[00:22:25] I meant "better geographical representation" above. * [00:23:07] Dispenser: There's a difference between ?action=purge and a null edit, though. A null edit will purge all the caches and rebuild the links entries (AFAIK). [00:24:16] All I know it GeoHack and python get old cached data, and I've determined that something's not being cleared [00:24:31] How are you retrieving? [00:25:00] The API should always be returning the newest wikitext with ?rvprop=content, for example. [00:25:09] If you're hitting HTML caches, it's a whole different story. [00:26:01] HTML, because I don't want to write a routine to handle language and edit/history links [00:27:13] You can't be surprised if the HTML is old and cached. [00:27:27] If you're logged in, it would help. [00:27:31] Might also be a lot slower, though. [00:29:09] I might be more inclined to rewrite it, if Kolosso didn't keep added undocumented OSM and de specific hooks [03:32:21] robla: Ping. [03:32:30] Ashlee: pong [03:32:38] > Was the Wikimedia community invited to join this conversation? [03:32:38] Yes, at Thu, Oct 7, 2010 at 8:28 PM [03:32:41] I don't know what that means. [03:33:05] Oh. [03:33:07] well, I should have actually cut and pasted the timestamp from my mail [03:33:20] Yeah, I'm not sure you want to play the snark game with me. [03:34:28] anyway, any comments on the substance? [03:38:48] FlaggedRevs is an overbloated POS? [03:39:00] I'm not sure what more there is to say. [03:39:06] I don't think UI is really what's annoying people about it. [03:39:23] The fact that it's slow as shit seems to be a much greater concern. [03:39:56] Ashlee: which part is "slow as shit"? [03:40:16] Diffing, as I recall. [03:41:00] hence why Priyanka is working on that right now [03:41:28] diffing is slow with or without Pending Changes, btw [03:41:39] it's actually not diffs, it's display of old revs [03:42:17] If you say so. [03:42:20] Where's the relevant bug? [03:42:29] Never mind. [03:43:05] https://bugzilla.wikimedia.org/show_bug.cgi?id=25289 [03:43:11] yup, that's the one [03:44:27] I'm looking for a link. [03:48:29] http://en.wikipedia.org/w/index.php?title=User_talk:Iridescent&diff=386624914&oldid=386622082 [03:50:34] I trust Gurch (more than trust most of you all, at least). [03:51:10] There's also http://en.wikipedia.org/w/index.php?title=User_talk:Risker&diff=prev&oldid=386185420 [03:59:16] robla: Your conversation skills are second-to-none. [03:59:18] :-) [04:17:03] I don't think Gurch's characterization of how PC deals with templates is correct, actually. if it is, I'm pretty sure it's a bug [04:17:25] It seemed accurate to me. [04:18:23] testing now.... [04:23:43] Ashlee: if I'm reading Gurch correctly, this page http://prototype.wikimedia.org/flaggedrevs/Test_page ...should probably say "Template version 1" on it, but it says "Template version 2" [04:25:04] That sounds like a different bug. [04:25:09] Aren't templates supposed to be frozen? [04:25:35] Nope, not in this config [04:25:46] What? [04:25:54] Is that how it is on en.wiki? [04:26:04] yup [04:26:07] That doesn't make any sense. [04:26:15] If someone vandalizes a template, their vandalism is shown immediately? [04:26:22] What the hell is the point of FlaggedRevs, then? [04:26:44] well, presumably, the templates have appropriate protection on them [04:27:28] Any unprotected template is equally damaging.... [04:27:36] All you need is one. [04:27:50] leaving PC out for a sec: what happens when there's a semi-protected page and an unprotected template? [04:28:04] The template stays unprotected. [04:28:18] And the page itself is semi-protected. [04:28:37] Why do you ask? [04:29:18] *robla runs another test [04:29:49] http://prototype.wikimedia.org/flaggedrevs/Test_page [04:33:14] http://prototype.wikimedia.org/flaggedrevs/Test_page_2 [04:33:27] that's a fully protected page, with an unprotected template [04:34:03] pending changes affords the same level of protection to a page that semi-protection