[00:00:14] the first is an array( 0 => 1 ) [00:00:26] array(1) { [1]=> string(3) "Foo" } [00:00:26] array(1) { [0]=> int(1) } [00:00:27] yeah [00:00:39] So you can do array (1 => irrelevant value ) [00:00:47] Right [00:00:49] Or, if you have a lot of stuff [00:00:56] array_flip() , which flips keys and values around [00:01:13] Hence saying, it might be worth just putting the title in there, for human ease? [00:01:20] So array_flip( array( 123, 456, 789) ) == array( 123 => 0, 456 => 1, 789 => 3 ) [00:07:19] Meh, commited [00:07:21] What to do now [00:07:36] Bugs and code inspection.. [13:20:21] there is a bug in the deployed code. [13:20:51] for some reason the wikiEditor isn't loaded when you switch from vector -> monobook [13:21:24] WikiEditor.combined.min.js:3TypeError: Result of expression '$j.wikiEditor' [undefined] is not an object. [16:47:18] Reedy: what up what up [16:47:40] just sent an email re: AA status with a couple questions for you [16:50:01] Morning [16:50:03] Just reading [16:51:01] adam_miller: question [16:51:07] (i am responding to your email) [16:51:22] jorm: (ok) [16:51:26] when a user is selecting their stars, and before they hit submit, what color are you leaving them at? [16:51:32] RED [16:51:33] (they should be red until submit, then blue) [16:51:35] just changed that [16:51:35] excellent. [16:51:51] i wasn't sure if i explicitly called that out in documentation. [16:51:55] you sure did [16:52:01] and i actually read it this morning [16:52:11] instead of just looking at the pretty pictures [16:52:57] jorm & Reedy - just committed my work from this morning [16:53:13] I'll have you comps for the thanks message and the error message soonish. [16:53:29] ok, can you reuse your styling for the stale message? [16:53:33] just like different colors maybe? [16:53:36] yes. [16:53:40] that was going to be my plan. [16:53:44] perfecto [16:53:51] i thought that may be your plan, so i have the code in place already [16:53:55] just need the css styles [16:54:02] am i that predictable? [16:54:08] nah, just makes sense to put them there [16:54:34] and you seem like a sensical guy [16:54:52] or sensible is maybe what i mean [16:55:11] harh. [16:55:45] most of the time my design justifications start with "I've had some whiskey, and I've been thinking..." [16:56:31] sounds good to me [16:56:58] Reedy: do you think we should send back updated ratings data when a user submits a rating? [16:57:21] i can probably just update the values with js....trying to decide if that's dirty though [16:57:31] Well, someone else "could" have rated [16:57:42] It's whether the api returns the updates on update [16:57:45] or we just do another call [16:57:59] but that's all the api would be doing (calling the other module to append to the results) [16:58:00] i'd vote for returning the updates on update [16:58:19] one http request > two http requests [16:59:45] mmm [17:00:02] just need to have a poke and see how we get it to use the other module to add on the results [17:00:12] might just need a bit of refactoring to reuse the functions [17:02:55] is it possible for us to know, when a user submits a rating, if this is a new rating or a rerating? [17:03:02] because i'll want to change the message if so. [17:03:48] not easily [17:04:04] in a nice world, insert or ignore would tell us if there was existing data [17:04:17] it's not a killer. [17:04:20] but ignore is just like some error su[21:20:04] Reedy: how we doing with those additions to the api? [21:20:11] which? [21:20:18] all of them [21:20:20] staleness [21:20:23] done [21:20:24] userid [21:20:26] done [21:20:28] committed? [21:20:29] usetoken [21:20:30] ypu [21:20:48] alrighty [21:20:52] imma take a look [21:20:59] jorm: how we doing with those extra comps? [21:21:38] i got two of them up already. are they insufficient? [21:21:41] i'm in a meeting right now. [21:22:21] and i can get anything else to you afterward. [21:22:47] imma take a look [21:23:43] jorm: good, i think they'd look better with a 1px border, but they work [21:23:50] Reedy: aanoanontoken? what do i need to do with that? [21:24:01] just pass that 32 character thing [21:24:04] i am [21:24:09] it's still yelling at me [21:24:10] try them with 1 px and see. i did two so they stand out more. [21:24:16] it's okay to use 2 px on all borders. [21:24:44] aaanontoken# [21:24:46] aaanontoken [21:25:18] oh. i geddit. *a*rticle *a*ssessment *anon*ymous *token* [21:25:27] Reedy: maybe i need to update the db? [21:25:31] trying that [21:25:41] errm, an update will work [21:25:42] ish [21:25:52] either do like i've said in the commit summaries [21:26:01] or drop the tables and run update to recreate them [21:26:45] cause i've done some schema changes, but didn't see any point doing patches for them, purely cause it's still in active development [21:27:14] i dropped and recreated, so it's working now [21:27:28] but when i submit a rating, it's still just returning 'SUCCESS' [21:27:44] i want it to tell me that, AND what the ratings are now [21:27:54] nice blue shirts [21:31:09] Reedy: ^^ [21:31:29] adam_miller, right, not worked on that one [21:31:44] RoanKattouw, got a minute? (quick question) [21:32:14] Shoot, busy fixing the site so might be slow in answering [21:32:31] sure [21:33:09] Want to use one api module in another, just refactor out the result getting, and pull it accross rather than letting it output itself? (roughly) [21:33:27] You have 2 modules that partially output the same thing [21:33:33] Then yes, factor out the result getting to a common place [21:33:56] yeah, want the submit to dump