[00:23] <djbclark> FYI problem was with crap tunneling via openvpn in an inopportune path, not bzr :-)
[00:25] <Peng> scorp007: P'raps with "bzr reconfigure" (but back up first).
[00:29] <scorp007> ok
[01:09] <Jordan_U> How do I continue a merge after resolving all conflicts?
[01:31] <maxb> It's not clear what you mean by "continue"
[01:32] <Jordan_U> maxb: Nvm, what I wanted was bzr commit ( and what I really wanted was bzr rebase :)
[02:56] <lifeless> james_w: around?
[02:56] <james_w> hi
[02:56] <lifeless> bzr builddeb question
[02:56] <lifeless> I'm debianising something
[02:56] <lifeless> whats the right way to get the upstream tarball recorded
[02:56] <lifeless> specifically, as merge-upstream writes to debian/changelog
[02:57] <lifeless> what version should I have in debian changelog, if the upstream release is N
[02:57] <james_w> yeah, that's currently broken
[02:57] <lifeless> there are currently no commits on the debian branch
[02:57] <lifeless> ok, so given its broken, do you have a recommended workaround?
[02:57] <james_w> you can trick it by setting an upstream- tag and then putting that version as the top debian/changelog
[02:57] <lifeless> [this is why I think an import command that can do a merge would be clearer than a merge command which can do an import]
[03:00] <scorp007> does bzr compress the stream as you check it out?
[03:03] <scorp007> and does bzr show no progress information *at all* when you branch by default?
[03:10] <cody-somerville> scorp007, yes
[03:10] <scorp007> damn. If I'd known I'da used -v.
[03:11] <lifeless> huh, it shows progress by default
[03:11] <scorp007> if I kill it now I'm gonna have to start all over again.
[03:11] <scorp007> well, it don't show me anything.
[03:11] <scorp007> taking over an hour and still going...
[03:17] <scorp007> how do I merge in a patch from lp?
[03:22] <lifeless> james_w: this is annoying too :(
[03:22] <lifeless> :!bzr merge-upstream --version=0.1 ../testscenarios-0.1.tar.gz ../trunk/
[03:22] <lifeless> bzr: ERROR: There are uncommitted changes in the working tree. You must commit before using this command.
[03:23] <lifeless> james_w: pretty sure I filed bugs covering this.
[03:23] <lifeless> scorp007: bzr merge BRANCH; bzr commit - I suggest you read the docs though, as it sounds like you're a new user.
[03:24] <james_w> lifeless: why is that annoying?
[03:24] <lifeless> james_w: because I'm trying to debianise this
[03:24] <james_w> bzr merge does it
[03:24] <lifeless> james_w: bzr merge has --force
[03:31] <scorp007> lifeless, I'm not trying to merge a branch though
[03:31] <scorp007> just a patch: http://lists.mysql.com/commits/93844
[03:32] <scorp007> erm, I think I got it actually
[03:39] <lifeless> james_w: bug 498402
[03:40] <lifeless> james_w: If you can suggest a workaround that would be great.
[03:44] <james_w> debuild -S -i -I
[03:44] <scorp007> does launchpad support the bzr:// proto?
[03:44] <james_w> no, that won't work
[03:45] <scorp007> what's the fastest one I can use?
[03:46] <scorp007> arg. I keep getting bzr: ERROR: Cannot lock LockDir when I try to bzr pull from lp.
[03:46] <scorp007> http does not support mkdir()
[03:46] <lifeless> scorp007: you did 'bzr checkout'
[03:46] <lifeless> to update a checkout, use 'bzr update'
[03:47] <lifeless> if you want an independent branch
[03:47] <scorp007> oh
[03:47] <lifeless> you can run 'bzr unbind'
[03:47] <lifeless> james_w: do you mean 'use import-dsc' ?
[03:48] <scorp007> I'm already at the head. I'm trying to undo a local edit so I can merge a branch, but I can't even do a bzr revert.
[03:48] <scorp007> update just tells me I'm up to date.
