[00:07] spm: (and concluding with the doc that ties all the threads in this channel together: http://svn.red-bean.com/repos/main/3bits/utf8_xml.txt) [03:40] wgrant: around? [03:40] lamont: Indeed. [03:41] did bigjools dump the lp-buildd faceplant on you? [03:41] gimme a minute and I'll pastebin it [03:41] He emailed me about it, but he private-pastebinned it, so I can't see it. [03:41] (I replied as such, but I doubt I'll get a response until next week) [03:42] actually, I did that - sorrty [03:45] http://paste.ubuntu.com/407378/ [03:45] OK, so, firstly... WTF. [03:45] Secondly: WTF [03:46] Thirdly: Did you not run 'debian/rules package' again? [03:46] It looks like you have a pre-Wellington buildd-slave.tac, but a post-Wellington debian.py [03:48] oh hrm... [03:48] I did it both ways [03:48] I think [03:48] What does your buildd-slave.tac have in the way of registerBuilder calls? [03:48] One or four? [03:49] well, i did it both ways, that one is probably the 59 with 56's launchpad-files [03:49] That explains this failure. [03:49] we really need to break lp-buildd into it's on bzr tree [03:49] How does it fail the other way? [03:49] well, it failed the other way too [03:49] We're about to, yes. [03:50] It fails the same way the other way? [03:50] lookiung [03:51] exceptions.OSError: [Errno 17] File exists: '/home/buildd/build-1361010-2804267' [03:51] sigh. [03:51] Ah, well, that one is obvious. [03:51] we should handle that case better [03:51] There's a bug for that™ [03:51] But it may become less relevant when we change the suffix soon. [03:51] let me push another true-59 pile and see how that does [03:52] I noticed that 59 probably needs a config migrator. [03:52] Also, https://code.edge.launchpad.net/~wgrant/launchpad/get-translations-jobs-working/+merge/22584 has a fix to get translations jobs working. [03:53] ah - we have a new set of hardware coming on for the non-package-build buildds - my primary motivator for this roll is getting it so that karmic and later twisted works [03:53] The armels are already Karmic, right? [03:54] old armels are karmic, new and shiny armels are lucid [03:54] Ah [03:55] close enough and didn't see much point to leaving them on karmic [03:55] lamont: Nothing later than 56 is in prod, right? [03:56] correct [03:56] though ferraz et al have 59-ish [03:56] Right, but they are not production. [03:56] They are SLOW. [03:57] Anyway, rolling 59 without the config migration sounds like a recipe for disaster later on. [03:57] http://paste.ubuntu.com/407382/ [03:57] Gnrgggh [03:58] Oh, that's in abort() [03:58] That's well known. [03:58] afk for a few [03:58] Why was an abort attempted? [03:59] (I had a fix for that, but I think it has probably been lost somewhere) [04:02] meh. could be stale state [04:02] A restart should fix it. [04:06] should is the relative term [04:06] just finishing up a stop, scan for timeout fail, start round [04:07] * lamont watches the log [04:09] 2010/04/01 04:08 +0100 [-] Iterating with success flag 0 against stage CLEANUP [04:09] we can maybe haz win [04:09] back in a couple [04:09] (the abort thing has always been problematic; it's not new) [04:09] Yay. [04:28] wgrant: so yeah, uploaded and happy. Is there more that needs to land before I roll 59 out to production? [04:28] that is, do you want to do a config migrator for it? (can I have it in the next 8 hours?) [04:29] I'd prefer to roll the buildds at the start of my workday (while london is still online, etc), and not on a friday [04:31] lamont: The config migrator is a trivial appending one... so I can have it to you in a few minutes. [04:32] lamont: The buildd hunk of the branch I mentioned earlier is also relevant, though it's unclear when we'll start or using or whether further changes will be required. [04:32] Errrr. s/or using/using it/ [04:32] right [04:33] https://code.edge.launchpad.net/~lamont/launchpad/launchpad-buildd_59 <-- wgrant fwiw [04:33] seems twisted/hardy didn't like the --umask option [04:33] oops [04:33] I was about to ask about that. [04:34] so that's what the current 59 was built from, feel free to base off that and push to it if you want, or whatever [04:34] Yep, pulling that down now. [04:35] the other question is if you feel a need for heavy testing tonight, or if I