2012/01/24/#bzr.txt

[00:16] <sidnei> anyone around to help me pushing a bzr branch to github?
[00:16] <sidnei> im getting an error: http://paste.ubuntu.com/814914/
[00:17] <sidnei> jelmer, ^
[00:28] <poolie> hi sidnei
[00:28] <sidnei> heya poolie
[00:28] <poolie> i don't recognize that
[00:28] <poolie> suggest you just file a bug
[00:29] <sidnei> oki.
[00:29] <sidnei> i assume if i do a fast-export it won't be compatible with doing dpush after that?
=== ISK_ is now known as ISK
[00:46] <jelmer> sidnei: I would suggest trying lp:bzr-git (asuming you're running precise)
[01:09] <sidnei> jelmer, as in, the branch not the package
[01:12] <sidnei> jelmer, slightly different error
[01:12] <sidnei> jelmer, http://paste.ubuntu.com/814962/
[01:15] <jelmer> sidnei: try adding ,branch=master to the end of the URL
[01:16]  * jelmer wonders if we should do that by default
[01:16] <jelmer> we used to, but that's slightly inconsistent with git itself
[01:16] <sidnei> jelmer, seems to be progressing
[01:25] <sidnei> jelmer, failed at the end, but seems to have pushed fine http://paste.ubuntu.com/814968/
[08:04]  * jelmer waves
[08:05] <vila> hi all !
[08:05] <AuroraBorealis> hi
[08:05] <AuroraBorealis> can i ask where bazaar explorer gets the 'names' it uses for its branches? i think its bzr nick but setting it doesn't seem to change anything
[08:05] <mgz> morning
[08:08] <poolie> hi mgz
[08:12] <AuroraBorealis> brb
[08:14] <jelmer> ready to hang out when you are :)
[08:20] <wgz> hm, does someone need to setup the thingy?
[08:20] <vila> chromium is up, waiting for the hangout to start
[08:24] <wgz> vila: easy to miss notification top right
[08:24] <vila> got jelmer's one
[10:00] <wgz> sad robot.
[10:12] <jelmer> wgz: :(
[10:12] <jelmer> vila: looks like my pre-merge trick backfired
[10:13] <vila> :-/
[10:13] <vila> how so ?
[10:13] <jelmer> vila: lp:bzr/2.5 is slightly ahead of bzr.dev, so feed-pqm won't suggest my branch for submission
[10:15] <vila> yeah, you landed your commit builder fix there ;)
[10:34] <mgz> just had a thought, where did the fsync changes land, and when is it enabled?
[10:37] <mgz> 2.4b5 and for all systems it seems
[10:37] <jelmer> that sounds about right
[10:37] <mgz> I wonder if that's the cause of exarkun's weird slowness yesterday evening
[10:37] <mgz> windows arguably doesn't want it on by default, as ntfs is less flakey than nix filesystems
[10:39] <mgz> anyway, getting distracted, back to dealing with the weather
[12:13] <futch> hello, is it possible to change the history log of a bzr repo without using the rewrite plugin? (just out-of-the-box means) - without causing harm to the repos itself, of course
[13:02] <jml> hey
[13:02] <jml> colo
[13:03] <jml> that's what I want to talk about. I hear it's in a state for end users to try out?
[13:07] <mgz> futch: sure, just branch from the last revision you want, then replay or flatten changes by cherrypicking changes across
[13:08] <mgz> jml: jelmer will have more for you, but basically yes
[13:08] <jml> cool.
[13:08] <jml> where do I start?
[13:09] <mgz> `bzr init --development-colo` for testing, there's a branch to flatten that into the existing format which hasn't landed yet
[13:11] <mgz> then er... see <https://lists.ubuntu.com/archives/bazaar/2012q1/074217.html>
[13:12] <jml> mgz: are there recipes for migrating existing directory-style set ups into colo?
[13:14] <mgz> good question, I think no, but there's a bug for `bzr reconfigure` to do... something
[13:14] <mgz> bug 919227
[13:14] <ubot5> Launchpad bug 919227 in Bazaar "reconfigure --colocated" [Medium,Confirmed] https://launchpad.net/bugs/919227
[16:22] <henninge> Hi!