[03:48] <scorp007> I tried deleting the file I edited, but update won't get it again
[03:49] <james_w> lifeless: yeah, but that won't put your upstream branch as the base
[03:49] <lifeless> james_w: I'm fiddling with that direction already, however its broken in treeless repos
[03:49] <lifeless> I've filed a bug
[03:50] <james_w> I saw
[03:50] <james_w> I merged your branch to fix it
[03:52] <lifeless> james_w: I think that was merge-upstream was broken
[03:52] <lifeless> this is new
[03:52] <lifeless> as I'm already running my branch, Ithink :)
[03:53]  * james_w -> leaving
[03:53] <lifeless> ciao
[03:53] <james_w> please file bugs for anything still broken
[03:53] <lifeless> oh, I will :)
[03:54] <scorp007> does bzr revert not work with lightweight checkouts?
[03:57] <lifeless> it works
[03:57] <lifeless> you might like to provide more detail with these questions
[03:57] <lifeless> 'I experience XXX, is that expected'
[03:58] <scorp007> I sort of alluded to it earlier. Same problem with locks.
[03:58] <scorp007> D:\dev\mysql-5.1>bzr revert
[03:58] <scorp007> bzr: ERROR: Cannot lock LockDir(http://bazaar.launchpad.net/~mysql/mysql-server/mysql-5.1/
[03:58] <scorp007> .bzr/branch/lock): Transport operation not possible: http does not support mkdir()
[04:00] <lifeless> that appears to be a bug
[04:00] <scorp007> awesome.
[04:00] <lifeless> note that using lightweight checkouts over the internet will be very very slow: don't do it.
[04:00] <scorp007> that makes me even happier.
[04:00] <scorp007> what do you suggest?
[04:00] <scorp007> heavy checkouts?
[04:00] <lifeless> bzr branch http://bazaar.launchpad.net/~mysql/mysql-server/mysql-5.1/
[04:00] <lifeless> or follow the mysql docs they wrote up
[04:01] <scorp007> I did.
[04:01] <lifeless> there are several howtos
[04:01] <scorp007> as we speak, I'm running >bzr branch lp:mysql-server
[04:01] <lifeless> I'm pretty sure none of them suggest lightweight checkouts over the net
[04:01] <scorp007> it's been going for 2 hours.
[04:01] <scorp007> and the bloody thing won't finish.
[04:01] <scorp007> with no feedback, either.
[04:01] <lifeless> had you run launchpad-login ?
[04:01] <lifeless> scorp007: is it using network or cpu ?
[04:01] <scorp007> no, but is that necessary, just to view it?
[04:02] <scorp007> yes, it's downloaded 900mb of stuff.
[04:02] <scorp007> and occasionally uses cpu.
[04:02] <lifeless> scorp007: launchpad-login enables bzr+ssh which is a lot faster than http doing VFS calls
[04:02] <lifeless> for most projects it doesn't matter
[04:02] <lifeless> but mysql-server is very large, and it makes a great deal of difference there
[04:03] <scorp007> great... I wish they told me.
[04:03] <scorp007> they just say "use bzr branch lp:foo"
[04:03] <scorp007> which seems to default to http
[04:03] <lifeless> :<
[04:03] <lifeless> there is a discussion at the moment about making the default better
[04:03] <lifeless> its desired
[04:04] <scorp007> I'm sure. But this has killed over 2 hours of my time and I don't know whether there is a benefit of killing it and starting over.
[04:04] <scorp007> I already have a lw-checkout copy though -- perhaps I can convert it to a branch?
[04:05] <scorp007> (it's times like these I wish they used git... or even CVS for goodness sake.)
[04:06] <scorp007> with those old things like cvs or svn, you can at least abort the process and continue on, just by running an update.
[04:06] <scorp007> if I cancel this branch operation, it's gonna start from scratch.
[04:10] <scorp007> do you know if it's possible to put my lw copy to use?
[04:10] <scorp007> i.e. convert it to a "real" branch or something?
