[00:33] hmm [00:33] i'm using bzr_access to grant multipule users access to the same repository, any way i can make it play nice with bzr? [00:36] er [00:37] s/bzr\?/CIA/g [00:37] wow i'm a failhouse tonight [00:50] actually [00:50] let me rephrase that [00:51] any way to get bzr-cia working when using bzr_access? === ampelbein is now known as Ampelbein [01:02] Spaz: I don't see why they would be related or interact at all [01:02] lifeless, basically, i want to report commits to the main repo [01:03] sure, I'm just saying I don't see the correlation with bzr-cia [01:03] and bzr_access [01:11] did you have it working with bzr-cia and not bzr_access? === d6g|away is now known as d6g === d6g is now known as d6g|away === abentley1 is now known as abentley === abentley1 is now known as abentley [02:12] abentley: pokpokpoke you know william witteman?! [02:14] * emmajane also has a random bzr question about this thing called "redhat"? [02:15] just ask :P [02:15] for the gitvsbzr thingy someone was complaining that tests are breaking in redhat? Other than, "stop using redhat" is there an answer for that or someone who is somewhat responsible for that? [02:15] * emmajane had to phrase it, lifeless :) [02:15] or figure out how to phrase it rather. :) [02:16] I don't think any of the high-volume contributors to bzr itself develop on redhat [02:16] hrm. [02:16] so the answer is sort of switch to ubuntu? ;) [02:17] * davidstrauss still has 11 failures from selftest on CentOS 5 with bzr + BzrTools [02:17] davidstrauss: dude. ouch. [02:19] emmajane: no, its more - file bugs [02:19] * emmajane nods. kay. thanks, lifeless [02:19] on any platform we would expect [02:19] 'make check' to pass [02:20] win32 has some known issues [02:20] so make that any foss platform :P [02:20] heheh [02:20] bzrtools is part of the ecosystem, but we don't require zero-failures from it at any point in time like we do for bzr itself [02:21] The one that concerns me most is this: FAIL: per_lock.test_lock.TestLock.test_readonly_file(fcntl) IOError not raised [02:21] davidstrauss: don't run the tests as root [02:21] * emmajane blinks. [02:21] all mainline commits to bzr since it moved to pqm are done by pqm only after the test suite passes [02:22] lifeless: very nice :-) [02:22] lifeless: Indeed, the "root" issue makes sense. [02:22] lifeless: This is on my package building VM [02:22] what is pqm? sorry. :/ [02:22] emmajane: its a robot that takes a branch and merges it to the mainline for us [02:22] it enforces the tests-pass policy for us [02:23] lifeless: ah, cool. [02:23] * emmajane nods [02:23] makes sense. [02:23] without requiring every developer to remember, or be trusted to always remember [02:23] or have the reference platform (python 2.4, all dependencies, etc) [02:24] * emmajane nods. [02:25] davidstrauss: so if you can run the test suite (make check) as non-root, I'd love to know about failures [02:25] lifeless: for the bzrvsgit is it then, 'we love you all, but please submit bugs'? [02:25] emmajane: yes [02:25] lifeless: Trying now [02:26] lifeless: cool :) [02:26] lifeless: I was a little bit like, 'redhat? RLY?' [02:26] I'm reasonably sure the redhat packagers run the test suite w/o errors [02:27] and we'd certainly pay attention to any issue [02:27] I will note though, that I can only really check fedora to reproduce things, RHEL being proprietarish and all [02:27] * emmajane nods. kay. I'll check back if I have more useful information, but this is cool [02:27] reason #34 to be fully open source; people can debug on your platform [02:28] I'm heading off for a bit, ciao [02:28] lifeless: thanks muchly, as always. :) [02:29] lifeless: I'm using CentOS === abentley1 is now known as abentley [02:30] * emmajane tries poking abentley again. [02:31] are there any pages on the bzr wiki about google summer of code? [02:31] * thumper waves at emmajane [02:31] thumper: are you doing gsoc? [02:31] * emmajane waves to thumper [02:31] heh [02:31] I think that is a little unlikely [02:32] I'm not a student any more [02:32] nonnono Imean "you" not "you" [02:32] however I know someone who knows students, and I'm trying to pass on details [02:32] emmajane: yes bzr is trying to organise