2008/05/31/#bzr.txt

[00:25] <lifeless> fdv: So svn has multiple disk formats like most vcs's, but unlike bzr it silently upgrades on you - which is why you couldn't use the older client after using the newer one
=== mw is now known as mw|out
[02:53] <rockstar> Anyone know why I would be getting "ValueError: I/O operation on closed file" while trying to branch from a launchpad branch?
[03:54] <jam> rockstar: that is usually a secondary error
[03:54] <jam> it means we are doing something like 'flush()' on a closed file
[03:54] <jam> usually we get that after something else has gone wrong
[04:02] <rockstar> jam, seems like I'm getting it trying to branch from all lp branches.  A friend of mine is having the same problem.
[04:02] <jam> rockstar: what version of bzr, and what platform?
[04:03] <rockstar> 1.5, on linux.
[04:03] <rockstar> It's the same system I have launchpad.dev on
[04:04] <jam> rockstar: have you tried running with '--no-plugins' to see if that fixes it?
[04:22] <m3ga> ping lifeless
[07:05] <lamalex> Hey pals, I'm having problems pushing to my LP branch
[07:06] <beuno> lamalex, specifics?
[07:06] <lamalex> LP has to use bzr push lp:~alexlauni/do/do-alauni, but that doesn't work due the http mkdir() erro
[07:06] <lamalex> so I'm trying to do bzr push sftp://bazaar.launchpad.net/~alexlauni/do/do-alauni/
[07:06] <beuno> lamalex, bzr launchpad-login yourlpid
[07:07] <beuno> that will set it to use bzr+ssh
[07:07] <lamalex> that's a beautiful thing
[07:07] <beuno> ain't it?
[07:07] <lamalex> now i have bzr: ERROR: Target directory lp:~alexlauni/do/do-alauni already exists, but does not have a valid .bzr directory. Supply --use-existing-dir to push there anyway
[07:07] <lamalex> should I just use that switch?
[07:08] <beuno> uhm, that's odd, but yes
[07:09] <lamalex> heh
[07:09] <lamalex> it's pushing. we
[07:09] <lamalex> 'll see how it gos
[07:09] <beuno> :)
[07:09] <beuno> you're using bzr 1.3, right?
[07:10] <lamalex> indeed
[07:10] <lamalex> 1.3.1
[07:10] <beuno> I believe newer versions do a better job at warning you about pushing without a lp-login set
[07:10] <lamalex> are they in the bzr ppa?
[07:10] <beuno> I'd recommend updating from the PPA (https://launchpad.net/~bzr/+archive)
[07:10] <beuno> :)
[07:10] <lamalex> :) great minds right?
[07:11] <beuno> heheh, right
=== gour is now known as gour|afk
[08:14] <dennda> Hi. I try to bzr ignore a Django settings.py file. I got this django project on several machines and the settings.py differs slightly on each host. I still want it to be version controlled per host but I don't want to sync it between the hosts. (And if that is not possible, just plain ignore it completely)
[08:14] <dennda> However, bzr ignore settings.py gives me this:
[08:14] <dennda> Warning: the following files are version controlled and match your ignore pattern:
[08:14] <dennda> olympic/settings.py
[08:15] <Peng> dennda: You're already versioning settings.py, so ignoring it does nothing.
[08:16] <dennda> Peng: So how do I keep the file but disable versioning for it?
[08:16] <Peng> dennda: You can't.
[08:16] <Peng> dennda: You could commit to a separate branch with a default settings.py, then "bzr merge" it into each machine's branch.
[08:16] <dennda> Ok what about removing the file with bzr remove, creating it again and ignoring it?
[08:17] <dennda> I commit early and often, that'd just be a pain
[08:17] <Peng> dennda: Sure, you could do that.
