[01:02:57] spagewmf: Workable? [01:05:10] Riley, I don't know. but start by submitting a bug with what you'd like to see. [01:05:41] spagewmf, kk. Thanks [01:37:44] gn8 folks [03:06:50] Google can't crawl bugzilla even though its robots.txt allows it: https://www.google.com/search?q=site%3Abugzilla.wikimedia.org --- is that something that can be changed on our side? [03:08:48] not it doesn't [03:08:55] our robots.txt disallows basically everything [03:09:29] no even [03:10:12] paravoid: it does allow show_bug.cgi which is the relevant one: https://bugzilla.wikimedia.org/robots.txt [03:10:31] how would it find links to that though? [03:10:46] at least from mailing lists and wikis, it should [03:11:03] and then from duplicate / block / depend lists, etc [06:02:18] Riley: around? all solved? [06:02:31] (my original request was way simpler than that btw) [06:03:08] Nemo_bis: Well I don't like the change but I'm going to have to deal with it until I get consensus from ACC so everyones opinion gets heard. [06:03:26] A lot of us liked the old way because it was efficient. [06:11:50] efficient? [06:12:00] because you don't have to check another box? [06:12:10] Riley: and consensus for what? [06:12:56] 3:24:19 PM Okay, so can ACC get a custom signup page then? 3:24:26 PM ACC as in account creators [06:21:14] Riley: would it work if you could add a special url param to get the old interface back? [06:21:33] ori-l: Yes, that could work. [06:23:11] Well, that's not hard to do, since the interface modification is currently all client-side. I thought there was a magic param currently, but there isn't. Still, easy to add one. I'll write a patch now and deploy it on Thursday, ok? [06:30:15] Riley: if you're logged in and browse to http://en.wikipedia.org/w/index.php?title=Special:UserLogin&type=signup you should get the old interface [06:30:36] ori-l: Nope [06:30:39] o.o [06:31:00] hello someone with nickname similar to ine.. [06:31:15] ori-l: http://puu.sh/21iFB [06:31:36] If you could write a patch when you get a chance, that would be great. [06:31:47] Although, a custom view would still be best. [06:32:38] (if we had that option, more things than just the "by email" would be changed :P) [06:33:00] wait -- which change were you objecting to? [06:33:03] I may have misunderstood [06:33:31] ori-l: https://gerrit.wikimedia.org/r/#/c/38905/ [06:36:27] Riley: oh, I misunderstood you earlier. I could probably still help, though. So you just don't want that checkbox? [06:36:35] Yes. [06:36:56] Or have the checkbox automatically ticked. [06:37:04] One of the two. [06:42:41] Riley: would a user-script work? [06:43:10] Hm. [06:43:22] Sure [06:46:24] Riley: add this to your common.js: [06:46:27] $( function() { $( '#wpCreateaccountMail' ).prop( 'checked', true ); $( '.mw-row-password' ).hide(); } ); [06:47:14] ori-l: Testing.. [06:49:40] ori-l: Nope, doesn't work. [06:50:35] Are you sure it's not just browser cache? Can you try in a different browser? [06:51:46] Tried a different browser, doesn't work. [06:53:14] Hrm. [06:53:27] Susan, can you help me test? [06:54:55] * Susan looks. [06:56:04] ori-l: We covered this the other day. ;-) [06:56:12] Special:UserLogin doesn't load user scripts. [06:56:17] It's a security issue. [06:56:27] * ori-l facepalms. [06:56:29] Right. [06:56:39] ACC is awful. [06:56:45] It's awful in about twenty different ways. [06:56:59] Susan: uh okay.. [06:57:04] It shouldn't be encouraged, in my opinion. If the interface is now more difficult for ACC, I think that's a good thing. [06:57:21] Susan: wait, are you sure? [06:57:22] Maybe it'll finally shut down the whole bureaucracy. [06:57:27] Because it does work for me. [06:57:38] ori-l: Where'd you put the