[16:22] <henninge> What could make bzr ignore information in locations.conf?
[16:23] <henninge> I created a locations.conf to specify a push location but bzr push keeps insisting that "no push location" is specified.
[16:25] <henninge> (This is on OS X but so far that has not been a problem ;)
[16:27] <jelmer> hi henninge
[16:27] <henninge> Hey jelmer ;)
[16:27] <jelmer> henninge: how are you, enjoying your new job?
[16:28] <henninge> Very much, thank you!#
[16:28] <jelmer> henninge: cool :)
[16:28] <jelmer> henninge: "bzr config" should tell you what it thinks the push location is, and where it got it from.
[16:28] <henninge> oh, right
[16:29] <henninge> I am not the one using OS X, btw ...
[16:31] <henninge> jelmer: bzr config returns the expected values, as configured in locations.conf.
[16:34] <henninge> jelmer: http://paste.ubuntu.com/815496/
[16:35] <jelmer> henninge: hmm, that's really odd
[16:35] <jelmer> what's the command being run - just "bzr push" ?
[16:37] <henninge> yes
=== yofel_ is now known as yofel
[16:38] <henninge> Inside a lightweight checkout of a branch, both located inside the repository.
[16:38] <henninge> Pushing directly from the branch is no different.
[16:39] <jelmer> henninge: and specifying a location works?
[16:40] <henninge> yes
[16:41] <henninge> jelmer: I have to say that I have not been entering the commands myself but have been given instructions through skype chat ...
[16:42] <henninge> But I doubled checked that he did it correctly.
[16:42] <jelmer> henninge: it's a mystery to me what's happening here
[16:43] <henninge> jelmer: ok, we'll work with specifiying the location for now.
[16:43] <jelmer> henninge: I'm about to EOD - can you perhaps file a question/bug about it?
[16:43] <henninge> jelmer: I will, after I have tried it myself, I mean, really myself. ;-)
[16:44] <henninge> jelmer: Enjoy your EOD
[16:44] <jelmer> henninge: if you feel like debugging this then the most useful thing to look at would probably be bzrlib/push.py - just stepping through the main function there
[16:44] <henninge> ok
[16:44] <jelmer> henninge: thanks!
=== beuno is now known as beuno-lunch
=== beuno-lunch is now known as beuno
=== Quintasan_ is now known as Quintasan
[19:45] <SamB> is it just me, or has bzr recently radically shortened the output it gives when an exception occurs?
[19:47] <SamB> whatever is going on, the following is a bit *too* short:
[19:47] <SamB> Pulling branches/v8.2 from svn://scm.gforge.inria.fr/svn/coq/branches/v8.2
[19:47] <SamB> open_branch() got an unexpected keyword argument 'possible_transports'
[19:51] <wgz> SamB: that's bug 902539
[19:51] <ubot5> Launchpad bug 902539 in Bazaar Windows Installers "TypeError: open_branch() got an unexpected keyword argument 'possible_transports'" [Undecided,New] https://launchpad.net/bugs/902539
[19:52] <wgz> you need to update to bzr-svn 1.1.2
[19:52] <SamB> The exception message doesn't tell me (a) *what* exception occured, (b) *where* it occured, (c) whether it was fatal, or (d) any indication as to what component(s) may need attention
[19:53] <SamB> wgz: I guessed it might be something like that
[19:53] <wgz> is this on the console or in bzr-explorer, what platform, what version of bzr, installed how?
[19:53] <SamB> but, didn't exceptions used to give a lot more information?
[19:53] <wgz> it certainly should be prompting you to report a bug.
[19:54] <wgz> find you .bzr.log and the block this error occurred in and paste it?
[19:54] <wgz> !pastebin
[19:54] <ubot5> For posting multi-line texts into the channel, please use http://paste.ubuntu.com | To post !screenshots use http://imagebin.org/?page=add | !pastebinit to paste directly from command line | Make sure you give us the URL for your paste - see also the channel topic.
[19:54] <SamB> I guess the exception might be getting swallowed or something ...