[04:20] <lifeless> scorp007: you can convert it but it has to download the same amount of history
[04:20] <lifeless> scorp007: I would stop the branch, do launchpad-login, and branch again.
[04:21] <scorp007> so you think I need a launchpad account in order to get access to the code in a timely fashion?
[04:21] <lifeless> scorp007: as for using git, bzr and git branch at the same speed when using the smart server these days
[04:22] <scorp007> right, but what will a lp-account help me with here?
[04:23] <lifeless> scorp007: as I said above it lets you use bzr+ssh
[04:23] <scorp007> oh
[04:23] <scorp007> so the command would then be bzr branch bzr+ssh://lp:mysql-server ?
[04:23] <lifeless> no, still lp:
[04:23] <scorp007> oh
[04:23] <scorp007> how will I know what it's using?
[04:23] <lifeless> but if you've logged into launchpad it selects bzr+ssh under the hood
[04:24] <scorp007> oh
[04:24] <scorp007> and are you sure it's not using it now/
[04:24] <scorp007> ?
[04:24] <lifeless> it will ask for your ssh passphrase ? :P
[04:24] <lifeless> scorp007: have you done bzr launchpad-login before?
[04:24] <scorp007> no
[04:24] <lifeless> then yes, I'm sure
[04:24] <scorp007> so it never uses ssh anonymously?
[04:25] <scorp007> what about regular bzr:// without ssh?
[04:25] <lifeless> launchpad doesn't operate a naked bzr:// protocol server.
[04:25] <scorp007> oh.
[04:25] <scorp007> why's that?
[04:25] <lifeless> no reason
[04:28] <scorp007> hmm, I've reg'd an account, but how do I know what my username is?
[04:28] <scorp007> It only asked for a display name and email.
[04:29] <lifeless> its the bit after the ~ on your home page
[04:29] <scorp007> oh
[04:39] <scorp007> why don't I get any progress information with branch?
[04:39] <scorp007> even with -v@
[04:39] <lifeless> what do you get
[04:39] <scorp007> only:
[04:39] <scorp007> D:\dev\bzr>bzr branch -v lp:mysql-server
[04:39] <scorp007> Connected (version 2.0, client Twisted)
[04:39] <scorp007> Authentication (publickey) successful!
[04:39] <scorp007> Secsh channel 1 opened.
[04:40] <scorp007> that's it.
[04:40] <lifeless> might be a windows thing
[04:40] <lifeless> we normally show a progress bar
[04:40] <scorp007> hmm.
[04:44] <scorp007> do you know any small projects I can check out just to see if the bloody thing works?
[04:45] <lifeless> lp:testtools
[04:45] <scorp007> thanks
[04:46] <scorp007> oh so that works
[04:46] <scorp007> but still no progress information... horrible.
[04:46] <scorp007> especially for large projects.
[04:46] <scorp007> s/oh/ok/
[04:47] <lifeless> file a bug, assuming you're using some half-recent release
[04:47] <scorp007> Bazaar (bzr) 2.0.3
[04:47] <lifeless> yeah
[04:47] <lifeless> file a bug
[04:47] <scorp007> re: what?
[04:47] <lifeless> not getting a progress bar
[04:47] <scorp007> oh
[04:48] <lifeless> ciao
[04:48] <scorp007> thanks, cuya
[05:34] <jkakar> lifeless: commandant is packaged in my jkakar/commandant PPA.  The latest trunk version isn't package though, because it has some changes that rely on bzr 2.1 (I think).  I'll probably back out those changes so I can make another release sooner than later.
[05:56] <lifeless> jkakar: I mean in debian :)
[05:56] <lifeless> jkakar: PPAs are not at all as convenient or discoverable as distros proper
[06:56] <wgrant> Why does bazaar-vcs.org now redirect to bazaar.canonical.com?
[06:56] <wgrant> You guys needed to become *less* associated with Canonical, not more!
[06:56] <lifeless> wgrant: see the long thread about that on the list
[06:56]  * wgrant headdesks.
[06:56] <lifeless> wgrant: I share your concerns.
