[15:15:01] while my mom is napping, I can do a little work [15:20:46] <^demon> sumanah: I meant to ask you, when you send out your next e-mail to the students, is it going to contain info about applying for commit access? [15:21:07] ^demon: I have done so [15:21:28] <^demon> Ah ok good :) [15:21:32] ^demon: I sent a note today and included: [15:21:40] Welcome to Summer of Code 2011! You have a month -- between now and May 23rd -- to start learning how to contribute code to MediaWiki, and to get to know your mentor(s) and the MediaWiki community. As soon as possible, please look at [15:21:40] http://www.mediawiki.org/wiki/Commit_access [15:21:40] and request commit access to our Subversion repository. If you have any trouble or any problems at all, please tell me or your mentor(s) ASAP so we can fix them! Also, within the next three days, please look at [15:21:42] http://www.mediawiki.org/wiki/How_to_become_a_MediaWiki_hacker [15:21:44] and tell me, and your mentor(s), how far down that list you've already gotten. [15:22:07] <^demon> Awesome! [15:22:17] ^demon: I am trying to be, like, useful [15:22:34] <^demon> http://www.mediawiki.org/wiki/Developer_hub has a lot of good links from it [15:22:50] <^demon> Especially [[Security for developers]] [15:22:57] ^demon: yes! there's a link to it at the bottom of How To Become.... [15:23:13] <^demon> That last link is really important :) [15:23:21] <^demon> Especially for any projects that are hoping for a wmf deploy. [15:23:23] ^demon: so I plan on directing the young'uns to Developer Hub as soon as they are most of the way through How To Become [15:24:05] Thanks ^demon [15:26:39] <^demon> np [15:27:49] *RoanKattouw looks around for an unsuspecting victim [15:27:53] ^demon ! [15:28:04] Does directory.corp.wikimedia.org respond to ping for you? [15:28:32] <^demon> Atm, no. [15:28:42] <^demon> It was working last week when I had to do my ldap thingie. [15:29:41] It was working for me less than an hour ago [15:29:46] But it suddenly stopped working [15:30:08] <^demon> Hmmm. That's an office it thing. Ping Jon? [15:30:34] Already talking to Rayne about it [15:36:00] ^demon: you should see an email with some next steps for mentors [16:54:37] alolita: I might be a bit late calling into the standup cause I'm only just starting to eat now [17:10:43] alolita: TrevorParscal: howief: awjr: [17:18:50] Fatal error: Call to a member function isListed() on a non-object in /Applications/MAMP/htdocs/SVN/mediawiki/trunk/phase3/includes/SpecialPageFactory.php on line 341 [17:18:54] caused by PrefSwitch [17:20:14] hey guys, I'm a bit stumped on what's up with my UploadWizard deploy. [17:20:35] Yesterday, we found that if we did a few uploads with UploadWizard on test.wikipedia.org, our browsers would get slow and/or crash [17:20:58] Can't simulate this locally, even in 1.17wmf1 branch and enabling gadgets which are enabled on test.wp.o. [17:21:00] Ideas? [17:21:27] Are there any other things running on test.wp.o that might interfere with how a JS app runs? [17:23:28] fuuuuuuunky [17:23:31] neilk_, is it still on now? [17:24:05] no, Roan reverted [17:24:23] ^^brion [17:24:38] I backed it out [17:24:45] well could be something in the local site js or one of the other exts being run on it [17:24:46] But both Neil and Kaldari saw slowness [17:24:52] but we should debug in browser and see if we can find something [17:25:54] brion: I can't think of anything else to look for, though [17:26:06] nothing in console, no errors, and the Firebug profiling is similar [17:26:30] cpu, memory usage high in the browser? [17:26:48] not that I can see from my tests locally today, but I wasn't checking that yesterday on test.wp.o [17:27:01] but that is to be expected because the browser isn't crashing today. :) [17:27:08] heh [17:28:03] i'd keep an eye out for any mad memory leaks (like is it slurping all your files into a giant array, pushing them up, and then not quite getting rid of it?) or weeeeeird dom manipulation that never dies [17:28:14] but probably the main thing is: need to be able to reproduce it to test it [17:28:22] dont' be shy about gdb'ing into firefox too ;) [17:28:43] brion: no, no [17:29:22] brion: i mean, it's not doing anything with FileAPI -- your FileAPI changes are not enabled in Special:UploadWizard (yet). [17:29:27] it's just