=== jml changed the topic of #launchpad to: https://launchpad.net | bazaar.launchpad.net running slower than usual, we're fixing it now | Next meeting, all welcome: Thu 5 June 2008, 1800UTC #launchpad-meeting | Help: https://help.launchpad.net | Questions and spam reports: https://answers.launchpad.net/launchpad | Channel logs: http://irclogs.ubuntu.com [02:39] »»»¶;]««« [02:49] Slower than usual? It's always "slower than usual". You should redefine "usual". Then you could say "bazaar.launchpad.net is way faster than usual". :D === jml changed the topic of #launchpad to: https://launchpad.net | Next meeting, all welcome: Thu 5 June 2008, 1800UTC #launchpad-meeting | Help: https://help.launchpad.net | Questions and spam reports: https://answers.launchpad.net/launchpad | Channel logs: http://irclogs.ubuntu.com [02:50] Peng: slower than it normally is. [02:50] Peng: in general & for me, bazaar.launchpad.net performs about the same as other bzr+ssh servers on the side of the world [02:51] in any case, the issue affecting the server has been resolved. [02:53] What was wrong? [02:56] Peng: there's an issue with the way our SSH server does buffering. [02:59] jml: You're using a customy Twisted SSH server, right? [03:00] Peng: we use twisted.conch [03:00] »»»¶;]««« [03:00] Okay. [03:07] Peng: I feel a certain level of responsibility for both Twisted and Launchpad, so I'm reluctant to blame either :) [03:07] but signs point to conch atm. [03:10] I once had a website on an ISP, which was slow fairly often, especially at night. After a few months of calling tech support, someone forgot to hit the mute button, and I heard something about "torrents" in the back [03:11] not saying this is the case, just comes to mind everytime I see the word "slow server" :) [03:14] beuno: I'm pretty sure elmo would have strong words for anyone running torrents from the data centre. [03:14] we run torrents from the data centre [03:14] heheh, yeah [03:14] lifeless: do we download movies to the data centre? [03:15] probably don't have a 6mbit cap on the server though [03:15] and Ubuntu ISOs must get a much lower leven of leechers than $random_movie [03:15] no, different server too [03:16] we maintain the seeds list [03:16] I dunno what fraction of the actual data serving we do [03:16] * jml lunches. [03:16] yeah, I was joking of course. I just can't avoid remembering that everytime a random server is slow for a while, and then is magically fast again [03:16] ((NiSiNhA)) »»»¶;]««« [03:16] [c40s]: ¶;*****~ [03:17] ((NiSiNhA)) »»»viaja ae««« [03:17] man, I'm just seeing weird characters there [03:17] c40s: eita começei ja [03:17] phew, I thought it was just me [03:17] I'm guessing spam [03:17] [c40s]: kkkkkkkk [03:17] ((NiSiNhA)) »»»hehe««« [03:17] knock it off [03:18] I don't have op in this channel or I'd boot [03:18] c40s: NiSiNhA: please stop the blnkenflashentext [03:18] ((lifeless)) »»»ok««« [03:19] ha, very ironic saying ok but still doing it [03:19] I think you are about to be kicked [03:19] boot to the head [03:19] Isn't there a channel mode to disable colours? [03:19] jamesh: mode +b first, then kick [03:19] I haven't seen colors in IRC in years! I didn't even think people could do that anymore [03:19] cody-somerville: +b :P [03:19] lifeless, :P [03:20] »»»sorry jamesh««« [03:21] c40s, I think they want you to disable the "art". [03:22] cody-somerville ok [03:22] Thank you. [03:24] c40s: what IRC client are you using ? [03:25] lifeless in from Brazil,using Alive [03:26] * wgrant looks for it in Ubuntu, and files a removal request if it's there. [03:27] http://www.brothersoft.com/alive-java-chat-10509.html ? [03:28] Damn. [03:36] adeuss povaoo [04:05] please fix the hppa, sparc, ia64, and ppc chroots. thanksk. [04:17] lol [04:17] FIX THE WORLD. kthxbye [04:22] so the non-obsolete architecture chroots are working? :) [04:22] lifeless: that'd be nice :) [04:22] jamesh: i think so. [04:22] lifeless: either way, i can't do it from here, and with a bit of luck, lamont will. [04:25] Hobbsee_: borked how? [04:25] and 'tis best left for infinity [04:27] lamont: trying to remove