[06:57] <wgrant> WTF
[06:57]  * wgrant reads.
[06:57] <lifeless> however, close association with canonical works both ways - for some folk its a 'way cool' response, for others its a negative
[06:57]  * Kamping_Kaiser goes to find thread
[06:58] <wgrant> One of the major objections to bzr seems to be that it's owned by Canonical and only Canonical and only Canonical has anything to do with it.
[06:58] <lifeless> While I am concerned about the change, the association with Canonical itself is something I'm very pro, and I don't think its something we should ever feel ashamed of.
[06:59] <wgrant> Sure.
[06:59] <Kamping_Kaiser> its not. but using a .canonical domain reinfoces the 'just a product' meme
[06:59] <wgrant> But this new domain name shows that Canonical owns it, and you feel it is a Canonical roduct.
[06:59] <wgrant> Exactly.
[06:59] <wgrant> Not a community project.
[07:00] <lifeless> so
[07:00] <lifeless> grab my on jan 4th :P
[07:00] <lifeless> this is too close to being work :>
[07:00] <Kamping_Kaiser> lifeless: hehe.
[07:01] <lifeless> there are however many positives to being a product, not just a project :). As I've said I have concerns myself. Bring it up in the thread.
[07:02] <wgrant> I cannot yet find this thread.
[07:02] <Kamping_Kaiser> for the logs: https://lists.ubuntu.com/archives/bazaar/2009q4/065173.html
[07:02] <Kamping_Kaiser> wgrant: ^^
[07:02] <Peng> I've heard about these people that like the Canonical association, but I've never seen one.
[07:02] <wgrant> Peng: I presume MySQL does, but that's about all I can think of...
[07:02] <wgrant> Kamping_Kaiser: Ah, thanks.
[07:03] <Kamping_Kaiser> np. looks to be a dozen or so posts from the list history ( https://lists.ubuntu.com/archives/bazaar/2009q4/thread.html )
[07:05] <wgrant> Ah good, so everyone else disagrees too.
[07:06] <wgrant> So, bzr is even more fucked than before.
[07:08] <lifeless> don't whine here. Post to the thread.
[07:08] <lifeless> If you want to have an impact, that is.
[10:27] <ChrisMorgan> I'm wanting to install on a Gentoo machine, but http://wiki.bazaar.canonical.com/DistroDownloads#Gentoo just contains the code-formatted text "dev-util/bzr" - what does this mean?  I'm lost!
[10:29] <ChrisMorgan> Should I just download the tgz and work from that?  Any easy_install script?
[10:32] <luks> ChrisMorgan: dev-util/bzr is the Gentoo package name for Bazaar
[10:32] <ChrisMorgan> Oh... how does that get installed?
[10:32] <luks> I don't know, I don't use Gentoo, you do :)
[10:33] <ChrisMorgan> It's not my machine, it's an OSUOSL machine, just wanting to work out exactly how to do it so I can ask them
[10:33] <ChrisMorgan> It took me a while even to work out how to work out what OS it was...
[10:34] <luks> `emerge bzr` doesn't work?
[10:34] <ChrisMorgan> I'm not going to run it to see if it does
[10:35] <luks> well, I can tell you how to install bzr from sources
[10:35] <ChrisMorgan> I'm just wanting to know so I can ask them to do it
[10:35] <luks> but surely if y ou are going to maintain a Gentoo machine, you should know how to install Gentoo packages
[10:35] <ChrisMorgan> I'm not maintaining it.
[10:36] <luks> then anybody who maintains it will know how to install a standard package
[10:36] <luks> to just tell them to install the package mentioned on the wiki page
[10:36] <ChrisMorgan> OK, that's fine then :-)
[10:36] <ChrisMorgan> It's an OSUOSL machine hosting inkscape.org for Inkscape, and we want to switch from SVN to Bazaar for the website repository (Inkscape source has just moved to LP, so we naturally want the website to too)
[10:37] <luks> is this for accessing the repository via bzr+ssh?