can go crash for 6 hours before my start-of-day [04:36] I can purge-n-reinstall 56 on ferraz, and confirm that the upgrade to 59 makes for a working buildd, at least from a package building perspective (and will before rolling to production with 59) [04:36] I know the binary and recipe slaves work for success cases and at least CHROOTWAIT. I haven't tested the other failure cases aggressively. [04:36] I also plan to poke bigjools about the translations thing to see how he wants to go with that for this week [04:37] bigjools is flying tomorrow. [04:37] He probably won't be around. [04:37] meh. slacker [04:37] right. will pester noodles then [04:37] that would explain why he told me to poke noodles to land my changes [04:37] Heh. [04:37] Yeah. [04:38] so yeah.. known bug fixes and such - email me the details and I'll see how much of it I'm comfortable with for the right-before-easter rollout, and how much wants to wait until next week [04:39] * wgrant is just testing the migrator. [04:39] but I should go crash now, if you don't need more [04:39] Yeah, I was wondering if this was part of a big break-everything-before-people-leave-for-four-days campaign. [04:43] it's a normal weekend in the US [04:43] as in I'm around both friday and monday [04:44] Oh. That's news to me. [04:44] pretty much the rest of the world takes a 4-day as a federal holiday [04:44] not the US [04:45] Heh. [04:48] lamont: lp:~wgrant/launchpad/buildd-59-migrator pushing. [04:48] It works. [04:52] ta [04:54] diff looks cool here. to sleep, perchance to dream, and will deal with all that in the morning [04:55] Night. === apachelogger_ is now known as apachelogger____ [11:41] has launchpad some problems with the login ? I can't login, I get "Invalid OpenID transaction" [11:42] ah, it's in read-only mode, thanks === effie_jayx is now known as keffie_jayx [12:13] is the login broken ? (Invalid OpenID transaction) [12:15] obergix[work]: it's in read-only mode, I think it's normal atm [12:16] ok === mrevell is now known as mrevell-lunch [12:33] Is it known that the package "build status" page shows inconsistent "Started X seconds ago"? [12:40] Daviey: Do you have an example build? [12:44] wgrant: not now, it's showing correctly.. but before the buildlog content is visable.. when refreshing for over a minute it jumped from 4 seconds, 14 seconds, 8 seconds [12:45] Daviey: That sounds like it was failing to dispatch, so it was retrying repeatedly. [12:45] Daviey: If you see it happening again, try to grab the build ID and builder name. [12:46] wgrant: well i just lurked on a build under /builders/* , and it didn't do that [12:47] wgrant: the one i noticed it on was https://edge.launchpad.net/~mythbuntu/+archive/0.23/+build/1595942 , follow pressing rebuild after a dep wait issue. [12:47] it's not really a major thing really.. just something odd :) === matsubara-afk is now known as matsubara === matsubara changed the topic of #launchpad to: http://launchpad.net/ | Read https://help.launchpad.net/ for help | Help contact: matsubara | Join https://launchpad.net/~launchpad-users | This channel is logged: http://irclogs.ubuntu.com/ | Launchpad is open source: https://dev.launchpad.net/ === daniloff is now known as danilos [13:26] * pmatulis waves to mneptok === apachelogger____ is now known as apachelogger === mrevell-lunch is now known as mrevell [13:45] https://edge.launchpad.net/builders/molybdenum/+edit <-- why does that page seem to always time out?? [13:46] I have a (the ?) http referer "problem", and strongly prefer not to enable http referer for all cases. How can I file bug reports? [13:46] is there a workaround for firefox? [13:48] Drakeson: i use the "RefControl" extension, it lets you specify policy per-site if want to [13:48] does it allow you to "only send referer from within the same domain"? [13:49] Drakeson: yeah [13:49] Drakeson: that's actually the policy I use, I don't have any per-site overrides [13:49] (it's the "3rd Party requests only", IIRC) [13:51] I wish it was the default behaviour. Currently I have network.http.sendRefererHeader set to 0, and I don't know how to find the documentation for the variables in about:config. [13:52] lamont, do you have an oops id? [13:52] Drakeson: http://kb.mozillazine.org/Network.http.sendRefererHeader [13:52] lamont, I see you're already talking to soyuz guys. ignore me :-) [13:54] Drakeson: (and in general , which is the first google hit for "about:config") [13:55] Please try again [13:55] Sorry, there was a problem connecting to the Launchpad server. [13:55] Try reloading this page in a minute or two. If the problem persists, let us know in the #launchpad IRC channel on Freenode. [13:55] Thanks for your patience. [13:55] spiv: thanks [14:44] I have a question in setting up my own instance of launchpad... I'm erroring out with error 1 after trying to do make schema, it was getting permissions issues on deletion and never got to creating the database "launchpad-dev" [14:45] I cannot find a resolution on the answers page, is there a way to solve this issue to allow make schema to continue and complete? [14:53] Byt|Eee, can you copy and paste the output? use pastebin.ubuntu.com please [15:00] http://pastebin.ubuntu.com/407581/ [15:01] That is the error out section and return to command line, matsubara [15:03] Byt|Eee, did you run make with a different user than you're running make schema with? [15:06] matsubara: No, I have simply provided my suso password with every entry so far for super-user access. I have always been logged in as lpuser. [15:06] ah, sudo I meant. [15:06] hello, is it here were should I report of someone abuse in lauchpad bugs? [15:07] Damascene, what's up? [15:07] Byt|Eee, what's the output for ls -alh /var/tmp/ ? [15:08] this man https://bugs.launchpad.net/vte/+bug/263822/comments/17 [15:08] Ubuntu bug 263822 in vte "RTL (right to left) support in terminal (BiDi)" [Low,Triaged] [15:08] if I select my [15:08] location to Jerusalem the settings of the calendar should act accordingly [15:08] (In Israel the rest days are Friday which is a half workday and Shabbat [15:08] which is a full nonworking day) [15:09] he ignore the international law and claiming the ownership of jerusalem on launchpad [15:11] he couldn't chose any other cities that have less sensitve improtant to Muslims [15:11] *city [15:13] Damascene, I fail to see how his comment is abusive [15:13] do you know what is jerusalem? [15:14] yes [15:14] and it seems to me he's giving an example for the settings he's using at his computer. not trying to be offensive or anything like that [15:15] this bug is important to muslim and arabs and thos people disagree on that point too much [15:16] matsubara, Sorry it takes so long, I'm running remotes and VNCs everywhere for this. ls output: http://pastebin.ubuntu.com/407588/ [15:16] his word is absolutely abusive to muslims and arabs [15:17] and he is using lauchpad to promote his idea [15:18] Byt|Eee, you can safely remove /var/tmp/bazaar.launchpad.dev from there. sudo rm -rf /var/tmp/bazaar.launchpad.dev [15:18] Byt|Eee, then try again [15:19] matsubara, Thank you for your help, hopefully that will solve my issue. [15:19] Byt|Eee, all right. let me know if you need further help [15:20] do I expect to get any help here or I should respond to him directly? [15:22] Damascene, I don't think we can help. he's not being abusive at all in the sense that we need to censor his coments [15:22] comments, even [15:23] Damascene, in any case, I think you can talk him and sort it out, but from my understanding of it, he's actually trying to help making the settings work for both hebrew and arab languages [15:27] what if England invaded paris for 100 year. England decided to call pairs an England city. an English man comes to Launchpad stating that he expect to see English software installed when he chose pairs as a city. why he couldn't chose London for his example [15:27] it's too clear. [15:31] matsubara, Thank you again, it completed schema and is running fine now. [15:33] oh right, he can't change his post. and you don't want. one way to solve the issue [15:33] Damascene, I think you're overreacting, he clearly states that depending on the language setting, the rest of the applications should behave accordingly. Using your example, both english and french would be installed by default and according to the language set, if you choose Paris, you'd get settings accoridng to the language set by the user. [15:35] Damascene, I understand that's a sensitive topic for you but he's trying