[19:54] <wgz> running `bzr version` will tell you where .bzr.log is
[19:55] <SamB> but I still think it's too terse
[19:57] <wgz> it's wrong, but without more details I don't know the cause :)
[19:57] <SamB> hmm
[19:57] <SamB> I think it might be from too long ago to be in .bzr.log :-(
[19:58]  * SamB was looking at mail from cron
[19:59] <vila> SamB: too long ago in .bzr.log ? That would mean a pretty intense usage or some high verbosity from bzr ;)
[19:59] <SamB> okay, no, that can't be it
[20:00] <SamB> my third-latest email has one of those in it ...
[20:02] <SamB> hmm
[20:03] <SamB> !pastebinit
[20:03] <ubot5> pastebinit is the command-line equivalent of !pastebin - Command output, or other text can be redirected to pastebinit, which then reports an URL containing the output - To use pastebinit, install the « pastebinit » package from a package manager - Simple usage: command | pastebinit -b http://paste.ubuntu.com
[20:04] <wgz> the whole of .bzr.log isn't needed
[20:04] <wgz> just copy out the part that you find that error in
[20:05] <SamB> I know, it just sounds more convenient than opening a pastebin site in my browser, given the amount of RAM available
[20:13] <SamB> http://paste.debian.net/153495/
[20:14] <SamB> I think multi-pull is eating the exception
[20:21]  * SamB looks for a command that compactly lists important details about bzr and plugins, like the proper exception output, only without the exception/traceback
[20:21] <wgz> yup, seems like it
[20:23] <SamB> I can see why it might not want to *just* fail, but this sure doesn't seem right, either!
[20:24] <wgz> try `bzr -Dno_apport --no-plugins assert-fail`
[20:25] <SamB> the trouble with that is that it doesn't list plugin versions :-)
[20:25] <wgz> but `bzr help commands` will tell you what multi-pull comes from so who to file the bug against
[20:25] <SamB> I got that from "bzr help multi-pull"
[20:25] <wgz> ^ah, remove the --no-plugins then
[20:26] <SamB> handy command, that
[20:26]  * SamB wonders why zsh doesn't know about it
[20:26] <SamB> hmm, didn't this used to show the paths to the plugins?
[20:27] <wgz> `bzr plugins --verbose` should.
[20:28] <SamB> I guess I could tell from the traceback if it really mattered, though
[20:28] <wgz> ^commands added just to cause a internal error aren't of general interest :)
[20:28] <SamB> you might be surprised
[20:29] <SamB> if someone is thinking "I know it shows this information when it fails", they might like a way to make it fail
[21:38] <wgz> SamB: so, did you file a bug somewhere?
[21:38] <SamB> wgz: am doing so
[21:38] <wgz> yell if you need help.
[22:39] <thomi> Hi - qbzr seems really broken in precise - I often get crashes with 'bencode' in the stack trace. I'm filing a bug about it, but I wonder if there's any additional information I can provide that might be useful.
[22:40] <wallyworld> poolie: bzr is quite borked on precise atm. i've filed one bug about bzr unshelve but bzr switch appears to have the same issue. are you guys able to look into it?
[22:41] <thomi> hi wallyworld - you too huh?
[22:42] <wallyworld> hi thomi. yeah.
[22:42] <thomi> I've got: https://bugs.launchpad.net/ubuntu/+source/bzr/+bug/921267
[22:42] <ubot5> Error: launchpad bug 921267 not found
[22:42]  * wallyworld looks
[22:43] <wallyworld> thomi: is that the right bug nr?
[22:43] <thomi> yeah
[22:43] <wallyworld> it must be private
[22:43] <thomi> it's currently marked as private since there's a stack trace.... I didn't realise that meant private only to me... but that would make sense
[22:43] <wallyworld> yep
[22:44] <thomi> wallyworld: try now
[22:44] <wallyworld> when you file using apport, that's what happens
[22:45] <wallyworld> thomi: mine is a different issue- a missing ListOption attribute on a config object
[22:46] <thomi> :(
[22:48] <Kamping_Kaiser> hi, i just ran 'bzr revert --no-backup -r -30' instead of using -r ..30 by mistake. am i out of luck, or can i get back rev 30 somehow?
[23:49] <jelmer> Kamping_Kaiser: "bzr revert -r30" ?