[00:46:02] lists.wikimedia.org is taking forever. [00:46:13] And giving me "no data received" errors when I try to clear a moderation queue. [00:46:16] Rage. [00:46:42] > Summary of junk emails blocked - 1 Junk Emails Blocked [00:46:44] Is that legit? [00:46:57] http://p.defau.lt/?AGy5dDfqkic8sSfLPgHxjQ [00:47:00] Those are the headers. [01:21:47] greg-g: Around? [01:23:48] hi, tools.wmflabs.org is down? [01:31:00] I think it is. [02:02:15] greg-g: http://lists.wikimedia.org/pipermail/wikitech-l/2013-August/071536.html [02:02:33] greg-g: The "Any help..." sentence was directed at you. ;-) [04:31:18] Elsie: on it. [04:31:23] :-) [04:31:38] I'm not sure how realistic the concern about VisualEditor confusion is. [04:31:51] me neither.... I was kind of surprised to see that line [04:31:57] I thought it was a jab at VE ;) [04:32:47] Well, [04:32:52] to someone who doesn't know the difference, [04:32:58] and who hates VE and thought it was disabled, [04:33:02] I could see the potential confusion. [04:33:07] true [04:33:11] Plus the user preferences are accessed in different ways. [04:33:28] yeah, confusing [04:33:34] A SINGLE EDITOR TO RULE THEM ALL! [04:38:34] Elsie: are you subscribed to the -ambassadors list? [04:40:19] Nope. [04:40:54] Elsie: k, mind if I cc you on my message there so they can hunt you down, I mean, ping you with questions? :) [04:41:18] Questions should go to [[m:Talk:CodeEditor]], I imagine. [04:41:20] Or [[m:Tech]]. [04:41:29] Though if people want to e-mail me, I can tell them that myself. ;-) [04:41:52] heh, fair [04:53:08] (sent) [05:28:12] http://lists.wikimedia.org/pipermail/wikitech-ambassadors/2013-August/000361.html [05:34:01] https://meta.wikimedia.org/wiki/Talk:CodeEditor#Legacy_toolbar [05:34:03] sDrewthedoff: ^ [05:43:33] ok, g'night [08:04:29] @notify qchris [08:04:29] I'll let you know when I see qchris around here [15:06:33] parentxyzq.. nope [15:06:56] no parent yet [15:07:05] coming now [15:07:12] ok [15:08:43] maybe we can start by going through the things that i did while you were away? (parent already knows those) [15:09:02] I did look at the backread here [15:09:10] ok [15:09:11] and also a bit at the diff dump specs [15:09:25] parent5446: hello [15:09:29] hello [15:09:31] Hey [15:09:47] so I have a comment about the diff dump specs actually [15:09:59] i'm listening [15:10:08] for the delte page change [15:10:18] This object describes a page that was deleted from the wiki. All of its revisions should be deleted too, unless they were already referenced in this diff dump (this can happen when a page is deleted and then undeleted). [15:10:31] why dn't we do that work ourselves so that the user app doesn't have to think about it? [15:10:53] that is, only write out page deletions that weren't undone by a later undelete [15:11:16] that might require another pass or some fiddling but it seems cleaner to me from the user's point of view [15:12:00] except undoing delete means creating a page with new id; are you saying there should be a change that changes the page id of a page? [15:12:25] I'm saying that instead of 'keep the revisions under this special circumstance' [15:12:38] we don't make any third party app that consumes these dumps have to do that work [15:13:16] if there's a page delete in the dump, it should be an actual 'delete everything' [15:13:25] without 'I have to check this other thing first' [15:13:34] see what I mean? [15:14:55] does that mean the the diff would have to contain all the undeleted revisions in full? [15:15:35] I don't know how you would want to do it [15:16:21] because if i delete all those revisions, i have to recrease them for the undeleted page somehow [15:16:26] *recreate [15:16:39] that's true [15:17:18] and if somebody deletes-undeletes a page with long history, i don't want to include all those revisions in the diff, that's wasteful [15:17:23] so let's think about this, can we actually