code? [06:57:40] What bureaucracy, its ACC? [06:57:52] http://en.wikipedia.org/wiki/User:Ori.livneh/common.js [06:58:13] And it loads at http://en.wikipedia.org/wiki/Special:UserLogin/signup ? [06:58:30] ori-l: You don't need that for ACC [06:58:52] Bsadowski1: Never said we *need*ed it. [06:59:04] Susan: at http://en.wikipedia.org/w/index.php?title=Special:UserLogin&type=signup [06:59:15] I'm talking to ori-l, Riley. Not you. [06:59:24] but not at http://en.wikipedia.org/wiki/Special:UserLogin/signup, oddly. [06:59:29] Sorry.. [06:59:37] Bsadowski1: don't need what? [06:59:39] ori-l: Odd. [06:59:58] ori-l: Are you using that for ACC? [07:00:07] https://test.wikipedia.org/w/index.php?title=Special:UserLogin&type=signup doesn't load JS for me. [07:00:14] But I use /monobook.js. [07:00:21] Perhaps /common.js has different behavior, dunno. [07:00:33] Bsadowski1: no; I'm just trying to help Riley w/a minor annoyance [07:00:42] oo [07:00:50] Riley: What issue? [07:01:05] Bsadowski1: …scroll up? :P [07:01:09] fine [07:01:11] * Bsadowski1 lazy [07:01:32] Oh [07:01:33] ori-l: /common.js doesn't seem to be loaded at Special:UserLogin/signup. [07:01:47] nemo's lame fixes for things that should stay the same [07:01:50] pfft [07:02:03] The old form was pretty bad. [07:02:27] Bsadowski1: don't be a jerk [07:02:34] I'm not. I'm annoyed [07:02:56] Riley: do you use Greasemonkey, by any chance? [07:03:05] No clue what it is. [07:03:12] All these changes related to group-related rights and interfaces. Like this one: https://gerrit.wikimedia.org/r/#/c/26902/ [07:03:13] We really shouldn't encourage ACC. [07:03:16] It was fine the way it was [07:03:44] Bsadowski1: I don't think anyone was discussing anything related to that in here. [07:03:54] Now non-OS-worthy things are in the suppression log [07:04:06] And for what reason? [07:04:08] ;s [07:04:20] I have to go now, Susan / MZMrBride. [07:04:22] bbiab [07:04:50] er MZMcBride* [07:05:05] Thanks for the correction. :-) [07:05:09] MrMZMcBride [07:24:46] https://bugzilla.wikimedia.org/show_bug.cgi?id=18060 was also before they introduced centralauth-suppress. Why didn't anybody comment about that? [07:26:07] VVV introduced the lock+suppress method of locking 2+ years after that ticket [07:26:16] Why was it changed now? [07:26:19] :/ [07:26:20] pfft [07:26:55] Before the only way to get rid of the name in the logs was to use a script and lock+hide and suppress the offending username [07:27:00] damn, forgot to close that one :/ [07:27:06] hoo, why? [07:27:07] Bsadowski1: maybe you should monitor bugzilla more often? Snowolf set it up so he gets cc'd for most relevant bug requests [07:27:08] :( [07:27:13] heh [07:27:16] I'm just saying [07:27:38] hoo, you also forgot to tell the rest of the stewards besides Snowolf [07:27:54] I had no idea about that "fix" until after I saw a username in there [07:28:17] Bsadowski1: Closed the bug now... [07:28:32] that change was coordinated with WMF and posted on -ambassadors [07:28:46] Where's the actual problem? [07:28:54] hoo: We were lockhiding usernames that needed to be hidden but NOT suppressed :P [07:29:10] like minorly offensive usernames not suppress-worthy [07:29:37] Bsadowski1: Yup, but as we have nothing between stewards and users this doesn't matter, does it? [07:30:05] hoo: But why are we changing logs without consulting the people that can see or are affected by it? [07:30:17] Bsadowski1: are you on the -ambassadors mailing list? [07:30:20] no lol [07:30:23] :< [07:30:24] you should be [07:30:28] Why? [07:30:28] Bsadowski1: What else shall I do? [07:30:34] thats where these "consultations" happen [07:30:39] hoo for stew :P [07:30:43] Bsadowski1: you're definitely complaining about the wrong