gsoc [02:32] thumper: the org alread has to be approved... [02:32] emmajane: I believe igc is handling it [02:32] lifeless, ah, i got it all working with a ghetto script [02:33] thumper: are you doing it under ubuntu? I think i read that they got in? [02:33] it basically cd's to where the repo is and does bzr cia-submit for all new revisions [02:33] emmajane: I don't know that much about it other than poking people to make sure bzr gets involved [02:33] thumper: I think you might have missed it if you're still poking peoople. :) [02:33] :( [02:34] emmajane: is there a definitive list of approved projects? [02:34] thumper: clearly poking is not so productive. less of the poking, more of the doing. [02:34] emmajane: I know that someone is doing [02:34] yay [02:34] emmajane: and I think that someone is igc [02:34] * emmajane checks the site. [02:35] http://socghop.appspot.com/program/accepted_orgs/google/gsoc2009 [02:35] thumper: idon't see bzr there :( [02:36] I lied, Ubuntu isn't there either. [02:36] maybe that was an application, but not acceptance. [02:36] oh FFS [02:36] :( [02:39] emmajane: I'm at a loss for words [02:39] thumper: that's sad. :( [02:41] Ubuntu was unaccepted, I think [02:41] It was there for a while. [02:41] * emmajane nods to wgrant [02:42] * emmajane tries to figure out how to use irssi. [02:42] which is completely unrelated to anything. [02:43] * thumper uses konversation [02:43] I tried irssi for about two days [02:43] heh [02:43] * wgrant was tempted to move to Quassel, but has been using irssi for almost three years now. [02:43] Originally I used XChat. [02:43] I'm on an EEE and don't have any in-depth trust that it will still be xandros tomorrow so is on her server using irssi. [02:44] and apparently I'm tired and can't figure out which person to speak in. [02:44] or something. [02:54] wgrant: what does [Act: 1,2,4] mean? [02:56] gah [02:56] network manager is running amok [02:56] thumper: did you figure out your adsl stuff? [02:56] thumper: I always have problems with my network dropping my irc connections at home. [02:58] emmajane: yeah, I got a clean cat-5 cable from where the telephone line enters the house, and now goes through a main filter there, all the way to my router [02:58] thumper: nice. [02:58] emmajane: not had any adsl problems since then [02:58] thumper: maybe just generally the internet hates you? [02:59] emmajane: my hardware hates me more than the internet does [02:59] heh [02:59] thumper: also: my wireless card causes kernel panics as of a couple of weeks ago. that was handy while on the road. [03:00] heh [03:00] :( [03:00] hence the eee [03:00] I'm going to make some cookies now [03:00] peanut brownies [03:00] mmmcookies [03:00] thumper: enjoy :) [03:00] emmajane: I will [03:02] ´ [03:02] hrm. /me spams the channel while she messses up her meta keys [03:03] brb [03:14] there is some confusion re: gsoc and ubuntu [03:14] wait a few days [03:15] yay! control-n control-p is your friend. [03:15] emmajane: [Act: 1,2,4] shows the windows that have activity. [03:16] Bold white means there's some new line of conversation. [03:16] * emmajane nods to wgrant [03:17] my alt key wasn't mapped corectly but someone at the sprint just toldme about control-n control-p [03:17] from what I've heard (on the GSoC list) Ubuntu has withdrawn from GSoC [03:18] jelmer: 'there is some confusion' [03:18] jelmer: 'wait a few days' [03:18] lifeless: that's different from "Ubuntu has withdrawn" [03:18] jelmer: yes [03:19] lifeless: The source here is Leslie, who is in charge of GSoC [03:19] jelmer: I know [03:19] jelmer: And yet, I'm saying that there is some confusion and you should wait a few days. [03:20] not being the ubuntu gsoc contact myself, or leslie, I can't really say more [03:22] * emmajane really ought to read some of the backlog of the gsoc list. [03:23] s/say more/know more/ [03:24] hrm. [03:24] on the list from lh "Lime Survey has replaced Ubuntu." [03:24] yes [03:25] lifeless: but that is not necessarily the case? [03:26] well clearly it is from LH's perspective, I have *no* idea on the Ubuntu front, other than people thinking we're in [03:26] at the least there is confusion, at worst there is utter confusion