see a delete and then an undelete in the same dump run? [15:17:34] certainly [15:17:48] page move [15:17:59] which isn't an undelete but it's the equiv [15:18:10] no, page move is simple, that changes page title, but not page id [15:18:22] right, the redir gets the new id [15:18:27] yes [15:18:47] sorry, I'm just not seeing where we would get an undelete after a delete in the same run [15:20:21] 1: create dump; 2: delete page; 3: undelete page; 4: create another dump; i think such situation can happen (especially in the case of edit warring) [15:20:21] talk me through it if you don't mind [15:20:37] so 4 happens after both a delete and an undelete [15:20:42] yes [15:21:12] we now have a new page id, the absence of the old page id, which is therefore marked by you as page deletion [15:21:46] yes [15:22:14] gah I am so used to the dumps from full stubs [15:23:35] so do you now think that my solution is reasonable? or do you have a better idea? [15:24:01] a 'page id change' marker? [15:24:29] right, but that would be hard to detect [15:24:34] or to make the user worry about it? [15:24:58] what do you mean? [15:25:43] well anyone who writes an app that reads these diffs (and there will be folks, as I judge already from users even of the 'experimental' adds/changes dumps) [15:26:03] will have to implement the same icky workaround for page deletes [15:26:11] when really we should take care of it on our end [15:27:26] so, how would we detect the "page id change"? by page title? by list of revisions? [15:27:39] well if you don't do page id change [15:27:55] which is fine, I'm not stuck on a particular way of getting it done [15:28:08] you could mark some changes as page + revision deletes [15:28:20] and some as page deletes (just the page row in the table) [15:28:59] and yes, pages with new page ids will get the full list of revisions written out [15:29:18] but presumably there aren't mass page undeletions generally [15:29:39] so any given diff dump under normal circumstances would still be a reasonable size [15:29:42] that's what i do now too, i think that's necessary, because the list of revisions can change (you can undelete only some of the revisions) [15:29:50] yep [15:31:39] i have to think about how exactly to distinguish the full delete vs. partial delete, but i think this will work [15:32:15] ok. if ou come up with another approach that's fine too, as long as we do it on our side [15:32:18] *if you [15:33:10] so 'delete revision change' [15:33:25] if someone suppresses (oversights) an edit [15:33:42] well, another approach would be to have only partial page deletes and then always include the full list of deleted revisions, but the approach you suggested is better [15:33:50] won't that look like a deleted revision from our perspective? [15:34:40] AFAIK, you can't suppress an edit, you can suppress only its text, comment and/or contributor; and that doesn't look like deleted revision [15:34:58] I need to check into that and make sure [15:35:36] that's what https://www.mediawiki.org/wiki/Manual:RevisionDelete says [15:37:54] but if a revision is actually deleted (e.g. by directly manipulating the database), then i should handle that properly using the deleted revision change [15:39:23] yeah, this is more a comment on the wording, I'll check on this and get back to you [15:39:30] ok [15:41:54] that was all I had on a first skim of the diff dump format [15:42:17] does the sha1 really go in as base 36 encoded? :-D [15:43:32] yeah, i didn't feel like writing the conversion code that at the time (36 is not a nice number); but i do have a TODO for that in the code :-) [15:43:40] ok :-) [15:43:52] I mean stuff like that is not tragic, it's just 'ok, clean this up later' [15:44:07] yeah, that's the plan [15:44:27] so using zdelta is next up? [15:45:30] yeah; i have already started a bit on that; my plan now is to write several versions of compression, to