regular boring uploads [17:30:00] You said it could be that you're doing something expensive in the tick() of a JUI animation [17:30:05] well it's something in the wacky browser :) [17:30:06] possible [17:30:07] that seemed logical, but I'm not. [17:30:13] and it's not reproduced locally [17:30:58] There isn't some request flood going on (in the Firebug net panel)? [17:31:03] common.js and vector.js seem empty on test [17:31:10] RoanKattouw: I hadn't ever popped open the profiler for this app before, so I wasn't prepared for some high counts on some functions. But they are the same locally, and they don't take more than a fraction of a millisecond even for a few K's of calls. Due to them being optimizable [17:31:43] er several tens of milliseconds anyway, for K's of calls. [17:31:45] neilk_, does the slowness survive after navigating away from uploadwizard stuff? or is it only while the page is open? [17:31:52] brion: we didn't try that. [17:31:59] it tended to crash :) [17:32:02] heh [17:32:11] also -- are you testing multiple browsers? or just one version? [17:32:25] brion: saw it in FF and Chrome, on my side. I think Kaldari was using Safari. [17:32:50] at first I thought it was some weird coincidence they both got slow / crashed. [17:32:54] hitting both gecko and webkit is a good sign, that probably means it's a bug in our code and not the browser [17:33:02] which'll make it much easier to find :D [17:33:26] y'all might have to just put it back until you can reproduce it [17:33:33] I thought it might be jquery version skew between 1.17wmf1 and trunk, but can't reproduce [17:33:44] yeah I'm starting to think that [17:33:46] 1.4.2 vs. 1.4.4 [17:34:06] there are various jquery ui things too, unless I reverted those... [17:34:15] *brion I recommend replacing the AE-35 unit and waiting until it fails, then we can isolate the fault. [17:34:52] I believe this problem can only be due to human error. [17:35:07] we must preserve the mission at all costs [17:35:18] for science. [17:35:58] i bet hal would have been happier with a weighted companion cube [17:39:46] TrevorParscal: /trunk with AFT setup seems to be Fatal erroring. Culprit appears to be PrefSwitch (commeting that out fixes it) [17:39:57] Fatal error: Call to a member function isListed() on a non-object in /htdocs/SVN/mediawiki/trunk/phase3/includes/SpecialPageFactory.php on line 341 [17:40:06] hmm [17:40:13] some core change, they took that method out> [17:40:14] ? [17:40:46] Dunno, been looking at it in a few ways for the past minutes. Can't find anything obvious [17:56:01] It's a non-object error, not a missing method error [17:59:12] Krinkle: Poke, meeting [17:59:36] RoanKattouw: Thx [18:48:20] RoanKattouw: I heard you mentioned something about "tech-days", I take it you didn't mean the Berlin hackathon ? [18:48:33] Krinkle: you changed from -dinner to non-dinner in 2 minutes [18:48:35] That was fast [18:48:38] The tech days are the Monday and Tuesday right after the hackathon [18:48:51] neilk_: turns out dinner isn't ready for me yet. I was ready for it tho :D [18:49:22] Krinkle: if you can it would be great if you were at tech days. Of course you'll have to ask somebody to extend your accomodation [18:49:46] I don't know what the Official Policy is [18:49:56] Might be a hard sell since they seem to be keeping contractors out of it [18:50:45] bleah [18:50:54] it's ridiculous to call you a contractor [18:50:57] (Roan) [18:51:03] neilk_: Aside from that, I'm not sure I can be away for 4 days from college (Friday, Mon, Tue, thus prolly begin of Wed. too) [18:51:17] Timo & Jan Paul, I guess they're pretty new, so whatever [18:53:00] neilk_: The reason I am invited, I hear, is because I'm slated to become an employee anyway [18:53:30] ok [18:55:50] Same thing for Sam but he couldn't come [18:56:08] ok Roan, so deploying again [18:56:15] Yes, part deux [18:56:40] Kaldari committed a few more changes recently, they are trivial CSS and HTML things. Can we sync to trunk again or do you want to do what we did yesterday exactly [18:56:45] prefer the former [18:57:18] I'll sync to trunk [18:57:46] ok [18:57:50] I reviewed them & marked them ok [19:04:21] nimish_g: Next time you deploy something, could you not merge it to 1.17wmf1 until after the deployment before it has finished? [19:04:44] It's slightly inconvenient