and things will change [03:26] its the weekend for doko, he's probably skiing or something [03:27] * emmajane nods. [03:27] the only thing I would bet on right now is that there is a lot of confusion [03:28] lifeless: which is often not a good way to start a projec.t ;) [03:28] there's nothing on ubuntu-devel, ubuntu-soc etc indicating a pull-out [03:28] * emmajane nods. confusion. [03:37] emmajane: as best I know, ubuntu was accepted [03:38] otoh, I screwed up getting an application for bzr in on time, so we not, at least as a stand-alone project [03:38] s/we/we're/ [03:39] igc: aww :( [03:39] it's very likely we'll do something though - either as part of ubuntu or just separately [03:39] igc: but we don't know how many slots yet, do we? [03:40] * emmajane is helping out with the drupal one too very very peripherally. [03:41] emmajane: we do have a good list of project ideas put together - see http://bazaar-vcs.org/SummerOfCode2009 [03:42] emmajane: so if you know anyone interested in doing those, students or otherwise, ask them to get in contact with me [03:43] * emmajane reads. [03:43] igc: drupal also has a VCS proposal, but I'm not sure what its status is. [03:50] igc: have you been talking with any students yet? [03:51] (regardless of ubuntu/bzr acceptance) [04:01] emmajane: That name doesn't ring a bell... [04:01] abentley: formerly william o'higgins; friend of mike audet [04:01] abentley: not sure on the spelling of mike's last name. [04:01] Oh, sure. He's good people. [04:02] abentley: we went to uni together. [04:02] Cool. [04:02] Which uni? [04:02] abentley: I'm actually in TO this weekend for a sprint. [04:02] abentley: UofT. [04:02] Yeah, I went there too, that's how I met Mike, and therefore William. [04:03] abentley: small. world. :) [04:06] abentley: I was there 95-99. [04:08] emmajane: So how'd you find out I know William? [04:08] abentley: he left a comment on my bzrvsgit post [04:09] abentley: he likes bzr because of you [04:09] abentley: and also because bzr is awesome [04:09] Ah. [04:09] Cool. [04:09] yah [04:10] and then someone else almost at the same time did one of those, You're both in canada do you know each other? things and... well..almost. :) [04:12] Just posted fresh RHEL/CentOS packages for 1.13: http://fourkitchens.com/blog/2009/03/22/bazaar-113-rpms-rhel-5-centos-5 [04:15] davidstrauss: did the tests pass then? [04:15] lifeless: Still getting the file size mismatch failures [04:15] lifeless: The scarier fails went away when running as non-root [04:16] file size mismatch -> please do file a bug === Ampelbein is now known as ampelbein [04:48] night all! [04:57] So everything i read about bzr says that it handles file renames with mv, but that does not seem to be the case in my branches [04:58] lamalex: what do you mean by 'handles file renames with mv' [04:59] lifeless: as in say foo.c is in my bzr tree, and i rename it to bar.c, if I do bzr stat the file is missing [04:59] but the bzr docs, and things people tell me, say that bzr should reconizze the rename [05:00] and i should see foo.c => bar.c [05:00] lamalex: bzr doesn't do that; either do 'bzr mv foo.c bar.c' to do the move, or do 'bzr mv --after foo.c bar.c' to tell it afterwards [05:00] the docs definitely don't claim a heuristic to figure that out [05:01] what does it mean in http://bazaar-vcs.org/BzrFeatures [05:01] first bullet in Intuitive [05:01] lamalex: it means if you have a _branch_ you can rename the branch with mv [05:01] ahh [05:01] ok [05:02] and cp will make a new branch [05:02] right [05:02] and rm will remove a branch [05:02] i knew that [05:02] thanks [05:04] there is a plugin that will apply a rename-detecting heuristic [05:04] abentley has been working on it, I don't know what its called or where to get it though :P [05:04] oh yeah? [05:04] ill google [05:04] thanks [05:05] he was doing it for the tarball import stuff; [05:05] also, has there been some LP weirdness recently? [05:05] its been getting stuck on "Transfering:Walking Content" [05:06] a bunch of people in my project have seen it [05:06] in bzr 1.13 we changed the pull/push logic to allow streaming which is much faster [05:06] when bzr 1.13 is rolled out on the server you'll