compare their sizes and speeds [15:46:05] (i will write the code just to test it out, it won't support updates at all at first) [15:46:08] right [15:46:32] and once again.. ah urgh [15:46:51] so I am going to be gone from tomorrow evening? (or friday morning) thrugh next monday, but strictly next monday this time, [15:47:12] I'm not going to be stuck on an island where the ferry schedule is once a day and the bus is once a day also and they fill up [15:47:35] anyways I was gonna say, for large benchmarks I cna do testing, which is true if I'm here (i.e. on piece X of en wp) [15:47:50] Also I'm not gonna be here for Thursday's meeting (I have another school meeting to go to) [15:47:56] ok [15:48:00] hmm maybe [15:48:09] I will ask if we cna move tomorrow's meeting up by [15:48:10] * apergos thinks [15:48:22] 90 minutes? [15:48:43] Would that be in the future or past? [15:48:46] this would permit me to possibly take a train tomorrow that gets in at 10 pm instead of 4 am (my friend will appreciate it more) [15:48:56] fine by me [15:49:00] ok [15:49:05] just for tomorrow [15:49:06] 90 minutes forward or backward? [15:49:13] 90 minutes earlier [15:49:33] OK. If I can manage to wake up I'll be there. [15:50:01] Actually, I'd still have to shower for my next meeting, so I probably can't make it. [15:50:02] ah, cool [15:50:05] oh :-D [15:50:18] ok well if you're there you're there [15:50:19] no worries [15:51:44] if that's all, see you tomorrow [15:51:47] so I said all I had to ay, anyone else? [15:51:49] *sya [15:51:53] grr *say [15:52:31] nothing from me [15:52:42] Nope [15:53:00] ok, then see folks tomorrow [15:53:06] See you Friday [15:53:07] and thanks for being flexible about the time [15:53:42] bye [16:41:39] qchris: good news: gerrit can also update with full instead of relative urls! :-) https://github.com/wikimedia/pywikibot-core/commit/e0838c4000ee45115ba9b248311bdbc21133a36c ; based on https://github.com/wikimedia/pywikibot-core/blob/master/.gitmodules [16:42:09] valhallasw: Cool! [16:42:52] valhallasw: but does that not force to https then? [16:43:06] valhallasw: The other should be protocol neutral. [16:43:18] valhallasw: (It least that's how I understood it) [16:44:21] qchris: hmm. Maybe // would also work, but at least this also works when someone clones from github [16:44:49] valhallasw: Meh. Github is evil anyways :-D [16:45:10] valhallasw: But you're right the ../.. breaks for github. [16:45:13] I'm not even sure whether gerrit supports cloning over http instead of https, to be honest [16:45:31] oh, wait, ssh vs https, not http vs https [16:45:35] The use case is ssh for me. [16:45:38] Yes. [16:45:49] there is a HTTP password method [16:45:57] should not be a huge issue because git review will setup ssh:// anyway [16:46:36] saper: Yes, there is http password ... but http is not ssh :-/ [16:47:10] git-review uses ssh and scp, so http is no go [16:47:28] But anyways, ^demon uses absoluted paths for the extensions parent repo. [16:47:38] So I guess that should be fine. [16:47:51] saper: git-review *should* use whatever the "gerrit" remote points at, IIRC [16:51:58] marktraceur: yes, and it sets this up from .gitreview, not from the current origin [16:52:16] valhallasw: Yeah, but you can manually override it after the setup step [16:56:56] marktraceur: there are certain subtle issues when using HTTP, let me find the bug [17:06:00] marktraceur: git review does "ssh gerrit query --json <>" [17:06:12] Oh my god. [17:06:55] for that reason one can't use git review to contribute to gerrit itself (or android:) [17:07:52] yes, there are two JSON queries and "ls-projects" is used sometimes as well [17:08:01] plus scp to fetch the hook [17:09:04] it became hell of a monster script :( [19:05:40] Hey, um.. you know how when you click a video, it expands to fill more of the screen? [19:05:44] When was that added? [19:15:05] AdamCuerden: I forget... relatively