[10:38] <ChrisMorgan> Possibly
[10:38] <luks> (which is confusing if you said you are using LP)
[10:38] <ChrisMorgan> Why is it confusing?
[10:38] <luks> why do you need bzr on the inkscape.org server?
[10:38] <ChrisMorgan> The website is updated via SVN currently
[10:38] <luks> oh!
[10:38] <luks> so you want to check out the website source code there
[10:38] <ChrisMorgan> (Sure, we have to trigger svn up manually, but it's in an SVN repo)
[10:38] <luks> yep
[10:38] <ChrisMorgan> Yep
[10:39] <ChrisMorgan> But I'm planning on having the new Inkscape website also modifiable from the web interface, and thus have it able to do committing and pushing as well as pulling and updating.
[10:40] <luks> yeah
[10:40] <ChrisMorgan> It's a Django thing we're (I'm) working on for it, Python, so that might make it easier, import rather than exec... I'll find out when I handle it
[10:40] <luks> just installing the bzr package will be enough
[10:40] <ChrisMorgan> So that's why it may need bzr+ssh
[10:40] <luks> their admins should know how to do that
[10:40] <ChrisMorgan> OK, thanks
[10:42] <ChrisMorgan> Is there a nice friendly API for that sort of thing?  something like import bzr, bzr.pull_update()?  :D
[10:44] <luks> there is an API, but it's more low-level
[10:44] <luks> you can call the command which handles "bzr pull"
[10:44] <ChrisMorgan> OK.  I'll take a look at it when I get to that stage.
[10:44] <luks> but then you have no advantage over actually calling "bzr pull"
[10:44] <ChrisMorgan> Except saving memory
[10:44] <ChrisMorgan> No need for a second Python interpreter to be loaded into memory
[10:45] <luks> true
[10:45] <ChrisMorgan> And if editing content through the website is done often, it's worth saving.  Faster as well, though for some things you'd want it asynchronous
[15:44] <nyu> hi
[15:44] <nyu> someone pushed his branch on top of server trunk and replaced the whole log with it
[15:44] <nyu> is there any way to fix that?
[15:45] <jelmer_> nyu: Hi
[15:45] <jelmer_> nyu: to prevent that from happening in the future you mean, or to get the old trunk back?
[15:45] <nyu> I don't think we have local branches without the new log.  "bzr pull" seems to have overwritten it
[15:45] <nyu> jelmer_: the latter
[15:45] <nyu> though the first would be useful too
[15:45] <jelmer_> nyu: if you run "bzr heads --dead" on the server it should list as one of the revisions your previous tip
[15:46] <jelmer_> nyu: you can then switch back to that revision by running "bzr pull -rrevid:<revision-id> ."
[15:48] <nyu> trying...
[15:52] <nyu> jelmer_: "bzr heads --dead" lists nothing
[15:53] <jelmer_> nyu: what about without --dead?
[15:53] <nyu> jelmer_: I think the revision is just not there.  it's not that he merged a branch we didn't want to merge, he replaced the whole thing with his local branch
[15:53] <jelmer_> nyu: even if he replaced the tip of the branch then the revision should still be there in the repository
[15:53] <nyu> so the log now lists merges from his local perspective ("merge from remote", etc)
[15:54] <nyu> ok, without --dead I get a commit id
[15:54] <nyu> *revision id
[15:56] <nyu> this rev id corresponds to a commit done after the mess
[15:57] <nyu> tried bzr heads --all, but this is all it lists
[15:59] <james_w> nyu: did they just push, or did they log in and rm -rf and then push?
[16:00] <jelmer_> nyu: Did they perhaps merge in the old tip before pushing ? In other words, does "bzr log -n0" list the old tip somewhere?