experience this [05:07] but with 1.6(?) on the server pushes take forever? [05:07] is the solution to just wait for the push to finish? [05:07] 1.11 on the server, and yes [05:08] ok [05:08] thanks for your help [05:08] the rollout is next week [05:08] awesome [05:36] give a file X in mu current directory, how do I tell if its under revision control or not? [05:37] bzr status X? [05:37] or bzr inventory | grep X if you want a brute force approach, or st X doesn't do what you want [05:37] status says nothing, just returns [05:39] "bzr inventory | grep X" just prints out X which I assume means that it is under revision control [05:42] yes [05:43] where do i lodge feature requests? [05:43] bugs.launchpad.net/bzr [05:44] m3ga: If X is versioned and unchanged, "bzr status X" will print nothing and return. [05:48] exactly which is why i submitted this feature request: https://bugs.launchpad.net/bzr/+bug/346624 [05:48] Ubuntu bug 346624 in bzr "Feature request : fileinfo command" [Undecided,New] [05:48] ah there it is [05:55] Well, actually, "bzr st ignored_file.pyc" will also exit with no output. Soo.. === Spaz is now known as H1tler === H1tler is now known as Spaz [05:57] exactly. in the fetuare request I asked for four different catefories of output. there may be more. [06:20] why does the lp: transport need my launchpad login? [06:21] and then it goes ahead and does what i want anyway? [06:31] m3ga: It needs your Launchpad username so that you can write to Launchpad, and also to use bzr+ssh to make things much faster. [06:31] But it will fall back to read-only HTTP if you don't tell it your username. [07:54] m3ga: because we haven't made every ui command indicate to the directory service if it needs a writable transport [07:54] m3ga: and launchpads http mirrors are read only [07:55] lifeless: i'm just a dumb-fsck user and you're telling me about lowlevel implementation details? :-) [07:56] m3ga: asif :P [07:57] i think this is actually a real failing of all dvcs. the problem is very hard and that difficulty ends up leaking into the user interface [07:57] m3ga: We can do better here, but ETIME. [07:58] this part isn't particularly hard - merge (for instance) never needs a writable remote transport, push does, branch's target does etc [07:58] you're already doing well but i'm gonna keep beating you up about it, because too many people (esp git users) don't even realise that this problem exists [07:59] oh sure [07:59] long as you keep using bzr you can keep beating me up :) [07:59] i did a fast-import into git today. a whole fscking world of surprises :-) [08:02] indeed [08:02] but it gives them to you fast [08:03] ciao [08:37] lifeless: I think I fixed that roughly by the time you mentioned it and laughed a bit doing that thinking about *you* laugh seeing my mistake :-P (The less funny part was realizing this wasn't covered by a test...) [08:44] lifeless: next step as I see it: add a test runner registry to bzr and register subunit/forked/subprocess there from a plugin that could die when each runner find its natural home, that left the problem of additional parameters to these runners open though. [08:59] vila: I've pushed new stuff; I don't see subunit/forked/subprocess as runners... [10:12] Hello, I have been experiencing some strange slowness (that wasn't present before) when trying to push a branch to lp. [10:12] I am using bzr-1.13rc1 (from debian unstable). [10:14] The local branch is in a shared repository with rich-root-pack format. [10:15] alf: there is new streaming code in 1.13 [10:15] When trying to push a branch to lp which includes small changes (eg two added 100-line files) the push takes about 2:30 minutes with constant ~50 kb/s upload [10:15] alf: which lp hasn't deployed yet, but bzr gets too far down the code path to use the local-filesystem optimised path [10:16] alf: on thursday there is a rollout of 1.13 to lp, it will get massively faster for you then [10:16] [10:17] lifeless: ok, thanks! I will just be patient then :) === Ampelbein is now known as ampelbein === d6g|away is now known as d6g === epsilon_0 is now known as princess_liah === princess_liah is now known as epsilon_0 === abentley1 is now known as abentley === nekohayo_ is now known as nekohayo === d6g is now known as d6g|away === abentley1 is now known as abentley === abentley1 is now known as abentley [15:25] I don't see subunit/forked/subprocess as runners... [15:25] Why ? [17:12] lifeless: nm, just got you rmail [17:15] arrr [17:16] LarstiQ: yarrrrrr [17:18] jelmer: what identi.ca client do you use? [17:18] * LarstiQ tried TwitterVim [17:19] LarstiQ: I use the haskell one and gwibber [17:19] twidge [17:20] jelmer: ooh, twidge is actually in lenny === abentley1 is now known as abentley === d6g|away is now known as d6g [19:57] leon@bespin:~/Desktop/bzrtools$ bzr shell [19:57] bzr: ERROR: [Errno 2] No such file or directory [19:57] What am I doing wrong? [19:57] ( Er.. It works in my home ¬¬ ) [19:58] Leon_Nardella: does ~/.bzr.log provide a hint? [19:58] LarstiQ: I went to my home then back to the folder I was and now it's working.. Really weird. [19:59] Leon_Nardella: did the directory still exist? [19:59] Yes. [19:59] Leon_Nardella: ie, did you delete or mv it between cd'ing there and starting bzr shell? [19:59] (so, inode, not path) [19:59] No, no. [20:00] hmkay [20:00] http://paste.ubuntu.com/135571/ [20:00] The interesting part from the log. [20:01] http://paste.ubuntu.com/135572/ [20:01] Now it's like this. [20:02] as I thought, it's the 'get current working directory' call that returns 'No such file or directory' [20:17] hi [20:17] I had a typo on a commit message, can I fix it somehow? [20:17] I still didn't push the message [20:17] s/message/revision/ [20:21] Turl: bzr uncommit; bzr commit [20:22] thanks LarstiQ [20:22] LarstiQ: it doesn't change the code, does it? [20:23] Turl: no, it leaves the tree state the same, but sets the branch to point at a prior revision [20:24] Turl: so if you then commit, you end up with a revision where only the commit message is different (well ok, the timestamp, revision id, maybe committer name if you changed that inbetween) [20:24] Turl: basically, you can fix mistakes like that :) [20:24] ok LarstiQ, thanks :) [20:35] so for my development version of loggerhead, viewing a revision page that adds 9000 files doesn't stress loggerhead too much [20:36] ff, on the other hand... [20:38] mwhudson: yeah. I ditched ff(3) because it made my entire system unusable (sloow ssd + sqlite fsync + ext3 = hsd) [20:38] LarstiQ: what do you use instead? [20:39] mwhudson: opera for now, since I have two colleagues idolizing it. [20:39] I don't care too much for the browser, but at least the rest of my system is usable. [20:40] LarstiQ: believe it or not, opera is real slow in here :p [20:40] LarstiQ: how does it do on the "fills one with burning rage" metric? [20:40] that's why I use firefox [20:40] mwhudson: it scores a 3. But I haven't been using it very long. [20:41] the first versions of ff3 were so fast [20:41] what happened? [20:42] mwhudson: let me look up two urls that hit me [20:46] meh, logging [20:46] mwhudson: https://bugs.edge.launchpad.net/ubuntu/+source/linux/+bug/317781 I believe [20:46] Ubuntu bug 317781 in linux "Ext4 data loss" [High,Fix released] [20:46] mwhudson: and from there a mozilla person blogpost [20:47] mwhudson: http://shaver.off.net/diary/2008/05/25/fsyncers-and-curveballs/ [20:49] LarstiQ: that doesn't really explain why ff3 has gotten worse IME [20:50] maybe it's just that my awesomebar.db or whatever is larger now [20:50] mwhudson: I don't know what got introduced when. [20:51] mwhudson: for me, I've only been using this Acer Aspire One since the second half of January [20:51] and I think I know why Acer ships FF2 by default.. [20:51] mwhudson: I was able to speed things up a lot of trimming and vacuuming the SQLite DBs. [20:51] s/of/by/ [20:52] I wonder if Firefox can be taught to use a RDBMS that doesn't suck. [20:52] wgrant: hmmm [20:53] wgrant: i think an oracle license for every ff user would get a bit expensive! [20:53] * mwhudson runs away, terribly fast [20:53] * wgrant trips mwhudson. [20:54] * LarstiQ dumps 'no we don't store support vectors for linear kernels' on mwhudson [20:55] I think they have been working on the performance of the awessomebar for ff3.5 [20:55] At least Firefox never gets as slow as OOo. [21:30] hi all, if I wanted