] there's the mobile side, and then the tablet side. [18:53:43] the mobile side will probably be VERY quickly. within a few weeks i'd expect parts to leak onto the beta site. [18:53:54] Ah, gotcha. [18:54:01] as for the main sites, i don't know. [18:54:38] if there is significant community push (like, from the editor community) and support from the developer community, it might raise in priority. i don't know. [18:54:48] sometimes our scheduling is a myster. [18:54:51] mystery. [18:55:14] do you want to talk more about this, or did you just have those few questions? [18:55:14] I don't think I've ever written a "the foundation is doing stuff without community input" article, actually, and I'm not about to start now, don't worry :) Thanks for clarifying everything. [18:55:29] No, I think that'll be fine but I'm sure there'll be more to talk about later. [18:55:31] there have been a couple. [18:56:43] Later as in next weeks, months, etc. [18:57:43] Thanks for the answers, I'll try and post back later with a sketchy article. [18:57:44] off the record, i don't think this is a "full roll out, no stop" type thing. [18:57:56] we're breaking too many normalized paradigms. [18:58:41] in my experience with handling change control, the changes we're talking about are pretty radical. [18:59:08] I guess it's a case of them being not so radical on the mobile platform, and then [18:59:43] you ask "why don't we do things like this all the time"? [19:01:22] Anyway, I have to go, but I think it's a good project and hope to be writing about it lots for a long time to come :) [19:01:32] right. they're *not* radical on the phone. [19:01:43] hrm. when did i ask that? [19:02:04] Sorry, I meant people in general. [19:02:32] We should really have a way of differentiating that in English. [19:02:40] Anyway, really have to go, thanks for the answers. [19:09:36] got the little bugger [19:10:10] had to specially hack maintenance/dumpInterwiki.php which means that script should be fixed up someday [23:32:36] K4-713: I just reviewed that second change you made to lp-filter.c, I marked it OK [23:33:13] this bit looks strange to my eyes: [23:33:14] int filtercount = sizeof(urls)/sizeof(*urls); [23:33:18] int url_length[filtercount]; [23:33:32] but I guess if it compiles on the version of gcc we're using then it's not a problem [23:34:38] Hello, TimStarling. [23:34:44] Yeah, that bit was a bit tricky. [23:35:01] if it was filtercount was a const then it would definitely compile [23:35:18] Without dividing by the pointer size, it was counting 16 filters. [23:35:20] then you'd just have to use a compiler that supports const, not one that propagates constant expressions in this way [23:35:46] you know arrays on the stack have to have a fixed size at compile time [23:36:07] yeah, you have to divide by the pointer size, sorry for forgetting about that [23:36:46] No worries. [23:37:58] So, do you think we should leave it the way it is? [23:38:49] I think you should add const [23:39:02] okay. [23:39:40] I'll get that in asap. [23:40:01] Should just be a couple minutes. [23:40:13] it's not urgent [23:40:48] I'm just saying that that's the right way to do it [23:40:58] O [23:41:25] awjr: I'm going to delete the copy of lp-filter.c from /a/squid/fundraising to avoid any confusion [23:41:50] TimStarling: sounds good [23:44:17] ok, it's deployed [23:44:29] TimStarling: awesome! thanks [23:47:04] Neat! [23:47:37] awjr: can I get a filtered log sample just to make sure it all looks okay? [23:49:33] K4-713: sure. i'm tailing the log now and i'm seeing requests for donate wiki showing up:) [23:49:48] Nice.