person [07:30:55] If I post on IRC no one cares, no one's reading the mailing lists as it seems, no ones following the tracking bugs, ... [07:30:58] I know, Nemo_bis. It was another bug I think [07:31:13] Bsadowski1: which one? the other one I didn't even comment on [07:31:16] I still admire your coding skills [07:31:33] Better than me :P [07:32:01] now you're talking to hoo I suppose [07:32:07] And also you know that Wikipedians don't like changes (*coughPendingChangescough*) [07:32:08] Bsadowski1: https://lists.wikimedia.org/mailman/listinfo/wikitech-ambassadors [07:32:48] http://lists.wikimedia.org/pipermail/wikitech-ambassadors/2013-January/000129.html that was the relevant mail, btw [07:33:05] I was just looking at it heh :P [07:33:07] * Bsadowski1 should [07:33:23] Riley: btw you're thinking only about the en.wiki-only way to do things, on it.wiki most account creations are not done by email [07:33:41] Nemo_bis: Yes, en-wiki only [07:33:43] Why didn't anybody forward this to the stewards mailing list? [07:33:49] D: [07:33:50] or CC [07:33:54] Bsadowski1: did you whitelist hoo? :) [07:34:05] for accept-only? [07:34:08] Riley: what's broken is only en.wiki's process ;) [07:34:15] Bsadowski1: yes [07:34:17] Nemo_bis: True ;P [07:34:21] I have no authority over the stewards mailing list. [07:34:28] :( [07:34:43] I could ask DerHexer or another stew. to cross post things in future [07:34:44] I know, it sucks [07:34:58] Riley: anyway, frankly, I didn't like the new interface at first either, but themn had to recognize Liangent had a good idea [07:35:00] Do you people read that more than wikitech-l and wikitech-ambassadors? [07:35:08] yes [07:35:31] Riley: plus, ticking a box is hardly a super-destructive requirement and surely an aid for super-user can be found [07:35:40] * Nemo_bis slaps hoo  [07:35:46] Nemo_bis: Why? [07:35:46] super-user? [07:35:48] it's their mailing list, what else could they read ? :D [07:35:54] :P [07:36:22] Nemo_bis: I only ever once got a reply to a mailing list announce of mine... and that was an API change [07:36:51] http://lists.wikimedia.org/pipermail/wikitech-ambassadors/2013-February/000144.html no one replied to this either (both on -ambassadors and -l) [07:38:22] -ambassadors is mainly read-only [07:38:49] writing to the stewards list should work; you can also use the [[m:SN]] maybe [07:38:52] Nemo_bis: Yeah... but still there are plenty ways to contact me [07:39:21] I mostly yet posted stuff in #wikimedia-stewards and yet hardly anyone cared [07:39:24] hoo: I mean that people there don't care/know/want to comment stuff [10:00:16] gerrit has internal errors [10:00:37] Someone kick its ass? [10:00:41] :D [10:01:35] Jeblad_WMDE: On what page do you get internal errors? [10:02:16] https://gerrit.wikimedia.org/r/#/q/(project:mediawiki/extensions/Wikibase+status:open)+OR+(project:mediawiki/extensions/Wikibase+status:merged),n,z [10:02:43] But me closing down, got to get to the office [10:02:56] I'll have alook. Thanks. [10:03:24] (There were some issues with slashes in the gerrit URLs before) [10:03:36] no idea if its related, but gerrit-wm stopped reporting in #mediawiki [10:03:58] and now it just started :/ [10:04:16] :-) [10:05:43] * Jeblad_WMDE remembers some code that defaulted to "user error" when some statement couldn't be parsed [10:05:53] ..it never had "internal error" [10:05:59] XD [10:06:57] ok gerrit-wm is reporting code review, but hasnt reported either of my last two patchsets [10:07:45] because it promotes bikeshedding, not actual work [12:41:48] hi, how can I view the current source code in git/gerrit(?) like here: http://svn.wikimedia.org/viewvc/mediawiki/trunk/extensions/WikimediaMessages/WikimediaGrammarForms.php?revision=95203&view=markup ? I would like to request a new bug on Bugzilla, but I can't find current code [12:44:04] Leinad: git clone https://gerrit.wikimedia.org/r/mediawiki/extensions/WikimediaMessages [12:44:29] https://gerrit.wikimedia.org/r/gitweb?p=mediawiki%2Fextensions%2FWikimediaMaintenance.git;a=summary [12:45:29] Leinad: go to gerrit.wm , click "Projects" -> "List" to see all projects.. and from there you can click the "gitweb" links [12:47:28] you will want the blob for files, and tree for folders [12:49:16] mutante: thank you, and then I need choose on the bottom of page "master -> tree" - this is a current code? [12:49:51] yea, sounds right [12:50:01] or just clone it and see what you get [12:51:09] "clone", now it's too difficult for me ;) [12:51:26] see the first line i wrote above [12:51:35] if you just copy/paste that to a shell [12:51:48] you should get all the current files and you can look at them with a text editor [12:52:14] ahh, shell, ok! :) [12:52:47] yea, whatever seems easier to you. i would prefer that over gitweb..but up to you [14:12:48] http://video.fosdem.org/2013/lightningtalks/How_to_hack_on_Wikipedia.webm [14:16:28] mutante: we need the same t-shirt but s/Edit/Hack/ [14:18:00] hashar: http://www.spreadshirt.com/ has a browser-based shirt designer [14:18:20] http://www.spreadshirt.com/design-your-own-t-shirt-C59 [14:21:49] maybe I should order a bunch [14:22:04] and distribute them to folks in SF ( I wrote a PHPUnit test! ) [14:22:39] hah, just ask before using the trademarked logos i guess :P:) [14:23:06] yeah [14:23:15] I would add a {{fair-use}} at the bottom of the shirt :-D [14:23:25] hehee, nice [14:23:50] btw, we need a free font [14:24:05] cant just download and install it [15:39:22] Hi, I'm new to gerrit / wikimedia-gerrit. Why i get a "Internal Server Error"-Javascript overlay when im trying to type a char in the add-reviewer field? [15:46:00] se4598: Known issue due to last nights upgrade [16:00:44] wikibits is giving 503 varnish errors. Guru Meditation: XID: 2038761257 [16:01:16] URL? [16:01:19] Where are you in the world? [16:01:29] wizardist: Bits is indeed slow now.. [16:01:40] for me sometimes [16:01:44] Reedy: Belarus. url: https://bits.wikimedia.org/be-x-old.wikipedia.org/load.php?debug=false&lang=be-tarask&modules=ext.vector.collapsibleNav%2CcollapsibleTabs&skin=vector&version=20130212T155330Z& [16:02:06] Wiki13: ok, just wanted to make sure that the staff is aware [16:02:09] Hmm, confirmed [16:02:46] same thing is happening right here in the Netherlands sometimes [16:03:12] Yeah, I presume belarus would hit esams [16:03:50] Reedy: no, it hits on equiad [16:03:56] because of google dns. [16:03:58] CSS is being stripped from pages on wikimania wiki too [16:04:00] heh [16:04:03] local provider's dns is hitting on esams [16:04:11] and fails as well :) [16:04:26] http://ganglia.wikimedia.org/latest/?r=hour&cs=&ce=&s=by+name&c=Bits%2520application%2520servers%2520eqiad&tab=m&vn= [16:04:27] *one moment* (maybe chrome's dns caching) [16:05:09] yeah, both eqiad and esams are affected as for me [16:05:59] They both use the same apaches [16:06:09] for the record: sometimes it makes to serving a page, sometimes it reports 503 error. [16:06:21] obviously [16:06:31] I've notified ops [16:08:51] :( [16:16:39] yay nostalgia [16:17:04] Nemo_bis: I assume you're talking about the lack of Vector on en.wp? [16:23:49] js and css issues + slugginshness [16:25:24] I'm not sure about you, but its loading superfast for me, just without the pretties [16:28:12] yeah - various reports in other channels about no css/js - but speed seems fine actually [16:28:21] <^demon> Maybe we should just get rid of all the pretty modern skinning. Wiki