to use bzr to create a trunk and some branches.. would I have to physically copy files to a directory like "myRepository/trunk" and then issue bzr init? [21:30] is that how it works? [21:30] also you think its a good idea to setup a "bzr" user on my CentOS machine so people with different accounts and userids can access it? [21:31] anyone? :) [21:35] wow no one helps out in this channel eh? [21:35] silence is the key [21:35] CBro2007: um [21:36] for the first, yes, i think you're right [21:36] ok [21:36] for the second, well, it depends what you're trying to acheive [21:36] well this is what I want... [21:36] I got a bunch of project files on a shared server that I want to share with 3 other developers [21:37] they all have their own unix accounts on this machine [21:37] with you so far [21:37] am wondering if there will be any problems if I am the owner of the trunk etc [21:37] ah [21:37] like the file ownerships would be in my name because I did an init-repo or something [21:38] i think you could probably use a unix group for this [21:38] so I want to give some thought to the dir sructure before I launch off [21:38] I have setup a UNIX group to which we all belong [21:38] :) [21:39] so if I was to do this... [21:39] then, so long as you use g+s or whatever the chmod thing is, i think it'll work alright [21:39] bzr init-repo myRepository [21:40] then create a dir for my project under there [21:41] and then under that project create a dir called TRUNK to which I copy all the original project code [21:41] ? [21:41] right so far? [21:41] yes [21:41] ok... [21:41] often times people create a repo per user, rather than all sharing the same repo [21:42] and in addition have a central repo from where releases/rollouts are made [21:42] mwhudson: is this recommended? [21:42] it depends on your workflow really [21:42] mwhudson: I suppose I am not super sure as to how we will MERGE our changes etc === ampelbein is now known as Ampelbein [21:42] mwhudson: just don't want people stuffing up the GOLD COPY you know [21:42] :) [21:43] CBro2007: one of the nice things about bzr is that it makes many decisions that are technical with, say, svn, social [21:43] how do you mean sorry? [21:43] CBro2007: unfortunately this means that the answers to many questions become "um, well, it depends" :-p [21:44] you mean it supports a lot of kind of workflows? [21:44] CBro2007: what version control do you use now? [21:44] CBro2007: yes [21:44] used SVN and CVS in the past [21:44] CBro2007: for "protecting the gold copy", there is PQM, which is a robot that merges branches for you [21:44] want to move away from it to the distributed version control and enforce it amongst other developers alike [21:44] and it can be configured to run tests before it accepts a branch [21:45] Nah I would like to do that manually [21:45] hang on, there must be a wiki page about this stuff [21:45] myself [21:45] ok [21:45] I think I remember reading that in the user guide.. where you have a gateway keeper or something that can automate the merges [21:45] I just want it a bit simpler where developers let me know when they think their branch is ready for a merge to the trunk [21:46] CBro2007: http://bazaar-vcs.org/Workflows [21:46] and then I can do it [21:46] CBro2007: sounds like http://bazaar-vcs.org/Workflows#head-f5c9069eb3a6991f31e6b909dbb3fb12a45f34a5 [21:47] you mean centralized? [21:48] I think I want the decentralized with human gatekeeper [21:48] I like that workflow best [21:48] CBro2007: no, "decentralized with human gatekeeper" [21:48] right [21:48] From what you said - I think Decentralized with human gatekeeper [21:48] that seems to be basically exactly what you're describing [21:48] Yeah :) [21:48] haha [21:48] thats what I want [21:48] :) [21:49] wow now how do I set this model up [21:49] this is exactly what I want to do [21:50] but the model doesn't show me a step by step scenario of how to do it :) [21:51] was wondering if I need to even use launchpad or can I just merge stuff locally [21:51] as it will all really just be UNIX accounts on the same machine [21:51] you certainly don't need to use launchpad [21:51] the docs are probably tilted towards open source-y things [21:53]