[08:17] <dennda> Then I will
[08:17] <dennda> brb, breakfast :)
[08:25] <bob2> dennda: bzr rm --keep settings.py
[08:31] <dennda> thanks bob2 and Peng. I hope it worked
[08:36] <dennda> bob2: I did that locally, committed and pushed to a remote branch where it deleted not only the version control but the file itself, too!
[08:37] <luks> bzr doesn't modify remote working trees
[08:37] <luks> if you have a checkout there, you need to `bzr update`
[08:37] <luks> er, wait
[08:37] <luks> ignore the second line :)
[08:38] <luks> it couldn't remove the file, because it doesn't touch remote working trees
[08:39] <dennda> I did update and the file was gone
[08:40] <luks> I think update would remove it only if it's identical to the versioned version
[08:40] <luks> so bzr revert -r X settings.py
[09:47] <Arby> I'm having a problem pushing to a branch. bzr push returns ERROR: Cannot lock LockDir(http://bazaar.launchpad.net/%7Ekubuntu-users/jockey/jockey-kde/.bzr/branch/lock): Transport operation not possible: http does not support mkdir()
[09:47] <Arby> this seems to be a case of bug 156462
[09:47] <ubottu> Launchpad bug 156462 in bzr ""cannot lock LockDir" on update over http" [Undecided,New] https://launchpad.net/bugs/156462
[09:47] <Arby> I've tried all the things suggested in that thread
[09:47] <Arby> without success
[09:48] <Arby> can anyone suggest what I'm doing wrong
[09:49] <luks> you can't push over http
[09:50] <luks> either use sftp: or bzr+ssh: link, or bzr launchpad-login and then lp: link
[09:51] <Arby> I've tried sftp and bzr+ssh both fail with Permission denied (publickey).
[09:51] <Arby> but I do have a public key on launchpad
[09:52] <Arby> I even tried to upload it again just to be sure
[09:52] <luks> does 'sftp bazaar.launchpad.net' work?
[09:52] <mwhudson> is it the same one as one the machine you're trying on? :)
[09:52] <Arby> mwhudson: yes
[09:52] <mwhudson> could this be a username issue?
[09:52] <Arby> luks checking
[09:53] <mwhudson> actually, i bet that's it
[09:53] <Arby> luks: no it doesn't
[09:53] <Arby> mwhudson: what is?
[09:53] <mwhudson> Arby: what's your launchpad id?
[09:53] <Arby> rbirnie
[09:53] <mwhudson> does 'sftp rbirnie@bazaar.launchpad.net' work?
[09:54] <Arby> checking
[09:54] <Arby> mwhudson: yes it does
[09:55] <mwhudson> Arby: cool
[09:56] <Arby> so because my user id isn't the same as the owner of the branch it fails?
[09:56] <luks> no, it's because your local username is not the same as your launchpad id
[09:56] <Arby> although I joined the team that owns the branch
[09:56] <Arby> ah ok
[09:56] <mwhudson> this doesn't really have much to do with launchpad
[09:57] <Arby> so what is the solution?
[09:57] <luks> bzr push bzr+ssh://rbirnie@bazaar... should work
[09:57] <Arby> trying
[09:58] <Arby> victory :)
[09:58] <Arby> is there a setting somewhere to prevent this happening in future?
=== brilliantout is now known as brilliantnut
[09:58] <luks> .ssh/config
[09:58] <luks> bzr launchpad-login and then use only lp: urls
[10:00] <Arby> luks: sorry I don't understand. .ssh/config doesn't exist
[10:00] <Arby> on this machine
[10:00] <luks> create it, and add this:
[10:00] <luks> host bazaar.launchpad.net
[10:00] <luks>     User rbirnie
[10:01] <luks> (.ssh/config is supposed to be ~/.ssh/config, just to be clear :))
[10:01] <Arby> ok, thanks for the help :)
[10:29]  * mwhudson attempts to be a mere user of launchpad, and fails
[10:29] <mwhudson> spiv, jml: here?