recently, like within the last month or so [19:15:24] marktraceur: Maybe you know this? ---^^ [19:16:30] Er [19:16:35] Few months ago, I think [19:16:46] IIRC kaldari worked on it [19:17:03] * marktraceur looks for documentation [19:17:31] probably not documented :( [19:17:47] but I can tell you about it :) [19:17:48] Ach. Just I'd like to include it in the Wikipedia Signpost [19:17:53] RoanKattouw: I'm trying to get the # of watchers of some of my templates/pages with the API, and I'm having to do multiple API calls for each page. [19:18:17] ...Eh, screw it. I'm slipping it in, and they can complain if they like. Also including that the gallery tool changes are live. [19:18:34] First call is to get page id, second call is to get watchers because page id is in path to watchers. [19:18:39] AdamCuerden: I think it may have been mentioned about a month back, but I can't remember specifically [19:18:44] Is there an easier way? [19:18:57] T13: How do you get watchers, exactly? [19:19:03] ++bawolff re gallery changes [19:19:38] AdamCuerden: In the multimedia theme, Commons got brand-spankin'-new Campaigns infrastructure! CC YuviPanda. [19:20:32] I'm using a jsonp ajax call and the path is something like result.query.page.(page id number).watchers [19:20:52] I see [19:21:00] Oh, you may be interested in ?indexpageids=1 [19:21:06] And... I have no idea what a Campaigns infrastructure is [19:21:06] I'm on droid do link is hard to get [19:21:19] Sorry! [19:21:19] What's that do? [19:21:28] It adds an array of all page IDs that are in the response [19:21:40] I'm still near the bottom of the learning curve =) [19:22:23] So you can do something like result.query.pages[result.query.pageids[0]].watchers [19:22:46] AdamCuerden: Campaigns are like Wiki Loves Monuments, it's just a way to organize people to help drive uploads [19:22:54] I'll play with that on [[Special:ApiSandbox]] [19:23:00] You can auto-fill details for them, group stuff in a category, that sort of thing [19:23:10] Yeah, and so we had a campaign for Wikimania pictures as well [19:23:24] With like the location and categorization pre-filled, only one license option, etc. [19:23:25] It's all stored in JSON now, thanks to YuviPanda, and there are major prettification patches in the pipeline [19:23:33] Oh, neat! [19:23:51] That deserves inclusion [19:24:14] Is there an information page about it I can link to? [19:24:20] Hrm [19:24:26] Yes, YuviPanda has a user page about it IIRC [19:24:35] AdamCuerden: https://www.mediawiki.org/wiki/User:Yuvipanda/Campaigns_namespace_proposal [19:25:05] I think he's in the middle of phase 4 now [19:27:27] Hmm. Are a couple issues with the video viewer technology. It's very hard to get from there to the page information that documents it [19:27:42] true [19:28:13] I mean, I can get the author and date easily enough, but if I try to click on them, nothing happens [19:28:17] the new multimedia team is working on a totally new media viewer that will include metadata in the viewing window under the video/image [19:28:37] AdamCuerden: Oh really? That sounds like a bug [19:29:34] yeah, looks like something's broken [19:29:38] it was working before [19:29:43] Well, crap. [19:30:19] Sorry to be the bearer of bad news. [19:31:10] AdamCuerden: That's really weird. The links are there but don't work. There must be some JS hijacking the click action on them. [19:31:47] Odd. [19:32:45] AdamCuerden: I'll file a bug on it. Thanks for the report [19:33:05] No worries. [19:33:33] As I said, I promised to work on the Tech news for En-Wiki's Wikipedia Signpost, so I have to look into a lot of things. [19:33:47] Right. I'm afraid I have a few more questions. [19:34:02] Does anyne ave the documenttion on the new mode types for Galleries? [19:34:47] And can alnyone tell me about the new search's improvements over the old search? [19:35:03] AdamCuerden: the infrastructure isn't a giant piece of