to address the use case for both arab, hebrew and persian languages. not making a statement that he owns Jerusalem or anything like that. [15:36] Damascene, in any case, what do you want us to do? to censor his post? I don't have the power to do that but I can escalate if you want [15:37] so you are suggesting to do nothing at all or to state that Arab calender should be installed too if he chose Jerusalem? [15:37] yes please, escalate it. [15:37] Damascene, I'd have to ask you to file a question here: https://edge.launchpad.net/launchpad/+addquestion and let me know the URL and I can ask the admins if this should be removed or not [15:38] his move is really sensitve. [15:39] Damascene, I don't know enough about this piece of software but it's seems sensible to me, if I have arab as my system language, to get a arab calendar. if i have hebrew as my system language I'd get a hebrew calendar, and if I have english I get the gregorian calendar. [15:40] Damascene, and I think that's exactly what Yaron is suggesting [15:40] matsubara, what the question would be about [15:41] "is it allowed to someone to use launchpad to promote his idea in a sensitive subject"? [15:41] Damascene, explain that you feel someone is abusing LP bugs just like you did here on IRC. it doesn't really need to be a question [15:42] Damascene, sounds good [15:42] ok [15:42] thanks === matsubara is now known as matsubara-lunch [16:03] matsubara-lunch, https://launchpad.net/~sh-yaron [16:03] sorry [16:03] https://answers.edge.launchpad.net/launchpad/+question/106198 [16:04] please tell me after you forwared it to the admins [16:32] matsubara-lunch: I think the bohrium builder is stuck [16:34] Damascene, you can subscribe to the question, I guess [16:35] what do you mean Ursinha? I'm the one who reported that === matsubara-lunch is now known as matsubara [16:56] Hi [16:56] I need to submit a bug for conkyForecast how can I do this? [16:56] This is the only link Ican find https://code.launchpad.net/~m-buck/+junk/conkyforecast === stub1 is now known as stub [16:58] Hi everyone! I have a problem reassigning bug to a different project [16:59] hey micahg, what's the URL? [16:59] Damascene, I'll let you know through the question itself. thanks for filing it [17:00] ok, thanks magcius [17:00] matsubara, [17:00] deuxpi: you have to add the project to "Also affects project" [17:00] matsubara: https://edge.launchpad.net/builders/bohrium It's a 5-10 minute build [17:00] deuxpi: and then set the bug to "Invalid" in the first project [17:01] micahg: I can't make the project appear by searching for it [17:01] deuxpi, do you know the name of the project? [17:02] micahg: If I search for ubuntu-mono, I get a few packages, but not lp:ubuntu/+source/ubuntu-mono [17:02] deuxpi, what's the bug/ [17:02] https://bugs.launchpad.net/humanity/+bug/552558 [17:02] Ubuntu bug 552558 in humanity "Simplifying the GTG monochrome icon" [Undecided,New] === salgado is now known as salgado-lunch [17:05] deuxpi, so, the bug is really on a ubuntu package? [17:07] deuxpi, if you're certaing the bug is on a ubuntu package, you should mark the task as invalid on humanity and then click the "Also affects distribution" link. Choose ubuntu and then the ubuntu package. [17:07] deuxpi, if you're not certain, I'd suggest you to join #ubuntu-bugs and ask for help on how to triage it appropriately [17:09] micahg, checking that. hang on [17:10] matsubara: Ok, if I understand correctly, unless I specity that it affects Ubuntu, I don't get ubuntu packages from the list. [17:11] deuxpi, exactly [17:11] I'll remember that :) thanks! [17:12] deuxpi, you can change projects though. do you know the upstream project for the ubuntu-mono package? [17:12] I think it's maintained by Ubuntu [17:13] matsubara: it's a Ubuntu-specific derivative of Humanity [17:14] deuxpi, right. so in that case I think the correct thing to do is to mark the humanity bug task as invalid and open a new bug task for ubuntu package [17:16] matsubara: done. thanks again! [17:16] deuxpi, np === deryck is now known as deryck[lunch] [17:18] hey pyther, apparently there's no package for that conkyforecast code [17:18] LP doesn't support bug reports against branches [17:19] pyther, my suggestion is to try and contact the