[10:39] <mwhudson> oh never mind
=== gour|afk is now known as gour
=== yacc_ is now known as yacc
[15:16] <philipp_> `bzr: ERROR: Transport error: Server refuses to fullfil the request` what does this mean? tried to do `bzr branch http://bzr.debian.org/apt/debian-sid apt-debian-sid`
[15:17] <dato> philipp_: try bzr branch nosmart+http:://bzr.debian.org/...
[15:19] <philipp_> k, this works, thanks
[15:21] <philipp_> hm, this isn't documented in the man page
[15:23] <dato> philipp_: maybe/probably because it's a bug in bzr
[15:25] <philipp_> http://paste.debian.net/4760/ as is this?
[15:26] <dato> philipp_: no, not that; you can't checkout over a read-only transport.
[15:27] <philipp_> k, but a simple error message would be nice ;-)
[16:17] <RainCT> Hi
[16:18] <RainCT> Can someone help me? I'm getting this error: https://code.edge.launchpad.net/~revu-hackers/revu/trunk/ and I can't get it to work on LP in any way, but locally it works
[16:18] <RainCT> (the guys in #launchpad couldn't help :()
[16:28] <Peng> RainCT: Tried pushing over sftp?
[16:30] <RainCT> Peng: yes
[16:30] <RainCT> Peng: and also to lp:~rainct/+junk/test
[16:31] <Peng> ...And he's also tried with bzr 1.3.1 and 1.5. Launchpad is still on 1.3, of course.
=== kiko is now known as kiko-afk
[17:44] <RainCT> Peng: uhm.. weekend? :P
[17:46] <Peng> RainCT: Yeah. Star Trek is on TV today!
[17:48] <RainCT> Peng: heh. where?
[17:55] <Peng> RainCT: Central Florida, U.S., North America, Earth, ... Local CW affiliate.
[17:55] <Peng> RainCT: It's Star Trek TOS.
[17:57] <RainCT> ah, thought it might be on german tv (as your surname is german) :P
[17:57] <Peng> It is? I thought it was Scandinavian.
[17:57]  * Peng shrugs.
[17:59] <RainCT> heh well might be :P
[17:59] <RainCT> s/is/looks ;)
[18:00] <Peng> What's it mean in German?
[18:06] <RainCT> Peng: Dunno, it might pretty well be Scandinavian :P. «Nord» is «North» and about «hoff» I'm not sure, Firefox says that it's correct but I don't know what it means (I first thought it's «courtyard» but that's «hof»)
[18:09] <Peng> RainCT: I think it's supposed to mean "north house". http://www.answers.com/topic/nordhoff says it's North? German, "from Middle Low German nord- ‘north’ + hof ‘farmstead’, ‘manor farm’".
[18:09] <RainCT> ah, I was right then :)
[18:10] <Peng> :)
=== fullermd_ is now known as fullermd
[18:29] <bjdooks> I'm trying a bzr push, using sftp, but after it asks for my password it just hanfs there and doesn't seem to be doing anything. anyone help?
[18:30] <Peng> bjdooks: You're sure it's not just slow?
[18:32] <bjdooks> ah, it just bombed out with a  Could not acquire lock LockDir(
[18:32] <bjdooks> hmm, how do I clean out the lock?
[18:32] <RainCT> bjdooks: bzr break-lock <branch url>
[18:32] <RainCT> (where <branch url> is sftp://..., bzr+ssh:// or lp:...)
[18:33] <bjdooks> right
[18:33] <bjdooks> held by ben@ivanova on host ivanova [process #19691]
[18:33] <bjdooks> locked 1731 hours, 31 minutes ago? [y/n]: y
[18:42] <bjdooks> ffs, Could not acquire lock LockDir again
[18:44] <bjdooks> hmm, deleting .bzr/repository/lock/held worked
[18:53] <Peng> bjdooks: 1731 hours ago? You sure it's not still being used? ;)
[19:06] <bjdooks> Peng: i think it was a failed session, my laptop doesn't usually get to stay on that long
[19:07] <bjdooks> right, now to work out what completely broke sftp support on my desktop
[19:07] <bjdooks> bzr: ERROR: exceptions.AttributeError: 'SSHSubprocess' object has no attribute 'get_name'
[19:08] <Peng> bjdooks: That's been fixed.