sysvinit, it looks like. [04:33] we need an MMIX ppa [04:33] for larks. [04:34] jml: what would you put in it? [04:34] besides TeX? [04:35] jamesh: I'm sure someone has ported intercal to mmix by now. [04:35] threaded intercal? [04:35] jamesh: and if not, I think all good and right-thinking people deserve an answer as to why not! [04:36] * jml delicately wipes away some froth from the corner of his lip. [04:38] well, you can always use http://entrian.com/goto/ in Python [04:39] I don't think it does multi-threaded come from though [06:01] * jml waves [06:02] :) [06:03] so, as I was about to say, most of the Launchpad hackers are swamped at the moment. [06:03] Ah, that certainly helps. :) [06:03] ToyKeeper: I'll be hanging out with poolie (the original filer of the bug) next week, that might generate some sort of momentum. [06:04] (of course we all know that momentum is conserved, and so something else will have to slow down) [06:04] I hate to nag about minor feature requests, especially if everyone is busy. [06:04] I'd just make a branch and propose it for merging, but I don't have access. :) [06:04] (bug 161187 for context) [06:04] Launchpad bug 161187 in launchpad "not obvious how to add a download file for a new release" [Medium,Confirmed] https://launchpad.net/bugs/161187 [06:04] ToyKeeper: :) [06:05] It seems like a 5-minute code change, but of course testing takes longer. [06:14] Do small UI changes go through a review process? [06:16] yes [06:17] everything does (for large values of everything) [07:36] hmmm thats ugly [07:36] help.launchpad.net says 'not logged in' [07:37] launchpad.net says 'logged in' [07:38] lifeless: we could solve this by not making help.launchpad.net look like launchpad.net ... [07:38] we've had other people note this inconsistency [07:38] is there a bug? [07:39] I like me-tooing [07:39] yes [07:39] feel free to search for it :) [07:39] we apparently have useful search now [07:39] at least search is vaguely useful now [07:39] oh its massively improved [07:39] only thing is it truncates topics :/ [07:42] and seems to include the chrome as search strings [07:42] e.g. searching for "not logged in" seems to get pretty much every page on help.launchpad.net [07:44] has /me/ been altered? [07:45] the ppa quickstart links to .../~me/+archive [07:45] but that does not work on edge [07:53] should be /+me/ [07:55] ah [07:55] who can edit help.launchpad.net? [07:56] depends on the page [07:56] some are locked down (usually if they are pointed at externally) [07:56] others aren't [07:57] anyhow, this is the buggy page: [07:57] https://help.launchpad.net/PPAQuickStart [07:59] you should be able to edit that [07:59] /+me/ goes to "Lost something?" also. Well for me. [07:59] https://edge.launchpad.net/+me/+archive/+builds?build_state=all&build_text= [07:59] ah. it probably needs to be /people/+me [08:00] should be better now [08:01] jamesh: yes that works, thankyou. [08:06] I've got a quick code-hosting question [08:06] is it at all possible to have a team branch that is *not* part of a project? [08:06] https://help.launchpad.net/NonProjectBranches seems to indicate no, but that seems rather odd to me [08:09] hmm, gotta run, i check again tomorrow [09:01] Goooooooooooooooooood morning Launchpadders! [10:44] cprov: What useful information is communicated by the unlinked diff non-links? [10:45] wgrant: unlinked means "the diff will be done in some minutes" [10:46] cprov: 'None yet' or similar would probably make more sense. [10:46] wgrant: which is better, IMO, than forcing you to ask: "Doesn't LP do diff for new uploads ? Where is it ?" [10:47] True. [10:47] But listing something unavailable under a heading of 'available' is strange. [10:47] the fact that we communicate it by not linking the test maybe too subtle [10:47] I suspect so. [10:47] Adding a (pending), maybe. [10:48] wgrant: yes, maybe. [10:48] Also, how is the latest ancestor to diff from detected? [10:48] detected/selected. [10:48] -proposed uploads are diffed against the latest release, regardless of distroseries. [10:51] wgrant: yes, it erroneously assumes linearity. [10:51] Aha. [10:53] It