[16:00] <nyu> james_w: I think they just run "bzr push", which failed with "diverging branches" error, then run "bzr merge <remote url>" and "bzr push" again
[16:01] <nyu> seen this happen before.  but last time we had a recent local branch as backup
[16:01] <james_w> nyu: ah, in that case you can just find the old tip in the log of your local branch
[16:01] <nyu> james_w: my local branch is destroyed
[16:01] <james_w> then run "bzr push --overwrite -r <revno>"
[16:01] <nyu> I run "bzr pull"
[16:01] <james_w> then grab from the server again
[16:01] <james_w> it's ok
[16:01] <james_w> the revision is still there
[16:01] <nyu> that's the problem.  the server one was replaced
[16:02] <james_w> no it wasn't
[16:02] <james_w> "bzr log -n0" and you will find it in the history somewhere
[16:02] <nyu> I'm confused now
[16:02] <nyu> bzr log does list commits from before this happened, but they date back to october
[16:03] <james_w> what's the branch URL?
[16:03] <nyu> afterwards it's all "merge from server" commits
[16:03] <nyu> sftp://bzr.savannah.gnu.org/srv/bzr/gnewsense/metad/
[16:06] <james_w> "Resync with gnash packages from getgnash.org"
[16:07] <james_w> was that the old tip?
[16:07] <nyu> sounds like it could be, yes
[16:07] <nyu> it's fairly recent
[16:08] <james_w> bzr pull --overwrite -rrevid:rmh@aybabtu.com-20091217202044-ze9vfdfi9r157uin
[16:08] <james_w> on the server
[16:08] <james_w> then pull --overwrite locally
[16:08] <nyu> when you say on the server, do you mean on the server filesystem directly (e.g. sshfs mount) or with <server url> as argument?
[16:09] <james_w> then "bzr merge -rrevid:karl@kgoetz.id.au-20091218212429-gx6ixjplvm5yo88u sftp://bzr.savannah.gnu.org/srv/bzr/gnewsense/metad/"
[16:09] <james_w> on the server
[16:09] <nyu> ok
[16:09] <james_w> or locally, followed by push --overwrite instead of pull
[16:13] <james_w> then put "[DEFAULT]\nappend_revisions_only = True" in .bzr/branch/branch.conf on the server to stop this happening again
[16:14] <nyu> pffiew
[16:14] <nyu> thanks a lot man
[16:14] <nyu> I got pretty scared, I thought I'd have to live with a fscked log
[16:15] <nyu> can this append_revisions_only setting be applied to the whole shared-repo?
[16:16] <james_w> it's per-branch
[16:17] <nyu> uhm a quick google shows some pages suggesting it should be made the default
[16:17] <nyu> IMHO it'd be a good idea.  it can be a terrible mess when this happens and people like me have no clue on how to resolve it
[16:18] <nyu> oh, I see your name there
[18:24] <ejat> can i push to 2 destionation with 1 branch?
[18:24] <Tak> yes
[18:25] <ejat> Tak: any url can i refer ?
[18:26] <ejat> u mean .. i can push to subversion n launchpad together ?
[18:27] <Tak> you'll have to do two pushes, but you can push to both places from the same branch
[18:29] <ejat> owh ok ..
[18:30] <ejat> i got it .. u mean .. i need to push manually from the same brach to destination branch right ?
[18:30] <Tak> yeah
[18:31] <ejat> ok thanks Tak
[21:16] <rocky> if i've been working with multiple branches on a project and then stepped away for a few months what's the easiest way to see which branches i already merged back to the main branch?
[21:59] <lifeless> rocky: there is a plugin, removable
[22:00] <lifeless> or you can just running missing
[22:04] <lvh> Hello :-)
[22:05] <lvh> Any IDEs that integrate well with Launchpad? I'm guessing no, but no harm in trying I guess :-)
[22:05] <lvh> (IE: the bug tracker, the bzr bit ...)
[22:05] <lvh> Normally I'm an emacs person but people around me keep bugging me to look at netbeans/eclipse so I decided to humor them
[22:22] <ebrahim> I need some speed benchmarks for Bazaar 2.0
[22:23] <ronny> lvh: that stuff is on my tod
[22:23] <ronny> *todo
[22:43] <lvh> ronny, I thought you were a vi person
[23:17] <ronny> lvh: pida stufff a on other things