[19:08] <Peng> bjdooks: A recent release of Paramiko (1.7.3?) changed the API a bit, I think.
[19:08]  * bjdooks wonders if debian has caught up yet
[19:09] <bjdooks> hmm, downloading 65.6MB of updates...
[19:12] <bjdooks> so far I like bzr far more than git
[19:16] <Peng> 65.6 MB of updates to what?
[19:16] <bjdooks> my debian machine... looks like that's fixed bzr
=== gour is now known as gour|afk
[20:28] <beuno> hrm, more people seem to be getting "ValueError: I/O operation on closed file
[20:28] <beuno> on LP...
[20:28] <Peng> Yeah, on sys.stdout.
[20:29] <beuno> Peng, any idea what triggers it?
[20:29] <Peng> beuno: Writing to sys.stdout after it's been closed.
[20:29] <Peng> Or, doing any other I/O operation, I suppose.
[20:30] <RoAkSoAx> beuno, ok
[20:30] <beuno> Peng, http://pastebin.ubuntu.com/16005/
[20:30] <beuno> and http://pastebin.ubuntu.com/16006/
[20:30] <RoAkSoAx> beuno, if you need more info, i just upgrade my system to the latest upgrades available for HH
[20:31] <RoAkSoAx> including the upgrade fixing the ssh key vulnerabilities
[20:32] <beuno> RoAkSoAx, could you try the checkout again, but with  --no-plugins?
[20:32] <Peng> The warning and error are coming from the server.
[20:32] <Peng> The client's plugins are irrelevant.
[20:32] <beuno> yeah, which is odd considering it's LP
[20:32] <Peng> I'm pretty sure that area of the code has been changed since 1.3.
[20:32] <beuno> rockstar had the same error yesterday
[20:32] <Peng> Like, the warning was removed, which would make the error go away.
[20:33] <Peng> LP will upgrade their bzr installation soon.
[20:33] <Peng> Anyway, I was gonna go 5 minutes ago.
[20:34] <beuno> RoAkSoAx, can you try "bzr branch" instead of checkout?
[20:34] <beuno> Peng, thanks  :)
[20:34] <Peng> beuno: I got the same thing on a "bzr branch", but I thought it was because the branch is broken.
[20:34] <RoAkSoAx> beuno, ok i will
[20:35] <RoAkSoAx> same thing
[20:35] <RainCT> <persia> I found it worked for me with http pull and bzr+ssh push, although it took a couple  tries.
[20:35] <RoAkSoAx> http://pastebin.ubuntu.com/16053/
[20:36] <RainCT> (persia, sistpoty and me have the same problem from paste 16005)
[20:36] <beuno> ah, right, should would work
[20:36] <Peng> The server-side forgot to unlock something, and a warning was emitted and it was automatically unlocked when it was garbage collected (which isn't done it anymore), but stdio had already been closed, so it tracebacked.
[20:36]  * Peng wanders off.
[20:37] <beuno> RoAkSoAx, bzr branch http://bazaar.launchpad.net/~planet-ubuntu/config/main planet-ubuntu
[20:39] <RoAkSoAx> beuno, seems to be working now
[20:39] <beuno> RoAkSoAx, good, you should be able to push back with bzr+ssh
[20:40] <RoAkSoAx> ok thanks beuno :)
[20:40] <beuno> RoAkSoAx, thank persia  :)
[20:40]  * beuno reports the bug
[20:42] <RoAkSoAx> beuno, how should i push it, bzr push bzr+ssh://yourusername@bazaar.launchpad.net/~planet-ubuntu/config/main  instead of doing a bzr commit?