crap? [19:35:07] :) [19:35:47] Heh [19:35:48] AdamCuerden: what Ryan_Lane is basically the gist of it, really :) [19:36:02] So, what will users notice about it? [19:36:08] the be more politically correct: the new infrastructure properly scales horizontally [19:36:09] If anything? [19:36:13] it will fail less often [19:36:30] aka "more reliable" to be positive sounding :) [19:36:31] I'm not sure if the current implementation has any added features [19:36:35] not yet [19:36:49] so, yeah. more reliable :) [19:37:12] Does it also make it more efficient? [19:37:12] well, except for templates, those are search better or something :) [19:37:20] AdamCuerden: https://bugzilla.wikimedia.org/show_bug.cgi?id=53493 [19:37:28] searched* [19:37:30] AIUI it'll also allow us to build a lot of new cool stuff on top of it [19:37:36] But that's for later [19:37:46] right [19:37:50] expanded template search [19:37:59] You need to get the foundation in order before you do that, and that's what's happening now [19:38:00] search in categories (I wondeer if intersection of categories too) [19:38:20] AdamCuerden: If you're looking for tech stories for the SignPost, we just launched user notifications on the mobile interface for all logged-in users. [19:38:21] Can I quote on that? [19:38:26] really excited about expanded template search, it means search for translations on a pile of wiktionaries suddenly returns something useful [19:38:31] Already in, Kaldari [19:38:32] (once it's live) [19:38:37] cool [19:38:47] Here's what I have so far [19:38:58] You'll notice one of them is not yet in English [19:38:59] * Videos expand to fill more of the screen when viewed, providing a better viewing experience. Try it out! [19:38:59] * Mobile team [[WP:Echo]] http://lists.wikimedia.org/pipermail/wikitech-ambassadors/2013-August/000357.html [[Wikipedia:Notifications]] [19:38:59] * Commons gains a new Campaigns infrastructure, letting projects like Wikipedia Loves Monuments easily preformat information, and document uploads related to them. Details are [[:mw:User:Yuvipanda/Campaigns_namespace_proposal|here]]. [19:39:00] * For those who work with CSS, javascript, and gadgets, [[:meta:CodeEditor|]] will likely prove a boon. [19:39:30] Galleries will be the lead story [19:39:43] @link [19:39:43] http://enwp.org/:meta:CodeEditor [19:39:49] oh you mean, not a full grammatically correct english sentence [19:40:02] Ha.. another bug.. [19:40:41] AdamCuerden: You can mention that the pop-up video player on en.wiki is a trial. If people like it, it will be turned on for other wikis as well. [19:41:04] That last one should read "* Improvements to the search infrastructure promise to make it more reliable. Now that the backend is fixed, new features, like the ability to search within a category, are planned. " [19:41:34] AdamCuerden: The actual threshold for the pop-up player is any video embedded at a width of 800px or less, if the width is more than 800px it still plays inline. [19:42:19] off to lunch... [19:42:36] * Videos thumbnailed at less than 800 pixels wide expand to fill more of the screen when viewed, providing a better viewing experience. Try it out! This is currently exclusive to English Wikipedia, should things go well, it will soon appear on other Wikipedias. [19:43:06] The echo notifications are just notes as to what to link just now [19:43:28] I'm trying to collect things together. [19:46:14] "* The mobile team has [http://lists.wikimedia.org/pipermail/wikitech-ambassadors/2013-August/000357.html implemented] [[Wikipedia:Notifications|notifications]] on all [[WP:Echo|Echo]]-enabled wikis. This means that mobile users will get the useful communication features such as talk page notifications and mentions that users of the desktop site have." [19:47:03] Sorry to be live-blogging here. It's just far easier to get this right when I know if I'm wrong =) [19:47:42] Anything else anyone would