for me that there's CUS changes in the branch while I'm trying to only push UploadWizard changes [19:05:37] RoanKattouw: ok, sorry. I didn't realize you were still deploying. I'm commiting the proper change to special:userlogin and I'll leave stuff alone till you're done [19:05:41] what is CUS? [19:05:54] CustomUserSignup [19:06:11] nimish_g: No worries, I'm using selective path updates to work around ti [19:06:41] Don't worry about leaving stuff alone, the "damage" is done already [19:07:33] neilk_: Alright, we're live on test now [19:07:55] ok [19:09:08] Simple test worked for me: http://test.wikipedia.org/wiki/File:Aafail.png [19:10:04] kaldari: What's the rev number(s) of that parser function change you were talking about? [19:10:36] there are 2... [19:11:00] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/86927 [19:11:09] for the multilanguage support.... [19:11:20] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/86805 [19:11:29] for fixing a bug with year parsing [19:12:03] neither are reviewed yet though [19:12:20] Alright, it's tagged [19:12:28] thanks! [19:12:47] Nikerabbit: You commented on r86927, what are your thoughts on its deployability? [19:13:22] ok first anomaly [19:13:33] I am seeing funny errors with the check for filename uniqueness [19:13:53] funny errors? [19:13:54] the API is returning the correct response, but the code that receives the result believes that the result is blank [19:14:03] it logs a message saying 'ok-but-empty' [19:14:17] that is, it got 200 OK but the result was empty, yet Firebug is showing a full response. [19:14:32] And it's same-domain? [19:14:33] RoanKattouw: leeet me see [19:14:48] Nikerabbit: (not urgent) [19:14:52] Kaldari: shall we talk.. love? WikiLove, that is? [19:15:11] RoanKattouw: was that to me? [19:15:15] Yes [19:15:27] it's just an API request to the server to see if the filename exists [19:15:39] Hm OK so yeah that should be same-domain [19:15:40] janpaul123: sorry, but I think it'll have to wait until after this UploadWizard deployment :( [19:15:43] RoanKattouw: well, it's a solution to problem with no immediate alternative [19:15:49] ok actually it's working now and I think the code was correct -- it was another API call that failed, I just was looking at the wrong call. [19:15:51] I went through the wizard just now and it worked fine [19:16:03] kaldari: alrighty, no worries ;) [19:16:09] Nikerabbit: So you're saying it's not optimal but it's what we have and it's fine for now? [19:16:12] janpaul123: if it's getting late we can regroup tomorrow earilier [19:16:14] I'm seeing nothing strange with my FF browser. [19:16:16] Trying in Chrome [19:16:28] Ok [19:16:28] I tested in FF4, was fine [19:16:43] neilk_: I haven't seen the error either [19:16:48] FF4 [19:16:58] I was trying FF3.x [19:17:10] now trying Chrome whatever-chrome-updated-itself-to [19:17:12] Testing Rekonq [19:17:15] do I need to try with a filename that is already on test.wiki? [19:17:27] RoanKattouw: it's known not to work on Konqueror [19:17:35] Oh heh [19:17:42] Gonna watch it fail for teh lulz then [19:17:47] kaldari: you can if you like, or when you get the change to change it do so [19:17:52] RoanKattouw: yes [19:18:05] Whoooa [19:18:15] The magic growing textarea grows a line on every keypress in rekonq [19:18:34] Other than that everything worked fine (!) [19:18:54] http://test.wikipedia.org/wiki/File:Badtoken.png [19:19:38] What's up with this template loop [19:19:39] http://test.wikipedia.org/wiki/File:Apposite_postarthritic.jpg [19:19:43] kaldari: I've put your multilang change on the 1.17wmf1 list, the other change was already on it. It'll hopefully be dealt with this week [19:20:09] neilk_: {{self|self|cc-zero}} [19:20:19] is that the code it's generating? [19:20:26] Where Template:Self seems to call {{{{{1}}}}} and {{{{{2}}}}} [19:20:34] hurr, that's weird [19:20:35] So {{self}} tries to call {{self}} and the parser barfs [19:20:50] but it is supposed to be just {{self|cc-zero}} [19:20:53] I think [19:20:56] When I try to upload a different file with the same name I get a duplicate content error consistantly. [19:21:33] FYI "upload more files" is known to be broken now -- I wanted to fix that bug this morning but was busy reproducing the error [19:21:59] let me try that. [19:22:45] actually, I get that