and full protection offer [04:34:18] http://prototype.wikimedia.org/flaggedrevs/Test_page_2 [04:34:33] If it affords the same level of protection, what's the point of using it? [04:34:36] It has a worse UI? [04:34:40] It's slower? [04:35:04] It uses worse terminology? [04:35:08] I'm lost. [04:35:42] I'm not going to try to sell the feature to you [04:36:17] What are you trying to do? [04:37:12] create an updated version for November [04:37:35] Well, there's a gaping security flaw in the current configuration. [04:37:40] So perhaps that should priority 1. [04:37:48] should be * [04:38:06] no more so than with semi- and full protection [04:38:26] You're not making any sense. [04:38:38] FlaggedRevs is an _extension_. [04:38:46] It's supposed to _extend_ the functionality of MediaWiki. [04:38:57] By allowing users to specify pages as "good." [04:39:25] If it doesn't do that, it doesn't matter if it has similar functionality to any core MediaWiki feature. Hell, at that point it's just redundant code that should be killed. [04:40:18] Saying "well, it's the same as what you had before" is mind-boggling. The point was to add a level of protection, not duplicate the currently existing options. [04:43:27] <^d> Yes it was. FlaggedRevs can operate in several modes w.r.t. template inclusion. [04:43:52] <^d> See the docs for FlaggedRevision::findPendingTemplateChanges() and the various FR_INCLUDES_ constants. [04:44:03] "can" and "does" are completely different. [04:44:11] <^d> Shut up and let me finish. [04:45:19] <^d> The pending changes style of configuration (ie: enwiki) was designed to be an alternative for semi-protection. Thus, it would follow that the behavior for template inclusion would be what happens with semi-protection. [04:45:33] <^d> Having FR_INCLUDES_CURRENT set has always been a part of the public configuration. [04:46:09] Your "thus" doesn't follow. [04:46:19] You have an alternative because there's a distinction between two choices. [04:46:28] Emulating a choice doesn't make it an alternative. [04:46:33] <^d> For you. This is the first time I've heard the complaint about templates not being stablized in the enwiki configuration. [04:46:51] All that time you spend at en.wiki. [04:46:54] :-) [04:47:34] I'm not really trying to be a bitch. But I genuinely don't understand how this makes any sense. [04:47:52] It makes a bit more sense now that you've explained you were trying to emulate semi-protection. [04:48:01] But the question then becomes "why would you ever try to do that?" [04:48:27] When people at en.wiki said they wanted an alternative to semi-protection, I think they wanted an... alternative. [04:48:31] <^d> I didn't come up with the requirements, or implement ~99% of it :) [04:48:33] Not the same thing but slower. [04:48:42] And deceivingly uneffective. [04:48:48] Ineffective? [04:48:52] I'm hungry and tired. :-( [04:51:42] <^d> I've been staring at code too long, my eyes need a break. [12:24:26] any ops for #mw? [23:11:45] adam_miller? [23:11:47] reedy? [23:11:52] ? [23:13:28] heh. guess what i just had a meeting about? [23:13:34] please god no [23:13:45] harhahahaahaahaahahah! [23:13:55] phase 2 design work starts next week! [23:14:22] yay? [23:14:34] actually, yeah, i think. [23:14:42] lol [23:14:54] i'll be working on it and probably ping you with some questions. over the next couple days. [23:15:02] mmm, alrighty [23:19:21] hmmm [23:42:02] jorm: Does staff-l allow for posts from non-members? [23:42:21] i have no idea. [23:42:24] Okay. [23:42:35] Someone was running their mouth about me. I thought I might set the record straight. [23:42:50] Of course, they wouldn't say it to me directly... [23:43:37] on the staff list? [23:43:42] Nah. [23:43:50] People wouldn't be _that_ stupid. [23:44:00] Elliot Spitzer had a quote about that, I think. [23:44:11] what did you hear? [23:44:20] Doesn't matter. [23:44:41] well, it clearly does, if you're wanting to set a record straight. [23:50:50] *jorm -> homeward.