[20:44] <beuno> RoAkSoAx, you should commit first
[20:44] <beuno> then, yes, exactly like that
[20:44] <RoAkSoAx> ok thanks;)
[20:46] <beuno> RoAkSoAx, np
[20:48] <beuno> RoAkSoAx, bug #236380
[20:50] <RoAkSoAx> beuno, ok cool ;)
[20:57] <pygi> poolie, you around?
[21:01] <Peng> push --remember so that you don't need to specify it.
[21:01] <Peng> It's also good to pull over bzr+ssh...
[21:06] <RoAkSoAx> beuno, http://pastebin.ubuntu.com/16058/
[21:21] <RoAkSoAx> RainCT, how did you push to planet, somthing like this:  bzr push bzr+ssh://andreserl@bazaar.launchpad.net/~planet-ubuntu/config/main
[21:21] <beuno> RoAkSoAx, try again
[21:21] <RainCT> RoAkSoAx: I have the problems with another branch, but yes, that should be right
[21:21] <RoAkSoAx> ok thanks guys :), really appreciate it :D
[21:22] <beuno> RoAkSoAx, I was reconciling the branch to see if there where any problems
[21:22] <RoAkSoAx> beuno, ok ;), it worked now thanks
[21:24] <RoAkSoAx> well, thanks guys, I'm going to watch the Spain - Perú soccer match, bye all :)
=== blueyed_ is now known as blueyed
=== pbor is now known as pbor|out
[22:26] <PriceChild> Hey there, this is a support channel, not just development right?
[22:26] <beuno> PriceChild, yeap yeap
[22:28] <PriceChild> Great :D Well I'm wondering whether this error: http://pastebin.ubuntu.com/16068/ is something I should be asking you guys for help with, or canonical to fix with launchpad? I am using 1.5.0-1~bazaar1~hardy1 from the bzr ppa on launchpad.net
[22:30] <beuno> PriceChild, I reported that error a while ago, bug #236380
[22:30] <beuno> it's a problem with bzr on LP
[22:31] <beuno> the current workaround I can give you, is branch over http, and then push bzr bzr+ssh
[22:31] <PriceChild> Thanks beuno. I upgraded to that ppa version because the one in hardy didn't work which was annoying.
[22:31] <beuno> yeah, LP is still on 1.3, which might be what is triggering that bug
[22:43] <RainCT> good night
[22:50] <emgent> heya
[22:59] <PriceChild> beuno: ok so i branched it, how do i commit it back to the right place? if i do bzr commit -m "message" it gets commited to the directory its in atm :)
[23:00] <beuno> PriceChild, right, now that you've committed, you have to push your changes back with:  bzr push location
[23:00] <beuno> (in location, use bzr+ssh URL)
[23:01] <PriceChild> great, thanks
[23:02] <beuno> PriceChild, you're welcome  :)
[23:03] <PriceChild> *crosses fingers*
[23:03] <PriceChild> tis looking good...
[23:03] <pygi> PriceChild, you bugging people again? :D
[23:04] <PriceChild> pygi: you know what I'm like :D
[23:05] <PriceChild> launchpad is weird, said i pushed 29 minutes ago. *waits for planet.ubuntu/com to get updated*
[23:12] <Daviey> PriceChild: you changed your head?
[23:13] <Daviey> maybe your browser is caching the old one, i'm seeing a newish one
[23:13] <beuno> PriceChild, the planet gets updated via a cronjob, so expect it to take a while
[23:13] <PriceChild> Daviey: supposedly can take 2 hours to update
[23:14] <PriceChild> what he said
[23:14] <pygi> PriceChild, haven't I told you that having your head scrambled is a bad thing? :P
[23:17] <PriceChild> Yay all done, thanks again beuno!
[23:17] <beuno> PriceChild, :)
[23:17] <PriceChild> pygi: shh :P
[23:18] <pygi> PriceChild, it is rude to shhh people!
=== J-Unit is now known as jdong