would be rather difficult to build a system which correctly determines the correct ancestor... it would really have to check by archive, distroseries and pocket and make lots of potentially nastily incorrect assumptions... [10:53] It would probably be much nicer to have a form to request diffs against version X, but that would be expensive. :( [10:55] wgrant: manually-requested diff are already modelled and yes, they are more expensive for the users than for us. [10:57] cprov: That's why the UI looks like it was designed to display multiple diffs? [10:57] wgrant: err, we have have to think about an easy way to release this feature, that's why it's not available for users yet. [10:57] wgrant: yes, exactly. [10:57] 'Easy way to release'? A UI? [10:58] wgrant: yes, an intuitively way to lookup for the desired ancestry in the UI, to be precise. [10:59] wgrant: this is a real weakness in the packaging system, it's not "branch friendly" [10:59] wgrant: not a form with archive, distribution, series, pocket, name, version ala bugzilla ;) [10:59] it's hard to be a derivative. and ppa's are effectively derivatives. [11:00] we need to have the same sort of ancestry tracking in packages that we have in bzr branches [11:00] sabdfl: Even between pockets it's an issue. [11:00] yes [11:00] It gets very ugly very quickly. [11:01] sabdfl: Actually interpreting the changelog somehow? [11:01] wgrant: yes, but pockets are complex but predictable, ppas are unpredictable at the package level, we my have to investigate origs (and yes, it won't work for native-tarball uploads ...) [11:03] cprov: What's special about PPA orig.tar.*s? [11:03] They either have to be named differently or match primary's, don't they? [11:03] wgrant: changelogs can be 'inaccurate' :( [11:03] Same as anything in primary. [11:03] cprov: And are regularly modified after the fact, right. [11:04] wgrant: yes, that's why we can use it to find out the right ancestor. [11:04] cprov: Oh, right, like that. [11:05] If only NMSP had actually happened... would make all this trivial. [11:05] but, it's just an idea, I've not spent much time investigating it yet. [11:05] The current setup is a good start, at any rate. [11:06] wgrant: NMSP is (far) future ;) [11:07] Heh. [11:07] wgrant: I will try to round more corners of the current package-diff and see how it does. Thanks for reporting bugs (2 so far) [11:07] cprov: Also, it won't cross archive boundaries unless it has to, right? [11:07] Diffing against some random PPA package is probably worse than useless. [11:08] wgrant: you help is highly appreciated ! [11:08] (I just noticed one of my PPA packages diffed against primary, because it didn't exist in my PPA before) [11:08] wgrant: yes, it only cross archives for the first PPA upload. [11:08] s/cross/crosses [11:09] cprov: Sounds good. Except if I delete a package from my PPA, and upload a new one a couple of years later, based on a later one from primary. But again, that's ridiculously difficult or impossible to work out. [11:09] and it's restricted to the current version in primary archive, it doesn't go across other PPAs. [11:09] Good, good. [11:10] wgrant: we may go for 2 diffs for PPAs, one against local ancestry and another against primary [11:10] That works. [11:12] But getting the defaults right isn't too important if a sane UI for user ancestor selection comes about soon. [11:12] does the new search in launchpad have special reserved characters? [11:13] let's say I would like to search for the german translation team in ubuntu, I search for '-de' [11:14] the search returns a huge number of results [11:14] thekorn: It looks like it uses Google as a backend. [11:14] So you'll be excluding anything with 'de' in it. [11:15] (or Canonical has developed a search algorithm equivalent to Google's, which I find unfortunately unlikely) [11:16] ah, this also explains why it is a real string search, "your location" also returns alot of results [11:17] wghich does not make any sense, imho [11:17] Yeah, they could do with excluding the global templates... === cprov is now known as cprov-lunch [11:19] hmm, it seems the 'choose...' dialog on the 'subscribe someone else' page does not use this search [11:20] Using Google to search a list of people like that would be a disaster. [11:20] AFAIK only the global site search uses Google. [11:20] searching for '-de' there returns an timeout-error with Error ID: OOPS-886EC25 [11:34] wgrant: interpreting the changelog is a bit of a crapshoot [11:37] sabdfl: Right, attempting to retrieve much useful information from it would be an exercise in futility. [11:38] In Ubuntu we regularly drop and reinsert entries into the middle of whatever we get from Debian... it's terrible [11:39] <\sh> wgrant, you mean during merge time? [11:40] \sh: Right. [11:40] And syncs. [11:40] Where we drop all the Ubuntu history. [11:41] <\sh> wgrant, yeah...for syncs it's really terrible [11:42] <\sh> adding the single changelog entries for new versions/revisions into a container per package, could help...but I think that makes the db logic even more complicated/slow [11:43] I don't think the changelog should be in the source package at all, really. [11:43] Like VCS logs aren't part of the actual data stored in the tree - they're metadata alongside it. [11:43] <\sh> well... [11:43] But I doubt we'll see that changing. [11:43] <\sh> debian/ dir is a type of metadata, too :) [11:44] True. But with what we do, it's not persistent. [11:44] It will vanish every second release. [11:44] Logs of changes should be persistent. [11:45] <\sh> yes..that's why it's better to create a history per package ... in debian that's easy...but for derivatives with source changes, it's hell [11:46] As sabdfl was saying earlier, yep. === thekorn_ is now known as thekorn [12:25] hi, is Christian Reis here? [12:26] dhart, indeed [12:29] kiko: oh hiya. I'm typing a response on answers, but thought it might be easier to chat [12:30] dhart, better on answers today as I'm a bit busy [12:30] kiko, okay, I'll write on answers. [12:33] greetings people. Could we please get the launchpad.net/bugs/NNNNN redirect fixed. [12:33] there are URLs in that format published 350,000+ times on the web [12:33] and currently they're busted [12:35] mpt: ^^your kind of thing [12:39] sladen: What about them is broken? [12:39] sladen, works for me. Can you give me an exact example that doesn't work? [12:39] https://launchpad.net/bugs/1234 [12:39] sladen: Error: This bug is private [12:39] Isn't that meant to be private? [12:39] I thought that always had been. [12:40] well, *do the redirect* first have have the canonical location say that it's private [12:40] not bomb out before hand [12:40] That's deliberate. [12:40] That was a bug. [12:40] It was revealing information. [12:41] Could somebody add the words "private" and "bug" to that page, neither appear [12:41] There's a bug on that somewhere. [12:41] * wgrant looks. [12:41] heh [12:41] that bug is actually fixed [12:41] I'll disclose it :) [12:42] sladen, reload [12:42] kiko: Good idea. It's the first one I often check, but then I remember it's private. I'm sure others are the same. [12:42] sladen, and, as I told you on the bug report, that redirect /is/ fixed and has been fixed forever [12:43] that bug is very funny [12:43] kiko: yup, replied on the bug report. [12:43] btw [12:44] that unmaintainable mess is actually going to run over debian soon! :) [12:44] kiko: Woo! [12:44] Nice. [12:44] (ah, nice old statuses in that bug history) [12:46] dhart, my apologies, I'm really getting old [12:47] kiko: no worries. apologies for what? :-) [12:48] I forgot what I was apologizing about ;) [12:48] oh, the 1.3.1 bzr update in ubuntu today was nice. fixed the ssh+bzr checkout weirdness [12:53] kiko: So, why are you importing Debian? I can see it being useful for package diffs, and so we can see it in the LP UI rather than heading to Debian resources, but you probably have more revolutionary things in mind? [12:54] wgrant, diffs but syncs, mainly [12:54] native syncs [12:54] proper changelog preservation [12:54] the works! [12:54] Yay. [12:54] Ah yes, less nasty syncing. Good idea. [12:56] On a related note... Is Debian bug importing