[00:00:34] <hoo>	 it will
[00:00:58] <ori>	 we should probably just override document.write
[00:01:04] <ori>	 to do something less destructive
[00:05:21] <hoo>	 Slap them in the face with an alert? :P
[00:10:47] <MatmaRex>	 ori: hoo: Krinkle_: for now, i sent an amendment to Krinkle's ambassadors list email. we should probably do something like this, yes.
[00:12:02] <ori>	 or just append the HTML
[00:12:17] <ori>	 basically insert it where the user probably expects it to be inserted
[00:16:08] <MatmaRex>	 document.write = function ( html ) { $( 'body' ).append( $.parseHTML( html ) ); };
[00:16:14] <MatmaRex>	 ship it~
[00:16:24] <MatmaRex>	 this is now https://phabricator.wikimedia.org/T108139 btw
[00:19:27] <ori>	 MatmaRex: submit a patch? :)
[00:20:06] <MatmaRex>	 ori: not today
[00:20:20] <ori>	 ok :)\
[00:21:53] <MatmaRex>	 actually, $.parseHTML seems to strip <script>s, so if we want that to keep working as previously, better would be just `document.write = function ( html ) { $( 'body' ).append( html ); };`
[00:30:23] <MaxSem>	 MatmaRex, rather { $('body').append('Your user scripts don't work anymore!' } ;)
[00:31:26] <MatmaRex>	 MaxSem: meh. if we can keep it working, why not? it's zero cost
[00:32:04] <MatmaRex>	 but we could just do a console.log() instead, sure. i don't really care.
[00:32:14] <MatmaRex>	 as long as it's not a blank page.
[00:39:02] <greg-g>	 Krinkle_: oddly, your breaking change annoucement when to my spam folder (max's reply didn't)
[02:10:17] <Katie>	 Hello.
[02:11:14] <Katie>	 <ori> I guess we'll never change our software, then.
[02:11:17] <Katie>	 What a shitty thing to say.
[02:13:06] <Katie>	 I don't think setting up a false dichotomy helps anyone.
[02:13:27] <Katie>	 Does anyone know why https://en.wikipedia.org/w/index.php?title=Special:Search&search=insource%3A%22document.write%22&fulltext=Search&profile=all seems to be missing thousands of results?
[02:19:30] <ori>	 Katie: troll!
[02:19:55] <Katie>	 I know you are, but what am I?
[02:20:56] <Katie>	 I really hope we didn't break the site for several thousand users without any warning.
[02:20:56] <ori>	 a self-appointed watchdog whose sense of smell is going dull, and barks at neighbors.
[02:21:12] <ori>	 no; this won't ride the train until tuesday.
[02:21:16] <ori>	 we'll have a fix by then.
[02:21:40] <Katie>	 Why'd it affect Ryulong above?
[02:21:57] <ori>	 good question!
[02:22:03] <Katie>	 Oh, I see.
[02:22:16] <Katie>	 <Krinkle> Oh, it does. Aye, because while user-js is document-written synchronously, it's execution is delayed until after the async 'site' module
[02:24:42] <Katie>	 Anyway, when the false dichotomies don't work, attacking someone is usually a safe fallback. Good work.
[02:26:59] <Katie>	 You can keep blaming the victim in your back pocket just in case. ;-)
[02:27:30] * ori hugs Katie.
[02:27:43] <ori>	 We'll fix it. Let me get some dinner first.
[02:53:50] * Katie hugs ori.
[04:07:16] <Katie>	 I filed https://phabricator.wikimedia.org/T108149 about insource.
[05:18:26] <Romaine>	 is there something recently changed in javascript? like importScriptURI ?
[05:21:07] <Romaine>	 a part of my .js is broken
[13:04:03] <Katie>	 Hi Romaine.
[13:04:12] <Katie>	 document.write(), perhaps.
[13:04:13] <Katie>	 Or sajax.
[13:04:37] <Romaine>	 thanks
[13:04:46] <Romaine>	 is solved alreadt, was bug in MediaWiki software
[18:56:44] <Pine>	 The #wikimedia-dev channel is getting lots of automated messages related to "Remove use of deprecated wfRunHooks". Is there a way to quiet that particular notification?
[19:05:47] <legoktm>	 Pine: you can put the bots on /ignore. Those are just patches being reviewed and merged.
[19:06:43] <Pine>	 thanks
[21:41:14] <ragesoss>	 are there edit filters that would screen out too-rapid edits even if it's just 3 in a row?
[22:29:54] <legoktm>	 ragesoss: it's possible to write filters that do that, yes.
[22:45:29] <ragesoss>	 legoktm: my question is more like: is there such a filter active on en.wiki?
[22:46:36] <ragesoss>	 I've got edits that are supposed to be made with my OAuth app, including one set where the user makes three related edits in immediate succession. Sometimes, the second and third have been failing for reasons I don't yet know.
[23:14:20] <legoktm>	 ragesoss: what error code is the API returning?
[23:15:11] <ragesoss>	 legoktm: I'm not sure that it's returning any, but we may not be looking for the right thing in the return value.
[23:15:32] <legoktm>	 action=edit returns errors in a different way than the rest of the API
[23:16:09] <legoktm>	 it's something like { 'result': 'Failure' }
[23:17:24] <ragesoss>	 legoktm: so it's still going to return a 200, just with failure data in the json?
[23:17:59] <legoktm>	 yes, the API doesn't use HTTP status codes for errors
[23:18:34] <ragesoss>	 cool. Thanks legoktm. I think that'll put in on the right track.
[23:19:12] <legoktm>	 ragesoss: also see https://www.mediawiki.org/wiki/API:Errors_and_warnings and https://www.mediawiki.org/wiki/API:Edit#CAPTCHAs_and_extension_errors (XML, but translates to JSON easily)