like announced at this time? [19:49:07] looks good [19:49:26] has anyone else clicked on links lately and wound up at the AbuseFilter [19:49:27] ? [19:49:34] pretty sure this has happened to me twice [19:50:27] Right. I'll need to check the VisualEditor news still, and work out the section on gallery improvements, and check with Gerard for Wikidata news... [19:51:11] But I think this is a pretty good sampling of major improvements to Wikipedia. Thanks, everyone! [19:52:48] kaldari: is there anyway to get notifications delivered to my IRC channel? [19:53:01] T13: i'm working on it! [19:53:12] T13: https://gerrit.wikimedia.org/r/#/c/80710/ [19:53:41] sweet [19:53:50] T13: I have some primitive code so it stalks the IRC feed, and if it sees a Log/thanks action, it'll check your notifications and give you a link for what you were thanked for, and who did it [19:54:51] And it polls the API every 5-10 min to check for link notifications and such [19:56:16] so no one else has seen that? because if that's happening, it's a major bug. and the only alternative is that I've somehow been clicking on links to foreign-language AbuseFilters, which seems très unlikely [19:56:42] PinkAmpersand: like you randomly clicked on a link to where and ended up at Special:AF? [19:56:48] somewhere* [19:56:51] legoktm: ooooh yay. so i'll be able to get my PinkAmpersand notifications at PublicAmpersand? [19:56:57] yep, pretty sure that's what happened [19:57:15] PinkAmpersand: Well no. That's https://bugzilla.wikimedia.org/48892 [19:57:28] This just exposes functionality in the API for it. [19:57:43] legoktm: I'm more interested in mentions than thanks. [19:58:17] T13: Right. The problem with mentions is there's no way to check in a "push" way, so it just has to be done with polling [19:58:17] was trying to get to https://frr.wikipedia.org/wiki/Wikipedia:Biskiir_Rochtsthemen, wound up at https://frr.wikipedia.org/wiki/Spezial:Missbrauchsfilter [19:58:28] wfm...? [19:58:37] PinkAmpersand: try again :P [19:58:46] worked fine the second time [19:58:55] and the thing is, i looked away from the computer between clicking and viewing [19:59:04] so it's *possible* i clicked something by accident [19:59:07] but I think it's unlikely [19:59:49] and I'm pretty sure the same thing happened yesterday, but I forget which wiki [20:08:00] https://commons.wikimedia.org/wiki/User:Ilmari_Karonen/Queries/Zombie_images [20:09:08] PinkAmpersand, yes, i too got redirected to an AbuseFilter page, after I was auto-logged into a foreign language wiki, a few days ago. I couldnt manage to reproduce, and have cleared cookies/history since then, so can't help beyond confirming. [20:10:29] legoktm: wouldn't the easiest way just to have notifications create an rss/atom feed? [20:10:43] Why can't I subscribe to Wikitech-ambassadors-l, anyway? [20:10:46] T13: You need authentication for that. [20:10:47] quiddity: yeah. well, if you're confirming, shall I file a bug anyways? [20:10:57] T13: Implementing this is the first step for that :) [20:11:26] PinkAmpersand, probably a good idea. [20:11:28] AdamCuerden: you should probably poke Thehelpfulone, he's the mailing list master [20:11:31] will in 1 sec [20:13:31] anyone here experience with the Central Notice? [20:13:36] quiddity: think this is MediaWiki, or Wikimedia? [20:14:24] Romaine: yes, please shoot [20:14:42] i generally make at least 2 errors in any bug report. wrong component is often 1 of them.. >.