deferred indefinitely? I recall it was targetted for around half-way through last year initially. [12:57] wgrant, no, it's actually being worked on [12:57] there are just some tricky bits that need to be sorted out first.. [12:58] Aha, good, it will be very, very useful. [12:58] Version tracking would make it even more automatically useful, but importing at all is always good. === salgado-afk is now known as salgado === mrevell_ is now known as mrevell-lunch === mrevell-lunch is now known as mrevell === barry_ is now known as barry === bigjools_ is now known as bigjools === cprov-lunch is now known as cprov === mrevell_ is now known as mrevell [16:10] woo hoo. ppa problems. [16:11] lesson of the day: never copy package from one series to another series in the same PPA before the debs are published. [16:13] hyperair: https://bugs.edge.launchpad.net/soyuz/+bug/236407 [16:13] Launchpad bug 236407 in soyuz "Remaining race-condition in the copy UI" [Undecided,Confirmed] [16:13] fascinating. [16:15] well now that i've got both debs and failed builds for the same architectures in the same source package, what should i do? [16:21] hyperair, reupload a new version? :) [16:22] T_T [16:22] just for that? [16:22] D= [16:22] would deleting and copying them back work? [16:23] i mean hardy is fine [16:23] just gutsy and intrepid [16:31] hyperair: copy again, this time including binaries. [16:31] mm i did. now it's failing to upload [16:31] meh. [16:32] hyperair: the previous source with failed-to-upload builds will be superseded. [16:32] i see. [16:35] cyphase: The following source cannot be copied: banshee-1 0.99.3-1ubuntu1~ppa2 in hardy (version older than the banshee-1 0.99.3-1ubuntu1~ppa2 in intrepid published in intrepid) [16:35] whoops [16:36] cprov: The following source cannot be copied: banshee-1 0.99.3-1ubuntu1~ppa2 in hardy (version older than the banshee-1 0.99.3-1ubuntu1~ppa2 in intrepid published in intrepid) [16:36] typed wrong name @_@ [16:38] hyperair: uhm, did you deleted the FTBFS versions in gutsy & intrepid ? [16:39] hyperair: it's a half-bug, it should allow you to copy the same version within the PPA, PPA sources are unique. [16:39] eh no? [16:39] hmm [16:39] so i should delete the failed to build versions in gutsy and intrepid [16:39] then copy back? [16:39] from hardy i mean [16:40] do i have to wait for some time before copying? [16:42] hyperair: yes, that's the workflow [16:42] hyperair: no, no need to wait [16:42] okay. [16:44] this is great. [16:44] it's not deleting now. [16:44] i check the packages. click delete [16:44] and then. [16:44] the status remains "published" [16:44] hyperair: ppa url, please ? [16:44] http://edge.launchpad.net/~banshee-team/+archive [16:46] hyperair: are you adding the 'deletion comment' when deletion packages ? [16:46] yes. [16:47] hey it worked now [16:47] weird. [16:47] how does the cvs imports work exactly? do I paste the url to my svn trunk into the Location for the trunk on my project? [16:47] lzhang, no, visit +source for your trunk series [16:48] cprov: copying the hardy package over to gutsy resulted in failed builds. [16:48] instantly [16:48] it just appeared there. failed. [16:48] kiko: thanks a lot I found it [16:48] hyperair: those are the old builds, don't worry. [16:49] so i just leave them as "failed"? [16:49] lzhang, what project is that, I can approve it immediately [16:50] kiko: coughphp, but I need to do a bit of setup on my subversion repo first [16:50] hyperair: yes, ignore them for now, I can remove them for you later. [16:50] lzhang, okay, ping me when you're ready [16:50] kiko: will do [16:51] cprov: thank you. [16:51] hyperair: the archive will be republished in 10 minutes, and should allow you to install the hardy binaries also in gutsy and intrepid. [16:51] hyperair: you are welcome. [16:52] cprov, how come you know so many things about this PPA machine? [16:52] * kiko is amazed [16:53] kiko: ehe, I wrote all the broken bits, didn't I ? [16:53] hoho. so that's why Ng asked me to speak to you [16:54] cprov, no, IIRC you wrote all the good bits! [16:55] those too, but people usually