author using LP [17:22] matsubara: ok thanks! [17:22] pyther, np === yofel_ is now known as yofel === salgado-lunch is now known as salgado === Ursinha is now known as Ursinha-lunch === deryck[lunch] is now known as deryck [18:21] pmatulis: heya! [18:21] mneptok: just waking up? :) [18:23] pmatulis: yeah. was awake until 3am dealing with EU colleagues and issues. [18:23] the MySQL User Conference is in a week, and i've been buried. [18:23] mneptok: fun stuff, where will it be held? [18:24] Santa Clara. the wasteland of Silicon Valley. [18:39] Hello, what e-mail address should I contact to request permission to be able to assign questions to certain people on Launchpad Answers? [18:40] NicholasHale, for what project ? [18:41] matsubara: Launchpad itself [18:42] NicholasHale, so you want to help triage launchpad itself questions? [18:42] matsubara: Yes [18:44] NicholasHale, cool. I guess I can add you to the team that takes care of that. what's your launchpad user id? [18:45] NicholasHale, I'll make sure that's ok first and then add you or let you know what's up with your request. I guess there's no problem but I'll ask around first [19:34] i'm not sure if this is normal but ghc has been building on arm for 20 days now [19:39] danopia: in qemu or native? [19:39] no idea [19:39] i don't run the bot, i just noticed it on hte build queue [20:13] https://edge.launchpad.net/ubuntu/+builds?build_text=ghc6&build_state=built [20:13] Last build took 42 hours [20:13] So 20 days suggests a problem I suppose [20:18] buildd 12526 0.0 2.3 14344 11124 ? S Mar11 0:46 /usr/bin/make -r --no-print-directory -f ghc.mk all [20:18] buildd 16422 0.0 0.0 1340 204 ? R Mar11 0:02 sh debian/watcher.sh 12485 /build/buildd/ghc6-6.12.1 /build/buildd/ghc6-6.12.1/build-stamp ghc\|gcc [20:18] ^ the current last two lines showing in the build output display, so someone can check back if it's changed at all in a few hours [20:19] lamont, can you check the builder please ^? [20:19] which builder? [20:20] oh, ghc6-armel? [20:20] yeah. that one's special [20:20] I'm purposely ignoring it to see if it finishes building before release [20:20] we're kind of betting "not" [20:21] if too many more people ask about it, I'll just hide the buildd from the UI [20:21] danopia, ^ [20:21] all right. I'll let you know. thanks lamont [20:22] the buildd in question is slated to be turned off and yanked out of the rack late next week or early the week after [20:23] why is that build special [20:25] danopia: it's FTBFS pretty much everywhere, there was some question of whether it was "just really really really really slow" or "dead" [20:25] it seems to make progress, just not very fast [20:25] I'm in the process of turning up another 5 armel buildds, and shutting down all the old ones (need newer hardware to build lucid properly), so it's not a build I've been worrying about much [20:27] danopia: that there are lots of builds queued up is a global issue atm, armel has actually been keeping up quite handily, with only 6-8 buildds [20:28] danopia: jaboticaba and 5 other buildds are sitting on manual so I can do some work on them once they finish their current build. somewhere in there I'll have to make a decision about jaboticaba and how much longer I get to ignore it [20:38] Whats the fastest way to subscribe a single team to 64 different branches? [20:58] cody-somerville, I think by using the API === matsubara is now known as matsubara-afk === salgado is now known as salgado-afk [22:25] Can anyone help me get ec2test running? I had it running just earlier today, but now it explodes on me: http://paste.ubuntu.com/407760/ [22:26] cjwatson: ugh! [22:26] i can only suggest "try again" ? [22:27] I did, twice :-( [22:28] the only thing I'm aware of changing is pushing another revision to the branch under test [22:30] did the ec2 instance change, perhaps to an ubuntu AMI based one? [22:30] cjwatson: you can run with -d and step through until it has started the machine if you want to poke at it [22:31] cjwatson, just pull from devel and it will be fixed [22:31] I'll send an email about it in a bit, need to go afk now [22:32] salgado-afk: ah - thanks [22:45] salgado-afk: thanks, sorted [22:50] strange failure mode! [22:55] lamont: Morning. How's the buildd stuff