> [20:14:49] haha [20:14:57] PinkAmpersand: file it under extensions -> AF [20:15:07] someone will come along and triage it properly :P [20:15:14] good idea [20:15:43] waiiiiit [20:15:50] I bet this is actually a CentralAuth bug [20:16:02] * quiddity pictures the scenes from M.A.S.H. whenever the word triage is used. [20:16:28] PinkAmpersand: andre__ will fix it. :) [20:16:49] i'm gonna file in CentralAuth on a hunch [20:17:23] can someone with some tech have a look into why the RfA tally isn't correct here: http://en.wikipedia.org/wiki/Wikipedia:Requests_for_adminship/Singularity42#Discussion [20:18:04] My76Strat: you would want to talk to cyberpower about that, he runs the bot [20:18:07] better link http://en.wikipedia.org/wiki/Wikipedia:Requests_for_adminship/Singularity42 [20:18:19] ok thankas [20:18:21] yeah. that's bot-driven, not software- [20:24:50] ...Hmm. That's weird [20:25:01] http://en.wikipedia.org/wiki/User:Adam_Cuerden <- Do #3 and #12 look odd to people? [20:25:16] Like Packed is giving them extra margins on one side? [20:32:53] quiddity: do you remember what page yours was? [20:35:10] fraid not [20:35:32] ok. i'm CC'ing you, if you don't mind [20:35:41] AdamCuerden: Asking bawolff in -dev [20:35:52] np [20:36:22] Thanks, Mark! [20:36:36] My pleasure [20:36:37] https://bugzilla.wikimedia.org/show_bug.cgi?id=53498 [20:37:07] Ah good [20:37:09] Ah, hey, bawolff. [20:37:19] Hi [20:37:45] Um... So, I found something that's very slightly odd, but I'm not sure if it's a bug or not, with the new gallery [20:38:03] http://en.wikipedia.org/wiki/User:Adam_Cuerden <- Do #3 and #12 look odd to you? [20:38:25] I think they have excess right padding, but I'm not sure. [20:38:42] they do seem like the padding's a bit off [20:38:48] I think what's going on is that the image is super narrow, so the caption for it is forcing the box to be wider, so we don't get a caption that has one letter per line [20:38:55] but they should be centered in that case [20:38:59] Right [20:39:00] which isn't happening [20:39:47] Hmm. Though if I remove the caption for one of them, same issue [20:40:19] http://en.wikipedia.org/w/index.php?title=User:Adam_Cuerden&oldid=570581073 [20:40:43] bawolff: On the image, apply the rules "display: block; margin-right: auto; margin-left: auto;" [20:40:54] The link is width: 100% for some reason [20:40:55] Obviously, these are extreme aspect ratio images [20:41:28] * bawolff thought he did that, guess not. There's been so many patchsets with the gallery, I'm losing track [20:41:52] By the way, while you're here: What's the full list of gallery modes on en-wiki? [20:41:58] bawolff: You put text-align: center on an element and I assumed that would work but apparently failed to test adequately [20:42:08] I want to lead the Tech report in the signpost with the gallery featured. [20:42:23] er, gallery feature featured [20:42:36] I have margin: 0 on the img elements, guess that should be margin: 0 auto; [20:42:49] Maaaybe [20:42:57] AdamCuerden: Currently: traditional, nolines, packed, packed-hover, packed-overlay [20:43:13] Thanks! [20:46:09] Right. Time to write up the feature. [20:46:23] ugh, I need better rewrite rules for thumb.php. Everytime I test thumb.php locally, I end up breaking instantcommons... [20:47:24] btw, secure login by default is now deployed.... [20:49:23] greg-g: really? I was expecting the sky to fall on that one, after all the fighting about it [20:49:38] * bawolff notes lack of flame war in his inbox [20:49:45] shhh, don't tempt the gods [20:51:21] * bawolff gives marktraceur https://gerrit.wikimedia.org/r/81571 [20:53:55] Joy! [20:54:02] Happy happy joy joy [20:54:35] bawolff: Are images block elements already, or do you need to make them so? [20:54:46] Ah, no, we're good [20:54:54] It seemed to work without me making them block [20:54:59] KK [20:58:27] https://en.wikipedia.org/wiki/Wikipedia:Wikipedia_Signpost/2013-08-28/Technology_report <- Here's the report so far. I still need to add VE, but the new report isn't out yet on that. [20:59:16] If anyone else has features they'd like reported on, let me know [21:00:29] bawolff: Doesn't packed-hover become packed-overlay on mobile? [21:00:37] greg-g, possibly useful details in this thread. https://en.wikipedia.org/wiki/Wikipedia:Village_pump_%28technical%29#Secure_site_only.3F [21:00:46] marktraceur: that's correct [21:01:14]