don't come here to praise me about them ;) [16:56] * laga praises cprov [16:56] i like my PPAs [16:56] * hyperair puts cprov on a pedestal [16:57] PPAs are useful. i certainly wouldn't be able to find anywhere else to host debs. maybe mediafire, but doesn't work for repos [16:57] hyperair: lol, wait until I get your bug fixed. [16:57] heheh [16:57] i'm waiting =) [17:04] hyperair: it's fixed, I will remove the build cruft for you. [17:04] cprov: thanks a lot =D [17:06] hyperair: done, all "green checks" now. [17:06] hyperair: sorry for the inconvenience, I get this bug fixed for the next milestone. [17:07] hyperair: meanwhile, remember to wait the binaries to get published in the archive before you copy them to the other series. [17:08] okay. thank you [17:08] =) === mrevell_ is now known as mrevell [17:21] hi [17:21] how can I merge 2 launchpad users? [17:21] they are both mine [17:22] visik7, https://edge.launchpad.net/people/+requestmerge [17:26] kiko: how long can I expect this svn import to be "Testing" ? [17:27] lzhang, until you tell me you have done everything and give me a URL :) [17:28] kiko: https://code.launchpad.net/coughphp/trunk [17:28] word [17:33] I've merged an anonymous user [17:33] but how to merge registered user ? [17:43] lzhang, we currently have a problem with self-signed https:// hosts [17:43] hmm [17:43] kiko: do I need to move it to non-ssl then? [17:44] lzhang, it would be ideal if you could === kiko is now known as kiko-fud === matsubara is now known as matsubara-lunch === salgado is now known as salgado-lunch [18:11] Hi guys... i have a "small" problem trying to get my blog into planet.ubuntu.com when i checkout bzr says: Permission denied (publickey). === salgado-lunch is now known as salgado [18:45] lzhang_, so! [18:45] santiago-ve, your ssh key isn't in launchpad. [18:45] kiko-fud, i have two keys [18:46] just registered one.. to try on other machine [18:46] kiko-fud: ya I'm waitin on my buddy awbush to move it over to port 80... so still waiting [18:46] lzhang_, gotcha [18:47] kiko-fud, http://imagenes.zarate.net.ve/varias/launchpad1.png [18:47] http://imagenes.zarate.net.ve/varias/launchpad2.png [18:48] tried the same procedure on both machines... and both fail === matsubara-lunch is now known as matsubara [18:51] santiago-ve, it's weird. nobody else has this problem -- how are you trying to check out? [18:52] bzr checkout bzr+ssh://469@bazaar.launchpad.net/~planet-ubuntu/config/main planet-ubuntu [18:52] santiago-ve, uhhhh. what's 469? [18:52] kiko-fud, well my luck this last months havent been good [18:52] kiko-fud, my username [18:52] are you serious? [18:53] https://edge.launchpad.net/~469 [18:53] wtf [18:53] errr... yes [18:53] santiago-ve, that might be the problem [18:53] ugh [18:53] it's a bug if so, but, well.. [18:54] santiago-ve, try bzr branch lp:~planet-ubuntu/config/main [18:54] santiago-ve, you can do a bzr launchpad-login before that [18:54] if the launchpad-login fails, then.. [18:54] kiko-fud, the same: Permission denied (publickey). [18:55] fantastic [18:55] santiago-ve, if you rename your account, does it work? [18:56] kiko-fud, gonna try [18:58] kiko-fud, yep [18:58] its a bug [18:58] kiko-fud: yo I changed it, it is non-ssl now [18:58] santiago-ve, please report it. thanks! [18:58] lzhang_, perfect. [18:58] on my way [18:59] kiko-fud: I gotta run to a meeting, bbl [18:59] kiko-fud: thanks for all your help! [19:00] lzhang_, that repository seems to contain more than one project.. [19:00] trunk/simpletest/simpletest_1.0.1beta2 [19:00] or is that right? === kiko-fud is now known as kiko [19:01] well, here goes [19:01] lzhang_, https://code.edge.launchpad.net/~vcs-imports/coughphp/trunk [19:01] santiago-ve, you know, we should probably ban all-ints names :) [19:01] kiko, will my 469@ubuntu.com be changed to santiago-ve@ubuntu.com automatically? [19:01] santiago-ve, I believe so, yes, though it takes some time to update [19:01] kiko, would be a great practice [19:02] 469 is such a crazy name. :) [19:02] kiko, nea its sexy :D [19:02] math related [19:02] :D [19:02] * santiago-ve hates math [19:03]