error with any upload attempt after the first one [19:24:38] kaldari: it can't actually give you a naming error upon upload, it is silently renamed to a random string. [19:25:00] the naming error will occur later, at the description page. Just verified this. [19:25:09] actually, let me test with images that aren't actually on Commons! [19:25:13] duh [19:25:16] :) [19:25:23] ok I have tested this over and over & nothing [19:25:31] I think it was a weird coincidence [19:25:36] forgot about instandCommons [19:26:06] RoanKattouw: would you allow me to fix a small bug before we deploy? If it takes more than 5 min I will be surprised. [19:27:11] Sure go ahead [19:27:48] ok [19:28:38] ok, duplicate name actually looks like it's working OK to me [19:31:51] neilk_: one minor bug: You can't actually click on the link it gives you in the duplicate file error. Clicking it just causes the error message to disappear. [19:31:55] nimish_g: codurr krinkle * /trunk/extensions/ClickTracking/modules/ (ext.UserBuckets.js jquery.clickTracking.js sampleCampaign.js): [19:32:23] kaldari: odd, I tried this and it worked for me. [19:32:32] kaldari: okay, please add to buglist [19:32:38] ok [19:32:41] attach the file so I can try it [19:32:56] it's on duplicate file name [19:33:00] in the describe step [19:34:10] neilk_: I guess Nimish will want me to yield the floor in 20 minutes for the CUS deployment; how long till you're ready? [19:34:25] less than 3 mins [19:34:29] OK [19:38:40] RoanKattouw: r86977 [19:38:49] that was 5 but I had a bug :) [19:39:04] amazing how I manage to put a bug in three lines of code [19:39:12] Dude [19:39:17] I once put two fatal errors in one line of code [19:39:24] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/869766977 [19:39:33] er, nm that link [19:40:03] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/86977 [19:41:29] Merged and deployed [19:42:20] RoanKattouw what does it mean that uploadwizard is now in the trunk state ? [19:42:37] It means that we're probably updating UW on commons in the next 10 mins [19:42:41] If neilk_ gives me the green lgiht [19:42:48] looks good to me. [19:42:49] done! [19:42:59] We've been doing weekly bugfix deployments for UW for like three weeks now [19:43:03] ^^ RoanKattouw [19:43:13] does it have an updated user interface ? [19:43:19] yes [19:43:34] happy when I blog about it ? [19:43:38] GerardM-: it's more updated this time than previously, although there are still some known issues. I think by next week I'm going to call it 1.0 [19:43:43] yes [19:43:51] So maybe hold off until we're happier with it? [19:44:06] GerardM-: what sort of blog post would you like to write? [19:44:18] I am open for suggestions [19:44:29] GerardM-: oh you mean for like Signpost or something, or your own blog [19:44:31] ? [19:44:36] his own blog [19:44:40] I prefer to write about the story of getting there [19:44:42] yes [19:44:51] it's a long story :) [19:45:06] I know I was at the inception in Paris [19:45:19] neilk_: Pushed to Commons [19:45:28] GerardM-: well sure, whatever you would like. If you want to talk to me, snag me here or neilk@wikimedia.org [19:45:37] so it is now live [19:45:43] GerardM-: yeah, it's been a pretty long road from there. (Grand soupir) [19:45:53] going to make some screenshots first [19:46:33] GerardM-: It's been live for longer, just not advertised [19:46:37] fyi thelmadatter will be back from a trip in Mexico ... she plans to test it in earnest [19:46:38] [[Special:UploadWizard]] [19:46:47] I kniw [19:46:52] know [19:48:24] RoanKattouw, Kaldari: looks great, thank you again. I think it was a false alarm, but let's watch the feedback & bug reports [19:48:39] will do [19:48:42] OK [19:48:47] How many pictures can you upload in one go ? [19:48:58] 10 I believe [19:49:11] although you still have to select them one at a time [19:50:03] can this number be increased for GLAM relations ? [19:50:41] I think that will be a good topic for discussion at GLAMcamp [19:51:05] MultiChill was asking about using it for Wikipedia Loves Monuments as well [19:51:52] I think we need a more robust stashing mechanism before we increase that number significantly [19:52:37] neilk_: Is ops working on any kind of better stashing system currently? [19:54:22] I'm going to try uploading an 80MB file through UW shortly... [19:55:43] Kaldari