going? [23:05] wgrant: 5 buildds to quiesce and upgrade [23:05] my launchpad-buildd_59 branch. [23:05] lamont: Only five? [23:05] Or are the rest already done? [23:05] I expect that next week, we'll do it again for the next round of fixes [23:05] 5 left [23:05] ourd buttercup hooker jaboticaba vernadsky [23:06] gourd, that is [23:06] mind you, those machines have been building the same package for at least 9 hours [23:07] and jaboticaba will eventually come down to a race between me upgrading it and turning it off [23:07] Hm, I haven't heard of buttercup before. [23:07] Started 21 days, 10 hours, 10 minutes, 30.3 seconds ago. [23:07] Must be armel. [23:07] gourd and buttercup are armel [23:07] so the list above is armel, armel, ia64, armel, i386 [23:07] Yeah, I noticed gourd a day or two ago. [23:08] acorn will show up once someone sorts out the network cable in the DC [23:09] lamont: No catastrophic failures yet? [23:09] chugging along quite happily, other than a little muppetry on my part [23:10] some of the ppas don't have ubuntu-minimal installed, and 59 kinda lacks an actual Depend: on the minimal-but-not-actually-Essential lsb-release [23:10] Yeah, I saw that last commit. [23:10] non-virt i386 is looking pretty sad at the moment. [23:11] nah - it was much closer to 1000 earlier today [23:11] Ah. It was about the same as it is now when I last checked about 12 hours ago. [23:12] yeah - those long private builds are a bitch [23:12] OTOH, even hppa has a build going now.... [23:14] thanks again for the help last night [23:16] No problem. Thanks for getting the new code out there now. It's good to know it works for the critical things before we need it for the new stuff. [23:19] wgrant: true that. and for bonus points, you can presume that rev 59 has been deployed. [23:20] lamont: Not quite. [23:20] Although I guess all except jaboticaba should be done soon. [23:20] On the downside, the hardy version is 59~0.IS.8.04, and it'd be totally bonus if upgrading to 0.61 (I kinda ate 0.60)... didn't reapply the 59 upgrade to config [23:20] * wgrant assumes, and prepares the first hack-removal branch. [23:20] and jaboticaba will be done or dead [23:21] lamont: It finds the old version by splitting on ~ and taking the first bit. [23:21] So that version is fine. [23:23] ah, cool [23:33] What determines killing a branch import? cause new gcc import almost managed to finish only to be killed while it was working normally [23:33] eg. http://launchpadlibrarian.net/42056577/vcs-imports-gcc-trunk-log.txt [23:34] it just needed a few more hours to finish ;-) [23:35] xnox: I believe that Launchpad 10.03 introduces incremental bzr-svn imports. [23:36] This means that it will import 1000 revisions at a time, so a failure like that will cause minimal loss. [23:36] And it's less likely to fail on small increments. [23:36] I want to report a bug on launchpad.net [23:36] but it's telling me to enable referrers [23:37] I WONT. is there no way reporting bugs on launchpad with referers disabled? [23:37] Not at the moment. [23:38] dauerbaustelle, add GPG key and use email interface with signed emails [23:38] woa [23:38] well then I won't report any bugs anymore, sad [23:39] it is intermediate protection from spam and account hijacking as far as I know. [23:39] It will hopefully change soon. [23:40] It was the most practical solution at the time. [23:40] lots of spam or what? [23:40] so, I haven't heard an explanation for the logic here [23:40] wgrant: do you know the theory ? [23:41] CSRF protection. [23:41] It's unrelated to spam. [23:42] and how does referer help ? [23:42] what's about generating a token and passing it via POSt oder so [23:42] *POST or so [23:42] dauerbaustelle: That is the correct solution. But it is a massive task on such a huge webapp. [23:42] So it would not have been implementable quickly. [23:43] I don't this but I think you know better than me. :D [23:43] lifeless: Forms cannot forge Referers. They can force omission, but not mangling. [23:43] (and XHR has non-braindead security) [23:44] dauerbaustelle: Launchpad has hundreds of forms, many external applications that use its forms, and lots of JavaScript that would probably also need to be changed. [23:44] It is not a small task. [23:48] oh well [23:57] bye