will be way faster. [16:28:36] <^demon> And we can remove so much code :) [16:28:51] I like where this is going [16:29:07] carrier fonts - DOS prompt - perfect [16:29:12] not pt.wiktionary. it's way slow [16:29:23] <^demon> varnent: Hey, ugly & simple has worked for Craigslist forever ;-) [16:29:30] commons and wikimania sites seem okay on speed [16:29:32] XD [16:29:33] heh [16:29:47] and reddit :) [16:30:00] ^demon: I'll admit - I'm biased towards Apple's simple and clean :) [16:30:14] but I do miss DOS sometimes..terminal holds me over..but it's just not the same.. [16:30:17] there was some article recently that said it was a good thing for community sites like those to be a bit ugly and old [16:30:46] if those same rules apply to their participants - then I'm all set [16:30:59] * varnent rimshot [16:31:06] <^demon> heh [16:31:42] you always know there's trouble when the number of ppl in this room suddenly increases [16:32:01] and what is the reason for those css and js issues? live update? [16:32:26] go the other way, send raw wiki text to the client always, and use a javascript parser to turn it into html and stylize the page [16:33:26] * varnent envisions a scientist in a nearby bush observing that in the same way dogs fleeing signals a storm - developers gathering here signals a crash [16:36:41] <^demon> malafaya: Not sure yet, ops is still digging. [16:36:48] <^demon> Not a code update though, no. [16:37:02] ok, i thought it was something predictable [16:39:31] today is the Wikidata-on-en-wp rollout and the mobile watchlists rollout [16:40:17] sumanah: you think those could be the cause - or maybe it's the activation process itself causing side effects? [16:40:55] I don't know. [16:41:11] <^demon> MobileFrontend doesn't deploy until afternoon PST. [16:41:14] <^demon> Hasn't happened yet. [16:44:09] argh, I'm seeing JS errors in CN, breaking the sites badly [16:45:03] ^demon: is someone working on the gitweb->gitblit rewrite rules already? [16:45:51] <^demon> Not yet. gitweb is still the primary git viewer, I haven't swapped it yet. [16:45:58] <^demon> (I saw at least one bug we'll probably want to fix) [16:48:03] hey folks; I see in the topic you're already aware of the site issues. I'm having computer trouble at the moment so IRC is difficult, but if there's a need to send a message to social media channels I can do it [16:51:09] Nikerabbit: l10n-bot has 30 % of the repo activity :D https://gerrit.wikimedia.org/r/plugins/gitblit/activity/ [16:53:16] ^demon: is the bug that submodules aren't clickable in tree view? :) https://gerrit.wikimedia.org/r/plugins/gitblit/tree/?r=mediawiki/extensions.git [16:54:43] <^demon> Ah, we'll need to set git.submoduleUrlPatterns for gitblit. [16:55:46] <^demon> Oh, nvm, that's not it. [16:57:02] <^demon> "Submodule references in a working copy will be properly identified as gitlinks, but Gitblit will not traverse into the working copy submodule repository." [16:57:06] <^demon> Blah, that sucks. [16:57:46] who's handling the broken bits thing? [16:58:22] in case it helps, I'm still getting style-less pages after clearing the cache [16:58:25] well if by "handling" you mean looking at it [16:58:31] jeff, me a little, roan a little [16:58:37] and we're trying to get mark [16:58:54] ok; just checking someone was still on it :) thanks [16:58:55] I'm just investigating at this point [16:59:01] It looks like memcached is messed up [16:59:15] RoanKattouw: checking the CSS issue? [16:59:32] Yes [16:59:43] cool, glad someone is on it [16:59:57] Hah, we only have 16 memcached boxes now? [17:00:12] Seems like Varnish is giving error 503 [17:00:34] it seems that only logged-in users have issues; I can load styled pages logged out [17:01:07] so I'm not going to send tweets about this just yet [17:02:43] gpaumier: Thank you for this observation, that may lead us to the cause [17:09:16] RoanKattouw: logged-in only is false. the behaviour for an anonymous client is as random (and 503-penetrated) as for logged-in [17:10:22] fixed? [17:10:30] maybe [17:11:03] maybe [17:11:56] 20s response. [17:12:15] yet better than it was [17:12:21] It seems the Varnish caches are mostly papering over this for now [17:12:27] Still seeing 503s, but only for user JS [17:12:46] Ironically the caches (Varnish) are rescuing us from havoc wreaked by the caches (memcached) [17:14:30] what did I miss? [17:14:51] yes, I also had problems with styles and scripts. [17:15:59] still happening on some wikis for me [17:16:10] still working on it [17:19:13] uhm, for me CSS is fixed but I don't have personal CSS or JS [17:30:51] just to chime in, still seeing 503's on bits.wikimedia.org when loading CSS [17:36:42] Are the bits 503s going away yet? [17:37:02] TIAS? [17:37:38] bit's is down for me.. in portugal-> europe [17:38:04] Alchimista: still have nostalgia skin here :o [17:39:18] aude: i was tryng to get my watched pages, so it's quite a nostalgic skin, indeed [17:39:46] wow, en.wp main page is quite awsome [17:39:47] right [17:44:38] still seeing them now [17:45:46] Still 503ing? [17:46:14] I'm getting all 200s in the US [17:46:18] Maybe it's EU-specific? [17:46:56] yeah I am having issues also [17:47:08] (with javascript and style sheets) [17:47:46] i'm in the US [17:47:52] intermittent though [17:47:58] Hmm no I'm hitting EU bits [17:48:32] US bits works for me as well [17:51:00] everything seems fine here [17:51:05] no css or js errors [17:52:20] Poland the same [17:52:27] (no errors) [17:53:24] Does anyone still have broken CSS? [17:56:39] (oh, to add, "here" is Germany, as well as per proxy (asking my wife) in Estonia) [18:03:38] appears fixed for me now [18:17:04] looks good RoanKattouw [18:17:11] Good [19:00:33] FYI, deploying WikibaseClient to enwiki in a minute or 2 [19:01:30] Fasten your seatbelts! [19:01:34] \o/ [19:01:43] hoo! [19:02:06] * hoo :D [19:09:03] seems working at the first sight [19:17:51] Hi, I'm an admin at es.wikivoyage. We want to change the look of the table of contents in that wiki, now it is a floating table, and it should be a normal one, with the text only over and above the TOC like in the other Wikimeadia projects, not around it. You could see http://es.wikivoyage.org/wiki/Lyon as an example of how is now configured the TOC. I can't resquet it in bugzilla because I don't have an account there. [19:18:42] Could someone please help me to know how to change that or resquet it at bugzilla? [19:19:32] wikiviajero5474: Looks like es.wikivoyage is using a custom TOC extension. [19:19:50] https://www.mediawiki.org/wiki/Extension:TocTree [19:20:23] That value can only be changed via bugzilla, isn't it? [19:44:27] wikiviajero5474: it should be possible to change the style of the toc using MediaWiki:common.css [19:44:40] it's also quite possible that the "floating" stuff comes from there in the first place [19:45:16] We tried but it didn't work [19:46:31] wikiviajero5474: even using an "! important" modifier in the css rule? [19:46:44] did someone look which css rules apply? [19:47:18] I don't know how to use the "important" key [19:47:26] I'll show you the diff [19:47:56] wikiviajero5474: tocFloat seems to be the class in question. [19:48:20] .tocFloat { float: none ! important; } [19:48:37] !important will override other css rules and even inline style attributes [19:48:37] !important is a keyword in CSS to override conflicting styles [19:48:45] oh shutup mw-bot# [19:48:59] heh [19:49:06] That's