I would love to come ... but there is a lack of funding at my end [19:56:15] Kaldari, I have a 800 Mb image that I would like to upload [19:56:19] that's unfortunate, you should ask Liam if there are any grants available [19:56:44] it is a file restored by Durova and from the Tropenmuseum [19:57:55] so when you feel confident ... I am happy to test it for you [19:58:47] nice :D [19:58:49] 800MB? that doesn't sound like a reasonable size for an old photo [19:59:05] i assume it's an *absurdly-detailed* tiff [19:59:30] kaldari: well, it's not ops, it's up to us. It's all application code. We may get them to implement it slightly differently given the new options in file storage. [19:59:42] always start with the highest-quality original you can get, you can never add pixels back but you can always scale down for thumbs :D [19:59:45] they are aware we want to do this, but otherwise it's just temp storage. [19:59:46] GerardM-: The upload limit is still 100M [19:59:50] film grain even on a large-format camera would not create 800MB of actual image information [20:00:09] Roan I know [20:00:25] I can import things if you need me to, file a bug [20:00:25] it would have to be an overly-high-res scan of a photo [20:00:38] But I agree with kaldari that 800M for a single image is probably excessive [20:00:38] Kaldari they are stiched pictures of a storycloth of 15 meters in length [20:00:49] I watched a 1.5hr movie today and it was smaller [20:00:50] ah, ok! [20:00:58] daaaaaaamn [20:00:59] GerardM-: yeah, but in practice, who is going to use it as a single giant image [20:01:22] 16k x 16k at 24bpp would about do it [20:01:51] neilk_ the problem is that we are backward in images anyway ... for a mobile phone we should not send initially an image larger then its size [20:02:12] an image larger than its size [20:02:15] there are several open source applications that help with navigating big images [20:02:24] you mean larger than the screen of the phone [20:02:28] yes [20:02:30] aha [20:02:35] yeah that makes sense [20:02:54] although, sometimes I do want to pan around a big-ish image on the phone. You can't really predict what people are going to want. [20:02:59] people pay for all the data they receive [20:03:04] you can provide reasonable defaults [20:03:29] is this about panning through tiled images again? [20:03:30] neilk_ software like Djatoka and the stuff we have on Toolserver does allow for that [20:03:37] I actually use that as a feature... I download large images from Commons on my cell phone to test data transfer rates :) [20:07:06] neilk_ what is it that people will notice in this version ? [20:07:39] one change is that you can now enter arbitrary dates [20:08:31] ? [20:08:56] before it was limited by what you could choose in the datepicker [20:09:11] i.e. now you can enter non-ISO dates, like 'circa 1934' [20:10:09] that's the part I changed for the week, so that's all I know about :) [20:11:11] TrevorParscal: (copied from gtalk) So about tomorrow's AFT deployment - I'd prefer to deploy trunk, with the dashboard special page in development turned off somehow [20:11:37] TrevorParscal: Also, do you need any code review from me before then? I'm assuming you and Krinkle will have cross-reviewed the front-end parts [20:13:00] there's some stuff in the queue I think [20:13:55] http://www.mediawiki.org/w/index.php?path=%2Ftrunk%2Fextensions%2FArticleFeedback&title=Special%3ACode%2FMediaWiki [20:14:50] OK [20:14:55] I'll poke tomorrow [20:15:55] cool [20:16:23] nimish_g: Mind if I go ahead and fix /branches/wmf/1.17wmf1/extensions/CustomUserSignup so it's copied properly with svn copy which retains history? [20:16:47] RoanKattouw: ok go ahead [20:20:15] nimish_g: Done [20:20:51] nimish_g: For future reference, use something like svn copy svn+ssh://svn.wikimedia.org/svnroot/mediawiki/trunk/extensions/Foo ~/mediawiki/branches/wmf/1.17wmf1/extensions/Foo [20:21:09] and that retains the history? [20:21:29] Yes [20:21:35] Same for move/rename [20:26:56] kaldari neilk_ where do you want people to inform you about their experiences ? [20:27:13] there is a feedback tool built into the UW interface [20:27:38] UW stands for ? [20:27:42] UploadWizard [20:27:45] sorry [20:27:48] ah ok [20:29:02] kaldari neilk_ http://ultimategerardm.blogspot.com/2011/04/commons-has-uploadwizard.html [20:31:25] looks like thumbnailing is still failing :( [20:31:40] which is not due to UploadWizard [20:32:33] Failing how? [20:32:36] Some thumbs may 403 [20:32:49] Although Ariel was supposed to have fixed that like yesterday [20:35:18] thumbnailing seems to fail randomly in UploadWizard. I'm not sure exactly what the specific failure is, but they seem to come in waves [20:35:42] any server-side error messages that can be traced? [20:36:01] I think it's just a time-out on the API request [20:36:11] but I haven't troubleshot it [20:36:21] is that a word? [20:36:48] i say go with it! [20:36:54] perfectly cromulent [20:40:58] GerardM-: if thumbnailing ever becomes reliably speedy again, you'll see nice little thumbnail previews in UploadWizard. [20:43:14] kaldari blogging is about travelling a road ... I can blog again :) [20:43:35] OK, finally got my huge jpg saved. Here goes nothing [20:45:16] it ended up being 75MB [20:45:28] uploading now... [20:47:49] GerardM-: I wish you had posted screenshots where the thumbnailer had worked :) [20:48:24] GerardM-: anyway, I need to fix it ... our thumbnailing is flaky and doesn't give instant results, and the code needs to deal with that [20:48:29] neilk_ there is always a next time to do entheuasiastic [20:48:45] now you can have me blog twice [20:48:51] sure [20:49:04] please link to our feedback page, or post my email address [20:49:17] upload and thumbnails seems to have worked [20:49:24] they work. Usually. [20:49:26] can you give me the feedback page URL ? [20:49:30] sure [20:49:50] http://commons.wikimedia.org/wiki/Commons:Prototype_upload_wizard_feedback [20:50:05] there's a link in the subhead right below the headline to post there [20:50:56] done [20:51:43] thanks! [20:52:09] I don't suppose I could convince you to try uploading again where the screenshot does work? [20:52:50] GerardM-: also, what browser is that? [20:54:14] I use FF on Linux [20:55:14] GerardM-: does the thumbnail ever work for you? [20:55:59] I do not upload that much [20:56:17] GerardM-: I feel like you should not blog about it if it doesn't represent what 90% of people see [20:56:49] GerardM-: there is a bug with thumbnailing but it's that it doesn't deal with Commons' failed thumbnails well [20:57:02] :) it is just bad luck [20:57:53] In two days I can do this again with thumbnails (provided you get me the screenshots) :) [20:58:01] In two days? [20:58:03] I don't understand. [20:58:21] It is better to wait a few days for a next post [20:58:31] I'm suggesting you could update the post you have [20:58:47] that is likely already in the aggregators [20:58:51] I know it's a pain to make screenshots. If you want I will give them to you. [20:59:12] Get me screenshots and I happily use them in two days [20:59:37] I feel very unhappy about your decision [20:59:48] most people see my blog through an aggregator [20:59:59] and changes now do not get reflected [21:00:00] that caches images? [21:00:06] yes [21:01:14] two blog posts is much better publicity :) you should be happy [21:02:07] If you want me to be unhappy about this you are succeeding. [21:02:14] Fine, ttyl. [21:05:47] Hey which one of you nice young lads can remove a 'category' entry from db & memcache for philippe? :D [21:06:05] omg [21:06:12] or else deploy http://www.mediawiki.org/wiki/Special:Code/MediaWiki/80406#c16358 [21:06:32] Roan has gone... [21:06:34] awjr, around? [21:07:01] roan never sleeps! that boy's always online far as i can tell ;) [21:07:06] Reedy: that depends on what you want me to do [21:07:16] :p [21:07:17] If I merge r80406, can you push it to the cluster? [21:07:18] what's up [21:07:32] is it scary? [21:07:37] brion has reviewed it [21:07:40] It's not a big change [21:07:47] I just don't have full rights to push [21:07:47] it's a small fix so that deleted, unused category pages are properly marked as 404 [21:08:04] yeah should be no problem [21:08:06] needed to get spam/annoyance crap out of google cache [21:08:10] thanks guys! [21:08:13] no sweat [21:08:21] Reedy: lmk when you're done merging [21:08:23] *Reedy logs an RT ticket to give brion shell back [21:08:32] it's my favorite kind of fix -- the kind i thought was already in the code before and why didn't we do it that way already :D [21:09:09] RELEASE-NOTES conflicted [21:09:09] lul [21:09:20]