[02:12:19] <RoanKattouw>	 kaldari: So it sounds like the X-Frame-Options thing broke IE8 for you huh
[02:12:29] <kaldari>	 yeah
[02:12:38] <RoanKattouw>	 kaldari: I'll send an e-mail to Chris CC you suggesting we soften the restriction to 'SAMEORIGIN' rather than 'DENY'
[02:12:52] <kaldari>	 I guess we could just override 	$wgApiFrameOptions on Commons as a temp fix
[02:12:57] <RoanKattouw>	 Rather, I'll file a bug
[02:13:11] <RoanKattouw>	 Yes, but I'd rather not do that without approval from the security guy that wrote this code :)
[02:13:19] <kaldari>	 yeah
[02:14:12] <kaldari>	 Too bad WLM-US starts in a few hours
[02:17:58] <RoanKattouw>	 kaldari: Does only IE8 use an iframe, or does every browser submit into an iframe but is IE8 the only one choking on the DENY? What exactly happens here?
[02:18:27] <kaldari>	 they all use an iframe, but only IE9 and lower seem to be choking
[02:18:47] <kaldari>	 it gives a permission denied error
[02:18:48] <RoanKattouw>	 Do they submit a form inside an iframe, or submit a form with target="iframe", or what?
[02:19:15] <kaldari>	 I believe it submits a form inside an iframe
[02:21:38] <RoanKattouw>	 Hmm, looks like in modern browsers it doesn't use an iframe? It doesn't seem to in Firefox
[02:21:54] <RoanKattouw>	 I suppose you could use FileAPI to get the contents of the local file, and something else to post a file upload?
[02:23:39] <kaldari>	 changing $wgApiFrameOptions from 'DENY' to 'SAMEORIGIN' definitely fixes the problem
[02:24:21] <RoanKattouw>	 Yeah
[02:24:28] <RoanKattouw>	 We don't have Chris's phone number on the contact list unfortunately
[02:24:40] <RoanKattouw>	 So I doubt this is going to be fixed before Tuesday
[02:25:11] <kaldari>	 Well that means we'll need to cancel the WLM-US banners until then
[02:25:34] <RoanKattouw>	 Could you selectively show them for supported browsers?
[02:25:48] <kaldari>	 yeah, that's a good idea
[02:26:48] <RoanKattouw>	 I also strongly suspect that the reason IE9 works is not because it treats XFO differently, but because it has FileAPI & binary XHR and therefore doesn't use the iframe
[02:27:01] <kaldari>	 I believe IE9 doesn't work
[02:27:04] <kaldari>	 lemme see...
[02:27:29] <kaldari>	 yeah, this breaks IE9 as well
[02:28:02] <RoanKattouw>	 Right
[02:28:44] <kaldari>	 IE9 is about 10% of traffic BTW
[02:29:07] <kaldari>	 and IE8 + IE9 is about 17%
[02:34:04] <kaldari>	 if it isn't fixed before Tuesday, I know some people are going to be rather upset. There's a lot of media coverage of the event ramping up (besides CentralNotice), and WLM uploads only work through UploadWizard.
[02:34:23] <RoanKattouw>	 Right
[02:34:55] <RoanKattouw>	 Well if you can get a hold of Chris Steipp in a timely fashion and get him to agree, I can fix this for you
[02:35:08] <RoanKattouw>	 But I don't have his contact info
[02:35:09] <kaldari>	 I'll see if I can track him down
[02:35:30] <RoanKattouw>	 You might be able to track him down via RobLa, he's Chris's boss
[02:36:02] <kaldari>	 good idea
[02:36:06] <RoanKattouw>	 I'm about to finish writing the bug report (which CCs him), I'll give you the # when I'm done so you have a reference
[02:36:25] <kaldari>	 for the future I'll suggest we add our security guys cell # to the contact list :P
[02:36:30] <RoanKattouw>	 Yeah srsly
[02:37:25] <RoanKattouw>	 My cell # is on there too, I suggest you grab that too while you're at it. I don't need Chris to deploy anything, I can handle that for you if you like (or you can do it yourself), call me any time
[02:39:30] <RoanKattouw>	 I'm going on a couple bike rides this weekend so I might be up to 2 (Saturday) or 5 (Monday) hours away from my computer
[02:41:58] <kaldari>	 Rob says the security vulnerability this patches is relatively minor - he says we don't need to call Chris, he says it would be fine if we just change $wgApiFrameOptions to 'SAMEORIGIN' on Commons and fix UploadWizard later.
[02:42:14] <RoanKattouw>	 OK
[02:42:18] <RoanKattouw>	 Then I'll do it right now
[02:44:17] <kaldari>	 thanks
[02:44:41] <kaldari>	 are you doing a live hack or a commit?
[02:44:48] <RoanKattouw>	 A commit
[02:44:51] <RoanKattouw>	 Should be all set in 5 mins or so
[02:44:57] <kaldari>	 yay
[02:45:04] <RoanKattouw>	 Want to finish the bug report first so I can reference the bug #
[02:50:09] <RoanKattouw>	 kaldari: All done, see https://bugzilla.wikimedia.org/show_bug.cgi?id=39877 and https://gerrit.wikimedia.org/r/#/c/22290/
[02:50:43] <kaldari>	 awesome, I'll test in IE9...