the diff of our test: http://es.wikivoyage.org/w/index.php?title=MediaWiki:Common.css&diff=prev&oldid=60668 [19:49:11] !bequiet wm-bot [19:49:11] at least it was accurate :P [19:49:40] What should it be? [19:49:47] wikiviajero5474: that's wrong on many levels :) [19:49:54] http://www.w3.org/TR/CSS2/cascade.html helped me a lot to understand such stuff [19:50:44] wikiviajero5474: toc is a class, not an id. and you want tocFloat. and you need to set the float attribute, not allign. try .tocFloat { float: none ! important; } [19:51:29] oh, sorry, toc is also an ID. but still better to use the class [19:51:45] I suppose it would be nicer to tell to code not to add the tocFloat class, instead of redefining it. [19:51:46] #toc.tocFloat { float: none; } since #toc.float is more specific as .float afaik [19:51:58] saper: yea, that would work [19:52:01] I'll try in a minute [19:52:06] but with !important, it's not needed [19:52:43] hm, firebug doesn't show any rules being applied to #toc or .toc or .tocFloat. [19:52:51] mut be missing something, because it does float, after all :) [19:53:40] by the way, http://www.openstreetmap.org/openlayers/OpenStreetMap.js is loaded via http not https [19:53:55] DanielK_WMDE: chrome shows me #toc.tocFloat { float: left; } [19:54:12] margin: 0 2em 1em 0; [19:54:13] max-width: 20em; [19:54:14] also [19:54:21] saper: yea, i kind of expected that :) [19:54:28] where does it come from, though? [19:55:01] * DanielK_WMDE thinks tocFloat is a sucky class. It's a toc. the css gets to decide whether it floats. [19:56:16] it say "Lyon:1" which is ... fake :) [19:56:26] must have been dynamically added via js or something? [19:57:04] ugh. possible. [19:57:22] it's a user preference [19:57:25] I tried and it didn't work as expected [19:57:25] Maybe we should steal mobiles deployment window ;) [19:57:26] ah, now firebug sees it too. odd. maybe because it's dynamic?... [19:57:51] wikiviajero5474: works when i try it directly in the browser. [19:58:06] yes it's a user preference [19:58:11] wikiviajero5474: but if it's a user preference, as saper said, then you should probably not mess with it. [19:58:20] maybe file a request to change the default [19:59:27] Where? At bugzilla? [19:59:32] but why? [20:00:35] https://es.wikivoyage.org/w/index.php?title=Especial:Preferencias&uselang=en#mw-prefsection-misc "Floated table of contents" [20:01:52] Thanks, I see [20:02:20] I think it's a vector thing [20:02:35] I think something worked!! [20:02:58] o.O /me is saddened, no haves CSS... [20:03:30] ah no [20:03:33] it's some extension [20:03:45] I erased cache after I changed the CSS and nothing worked [20:04:01] But some minutes afted it just works [20:04:09] TocTree? [20:04:25] wikiviajero5474: you should ask for " $wgDefaultUserOptions ['toc-floated'] = false; [20:04:31] in Bugzilla really [20:05:16] That's what I did https://es.wikivoyage.org/w/index.php?title=MediaWiki:Common.css&diff=prev&oldid=61632 and unless another user have done another thing at the same time it worked [20:05:58] wikiviajero5474: https://www.mediawiki.org/wiki/Extension:TocTree read thi [20:05:59] this [20:07:42] looks like Toctree is there by default for all wikivoyages, why? [20:07:55] ftp://5.228.40.61/qq/diff-with-itself.png (attention to diff= and oldid= values). I never saw such amazing things before in Special:Watchlist of English WP. [20:08:14] saper: i suppose because it was there on wikitravel. [20:09:04] qq[IrcCity]: that can be caused by wikidata [20:09:48] qq[IrcCity]: wikidata will insert notifications about changes to the data item on wikidata.org into recentchanges/watchlist table on en.wikipedia.org [20:10:04] of course, no edit took place there, so there is nothing