basically what they list one does after a submit [21:33:58] cheers [21:38:53] hi everyone, I have a question concerning sitenotices - why are they linking to Meta? [21:39:44] this does not allow users to see description in their native language and turn new features off on their home project, not on Meta [21:39:54] nimish_g, ^ [21:42:00] NickK: they should be linking to {{SITE}}/Special:UsabilityInitiataivePrefSwitch , which site are you seeing this on? [21:42:14] ukwikipedia [21:43:32] nimish_g, I checked a few other projects in different langauges, all sitenotices link to meta [21:44:21] NickK: ok, I'm seeing that as well... [21:44:54] Oh for God's sake... [21:45:03] WHY does fullurl point to meta in CentralNotice [21:45:42] <^demon> shared cache? [21:46:01] isn't just [[Special:UsabilityInitiataivePrefSwitch|*text*]] much better? [21:46:46] link to meta makes local images and local bug report pages completely useless, and not everyone knows English [21:47:06] If you could do wikitext [21:47:09] But it's all HTML [21:47:14] *RoanKattouw stabs CentralNotice repetitively [21:47:52] but local sitenotices do support wikitext [21:48:13] is it possible to disable global sitenotice and enable local instead? [21:48:25] That requires duplicating it 689 times [21:51:34] the problem is that this time vector is enabled on both large (e.g. 1mln+ wiktionaries) with lots of local users who are able to maintain bug report page in their language and small wikis with hardly any activity and untranslated interface [21:55:07] NickK: ok, when the CN banners get regenerated, this should be fixed....thanks! [21:55:14] RoanKattouw: I did the /wiki thing [21:55:23] OK [21:55:28] That won't work on the secure gateway though [21:55:52] right, I can't think of an easy way to detect that right now [21:56:02] what if we had it /w/index.php?... etc? [22:55:47] Reedy: any progress on returning the ratings? [22:56:01] sorry to be a hound dogg, but i'd like to head out in a bit here [22:59:15] jorm: what you think about making the backgrounds a light variation of the border color on those messages? [22:59:16] http://i51.tinypic.com/nqzpzs.jpg [23:00:39] http://i53.tinypic.com/2dqq7vd.jpg [23:01:47] adam_miller, I'm thinking it's going to be easier and cleaner to just have 2 api calls from the js [23:01:59] It would be doable [23:02:08] but it's gonna need daft amounts of refactoring [23:02:14] or stuff RoanKattouw would shoot me for commit [23:02:18] *committing [23:02:24] oh. [23:02:26] so it's probably simpler just to do it twice [23:02:29] well thats shitty [23:02:56] cant you just duplicate code? [23:03:05] I could [23:03:05] or is that what roan would shoot you for committing? [23:03:07] But the same issue [23:03:12] That would occur also :P [23:03:18] It's a nested $j.ajax call, not too bad, I've done it before [23:04:04] yeah it's not hard to setup [23:04:11] just seems silly [23:04:22] It kinda is [23:04:28] But it's a rare occurence [23:04:34] And it makes your code way easier [23:04:39] but i don't understand wikimedia api programming so if this is the less silly thing, i'll do the double ajax call [23:04:45] You can also share the code for processing the JSON result now [23:04:53] I was just gonna say that [23:05:19] just functionise (if not already) the processing + display [23:06:37] done [23:20:56] did someone fix wikieditor yet ? [23:21:18] https://bugzilla.wikimedia.org/show_bug.cgi?id=25014 [23:25:46] cause live JS errors are a bit sloppy [23:27:00] Reedy: what am i to pass for aaanontoken when a user is logged in? [23:27:09] nothing [23:27:12] should i leave it off? [23:27:14] or just pass '' [23:27:42] just leave it off [23:28:47] saying that, it's ignored if not anon [23:28:53] so '' would be fine [23:31:52] it doesn't seem to return results to me w/o passing that thing [23:32:11] for a logged in user [23:35:22] Reedy: and if i leave off aauserrating, i get a DB error [23:35:31] wheee [23:35:36] unknown column aa_page_id in where clause [23:35:47] adam_miller: i would love to do light colors but they need to be websafe. [23:36:04] your stale treatment is websafe? [23:38:03] w/e i'll switch em back [23:38:48] it *should* be but right now i'm not betting it is. i pulled the colors from a specific usability initiative palatte. [23:39:11] but it wasn't vetted by me, and i didn't look at the hex values. [23:39:17] i am going to, though, you betcha. [23:39:43] adam_miller, think i've fixed that one [23:44:29] Reedy: which one? the db error? [23:44:49] i care less about that then the not returning any data for known users [23:48:11] Reedy: it seems to be only logged in users who have not yet submitted a rating for that article [23:48:24] once i submit a rating, i then am able to see results being returned on a refresh [23:51:37] sorry? [23:53:51] ok [23:54:10] so if i'm logged in, on a page I have not yet rated, and the javascript requests the ratings data for me, this is what is returned [23:54:11] {"query":{"articleassessment":[]}} [23:54:27] even though that article HAS been rated by three other user accounts [23:54:41] just not the one i happen to be logged in as at that time [23:54:49] ya dig? [23:55:54] hang on, just having to do some work as we're rewriting how we're doing inclusion [23:56:02] i am passing aauserrating everytime [23:56:15] not sure if i should be [23:56:48] but i feel like i should...i'd like the users rating if it's there, if it's not that's fine [23:56:54] yeahj [23:57:06] the left join should just not return user stuff if they haven't rated that page [23:58:19] cool [23:58:21] i like it