[02:51:46] <kaldari>	 seems to be working now
[02:52:28] <kaldari>	 thanks for the help!
[02:52:40] <kaldari>	 1000 WLM participants thank you :)
[02:54:53] <RoanKattouw>	 Yay
[06:42:02] <LukeDev>	 I need help, I'm using the python simplemediawiki api to query article data from wm
[06:42:35] <LukeDev>	 Does anyone know specific parameters to use to query the article text only?
[06:43:03] <LukeDev>	 I don't want wikitext or anything else.
[06:43:22] <LukeDev>	 just the article. Any help would be appreciated
[07:02:26] <LukeDev>	 Can I get some help!
[07:03:20] <LukeDev>	 I need some help making a proper media wiki api query!
[07:03:25] <RoanKattouw>	 LukeDev: There's no simple way to get "just the text". The API offers wikitext or HTML, your best bet is to grab one of those and try to strip it
[07:03:38] <RoanKattouw>	 (And shouting for help isn't gonna do much at midnight on a Friday night :) )
[07:03:56] <LukeDev>	 forever alone.
[07:06:18] <LukeDev>	 Thanks Roan, I just couldn't find the request in the documentation and was surprised that something so essential was missing- I thought I was missing something.
[07:06:38] <LukeDev>	 I'll use an nat lang toolkit to process the html.
[07:56:06] <yuvipanda>	 LukeDev: playing around with en.wikipedia.org/wiki/Special:ApiSandbox helps :)
[07:56:26] <yuvipanda>	 and if you want text action=parse is what you should be looking for
[09:14:17] <LukeDev>	 YuviPanda! I met you at Wikimania
[09:14:40] <LukeDev>	 If you'
[09:15:00] <LukeDev>	 re still on I've been pulling text however it's a mess
[17:59:07] <Vito-Genovese>	 we (turkish projects) seem to be experiencing a problem with dotted and dotless i's
[17:59:30] <Vito-Genovese>	 could a dev assist me in developing a solution for this?
[21:37:26] <LucianNovo>	 Could anyone help me understand the simline timeline?
[21:51:44] <LucianNovo>	 I need to understand how events are created and tagged when people read create timelines.
[21:57:41] <marktraceur>	 LucianNovo: Can you link to some code? I don't know what you mean, but maybe I (or someone else) could help you understand it
[21:59:36] <LucianNovo>	 Okay so the following timeline was created using json data fetched from wikipedia somehow:http://www.simile-widgets.org/timeline/examples/cubism/cubism.html
[22:00:06] <LucianNovo>	 as you might see, it took json data: http://www.simile-widgets.org/timeline/examples/cubism/cubism.js
[22:00:42] <LucianNovo>	 I want to understand how to fetch events
[22:01:23] <LucianNovo>	 What api do I use for that?
[22:01:27] <marktraceur>	 LucianNovo: It looks like this person compiled their own JSON data, they didn't get it from WP itself....I think
[22:01:36] <LucianNovo>	 really...
[22:02:29] <marktraceur>	 I don't see any requests to any domain but simile-widgets.org, so they aren't getting it from the client at least
[22:02:49] <marktraceur>	 They might be requesting it on the backend, but I can't really tell because they obfuscate their javascript
[22:03:08] <LucianNovo>	 yea. I see thatn
[22:04:09] <marktraceur>	 Yeah, it's almost certainly a compilation that this person did
[22:04:18] <marktraceur>	 Sorry :)
[22:06:09] <LucianNovo>	 for a timeline article like the following :http://en.wikipedia.org/wiki/Timeline_of_World_War_II_(1939) what mediaWiki query should I preform to get the headlines(ie:October 1939) followed by the paragraph formatted)content underneath?
[22:06:32] <LucianNovo>	 Any advice?
[22:07:32] <LucianNovo>	 nvm
[22:10:23] <LucianNovo>	 @MarkTraceur thanks for the information. Could you tell me if there is anyway to access semantic data from wikipedia articles?
[22:13:22] <LucianNovo>	 Like what if I wanted(my program) to determine what subject the article is about(an event in history, a person, etc), what would be the best way to access that data?
[22:14:44] <Reedy>	 You'd have to analyse what metadata your given
[22:14:49] <Reedy>	 Categories, infoboxes...
[22:14:57] <Reedy>	 There's currently no abstraction of this sort of information at all
[22:15:14] <LucianNovo>	 thanks
[22:15:50] <LucianNovo>	 Are there any good methods I can run to extract this data in the (python)nltk library
[22:15:53] <LucianNovo>	 ?
[22:16:35] <yuvipanda>	 LucianNovo: tried dbpedia?
[22:16:54] <Reedy>	 There isn't really any methods.. You can use the API to extract things like links, categories, images etc
[22:17:13] <LucianNovo>	 @yuvipanda nope
[22:17:45] <LucianNovo>	 @yuvipanda looking at it now
[22:17:50] <yuvipanda>	 LucianNovo: http://dbpedia.org/About
[22:17:51] <yuvipanda>	 yeah
[22:19:25] <LucianNovo>	 this site seems to have link rot
[22:19:39] <yuvipanda>	 but it has good data
[22:19:45] <yuvipanda>	 (I was told)
[22:21:42] <LucianNovo>	 trying to find a valuable querying method ...