to diff [20:10:20] no edit? thanks. [20:10:21] with th extension enabled, the diff link would be replaced with a link to the diff on wikidata.org [20:10:41] qq[IrcCity]: check the revision history, there should be nothing there for the respective timestamp. [20:10:45] aude: ---^ [20:11:03] aude: wikidata entries in recentchanges now get rendered in a strange way. can we just drop them? [20:11:20] what [20:11:31] on enwiki? [20:11:54] link? [20:12:42] https://test2.wikipedia.org/w/index.php?title=Special:RecentChanges&hidewikidata=0 looks fine to me [20:14:20] aude: the relevant entries would be further in the past, i suppose [20:14:51] qq[IrcCity]: can you give us the timestamp of the entry you saw for [[Transcomputational_problem]]? [20:14:55] hmm [20:15:04] ah, that's one i edited? [20:15:21] nah, mine was different [20:16:29] dunno, just saw the screen shot. maybe i misinterpreted [20:16:41] the thing is: the entries will have a nonsensical diff link, confusing people [20:16:56] nothing worse than that, i think [20:17:12] but still would be nice to just drop the disfunctional entries from recentchanges [20:17:20] sure [20:18:04] DanielK_WMDE, the timestamp is (today's) 07:25 (UTC). it is visible at the screenshot. [20:18:32] ah, i see [20:18:35] where's the screenshot? [20:18:41] we can delete those entries [20:19:44] ok, i do see entries in my recent changes [20:20:01] add http://en.wikipedia.org/wiki/Szklarka_Trzcielska to your watchlist :o [20:20:59] aude: there seems to be no way to unhide wikidata changes on Special:RecentChangesLinked [20:21:04] do we have a bug for that? [20:21:09] oh [20:21:18] it uses the same code as special:recentchanges [20:21:21] it just extends it [20:21:25] but could be a bug [20:21:31] i suspect so [20:21:52] https://it.wikipedia.org/wiki/Speciale:ModificheCorrelate/New_York [20:21:54] looks fine [20:22:00] has the toggle [20:22:30] it should just work, since related changes extends the recent changes special page class [20:23:04] * aude tries editing a related wikidata item [20:24:15] aude: duh! sorry. i was looking on enwiki. which of course doesn't have the extension enabled. silly me :) [20:24:43] right [20:24:50] * DanielK_WMDE should probably call it a day [20:24:58] if Reedy or ^demon can remove the entries, that would be nice [20:25:03] or binasher [20:25:25] it's weird to see entries in my watchlist when there is no wikibase client [20:25:49] aude: what's the exact condition? it's rc_type = .... what number? [20:25:50] What's the easiest way to select them? [20:25:55] rc_type = 5 [20:25:56] ^ [20:26:01] that seems simple :) [20:26:19] 11439 rows [20:26:28] sonds about right [20:26:30] doable [20:26:42] correlates with the job queue number approximately [20:26:45] yup, looks sane [20:26:48] not all the jobs = rc entry [20:26:57] aude: yea, though the stuff still in the queue shouldn't be in rc. [20:27:04] yep [20:27:09] mysql:wikiadmin@db1017 [enwiki]> delete from recentchanges where rc_type = 5; [20:27:09] Query OK, 11439 rows affected (7.59 sec) [20:27:12] yaya [20:27:14] yay [20:27:28] looks better [20:28:32] Reedy: thanks [20:57:42] apergos: ping [21:26:01] Vito: pong but nt really here [21:28:15] what is the link to wikimedia mrtg site? [21:30:04] apergos: pm [21:38:42] matanya: mrtg? I don't think we are still using it [21:55:05] matanya: Do you mean http://torrus.wikimedia.org/torrus ? [21:55:23] yes thanks Reedy [22:00:55] Was a database-user "metrics" added to the cluster short time ago (I see nothing in the server-log about) [22:01:03] ? [22:03:42] notpeter: ^^ [22:04:13] DaBPunkt: yep [22:04:16] sorry, should have logged [22:04:26] it was in rt! :) [22:05:17] notpeter: ok, no problem. [23:00:46] gn8 folks