[00:29] <BBHoss> anyone know why bzr would lock up pulling a large repo ~300 MB on a system with only 256MB RAM?
[00:56] <SamB> huh
[00:56] <SamB> some people!
[06:09] <suji1> Hi, I give the command bzr add filename , it shows an error the current directory is not a branch (bzr: ERROR: Not a branch)
[08:31] <cody-somerville> Anybody seen this before? http://pastebin.ubuntu.com/232826/
[08:32] <cody-somerville> ah, appears so
[08:42] <mneptok> cody-somerville: kneilsen in our team encountered the same behavior - https://bugs.edge.launchpad.net/bzr/+bug/375898
[08:45] <cody-somerville> guh
[08:45] <cody-somerville> And I've just been bitten by the bzr format incompatibilities :(
[10:55] <LarstiQ> cody-somerville: which formats?
[10:56] <cody-somerville> pack-0.98 and rich root pack I think
[10:59] <LarstiQ> right, non-rr vs rr
[11:00] <LarstiQ> cody-somerville: I'm happy to note that once we get 2a where we want it, there is a default format with rich roots, no more of that one-way-barrier after that (well, everyone will be on this side of the barrier)
[11:01] <LarstiQ> cody-somerville: right now it's a huge pain
[11:02] <james_w> we should probably improve that error message before 2.0
[11:02] <james_w> I imagine a lot of people will hit it and it's not the most helpful error
[11:03] <LarstiQ> james_w: you mean the trans_id one or the rr-format one?
[11:03] <james_w> rr-format
[11:03] <LarstiQ> right
[11:03] <james_w> the trans_id one should just be fixed :-)
[11:12] <jelmer___> moin!
[11:15] <LarstiQ> hoi jelmer___
[11:16] <LarstiQ> verterok, agrippa: does the new installer include enough to have qbzr working? ref bug 349143
[11:27] <jelmer___> LarstiQ: how's the 2a upgrade process going?
[11:32] <LarstiQ> jelmer___: I have bugs to file.  If we agree we can go ahead with the upgrade-guide approach before fixing them, a date should be picked for the upgrade.
[11:32]  * LarstiQ needs to find out some stuff about upgrading pqm first too
[11:32] <LarstiQ> jelmer___: it's moving, but no ETA yet
[11:38] <jelmer___> LarstiQ: ah, ok
[11:38] <LarstiQ> jelmer___: bug 390502 if you weren't aware of the bug number
[11:39]  * LarstiQ goes afk for a bit
[11:40] <LarstiQ> jelmer___: (I'm guessing something like next weekend for the upgrade, if we get agreement and announcement out this weekend)
[11:43] <lifeless> LarstiQ: I really don't like breaking *all* the metadata about existing branches
[11:43] <lifeless> LarstiQ: I'd /much/ much much much rather fix the bugs
[11:44] <lifeless> upgrading bzr itself won't teach us much if we haven't fixed the bugs such that we can do the process other people will to
[13:16] <mathrick> hiya
[13:16] <mathrick> does anyone know if bzr-svn supports history horizons yet?
[13:17] <mathrick> I know it does stacked branches, but that's a bit excessive in that it needs network for *all* access to old data, if I understand correctly
[13:28] <mathrick> jelmer___: poke
[13:28] <jelmer___> mathrick: hi
[13:29] <mathrick> jelmer___: does bzr-svn do history horizons yet? If I understand correctly, --stacked will want to access network for all past revisions, no?
[13:29] <mathrick> also, I have a repo that fail a stacked get with 0.6.2 here
[13:29] <jelmer___> mathrick: history horizons != stacked branches afaik
[13:29] <jelmer___> mathrick: what command are you trying to run exactly that fails?
[13:30]  * fullermd suspects getting history horizons in bzr-svn will be somewhat difficult before there are history horizons in bzr...
[13:30] <mathrick> jelmer___: http://pastebin.com/f1df8f6df
[13:31] <jelmer___> mathrick: stacked branches don't work in bzr-svn
[13:31] <jelmer___> mathrick: svn doesn't support stacking itself
[13:31] <jelmer___> mathrick: and bzr doesn't support stacking between different formats
[13:32] <mathrick> jelmer___: æh? But you tell to use stacked branches in the FAQ
[13:32] <jelmer___> oh, is that still there?
[13:33] <jelmer___> bzr-svn supported stacking for a while through a hack, but I removed that a while ago
[13:33] <mathrick> oh
[13:33] <jelmer___> mathrick: I don't see any comments about that in the FAQ, only " 30 In the future it should hopefully be possible to use stacked branches.
[13:33] <jelmer___> "
[13:33] <mathrick> it should just fail with "stacked branches not supported" then
[13:34] <mathrick> jelmer___: http://samba.org/~jelmer/bzr-svn/FAQ.html#cloning-a-large-subversion-branch-is-very-slow :)
[13:34] <jelmer___> mathrick: ah, I should probably remove that copy
[13:34] <mathrick> I believe it's also there in the bzr wiki
[13:35] <jelmer___> mathrick: I agree the error message could be clearer but afaik you get the same error message when trying to stack a bzr branch onto a different format bzr branch
[13:35] <mathrick> jelmer___: well, sure, but if you know svn doesn't support stacking, you could just catch it and tell the user outright, instead of after all the metadata is fetched
[13:36] <jelmer___> mathrick: nope, that can only be done in bzr itself, as bzr-svn doesn't know when a stacking clone is happening
[13:37] <jelmer___> mathrick: I've removed the FAQ from my homepage, can't find any references on the wiki though. Where did you see it?
[13:39] <jelmer___> mathrick: sorry for the confusion
[13:41] <mathrick> jelmer___: nah, I was wrong about the wiki
[13:41] <mathrick> and no problem, it's still a good solid plugin you have there :)
[13:45] <visik7> anyone using bazaar + buildout ? which parts of your project did you version ?
[13:46] <jelmer___> mathrick: thanks :-)
[14:09] <lifeless> visik7: the only folk I know doing that are the launchpad folk, someoe in #launchpad-dev may be able to tell you
[14:11] <fullermd> Hmm...
[14:11] <fullermd> Am I missing the magic to make branch: revspecs not blow up with a ReadOnlyError?
[15:09] <verterok> LarstiQ: no, the installer don't include PyQt nor Qt itself, but checks if PyQt is available and [dis|en]able the qbzr option
[15:27] <fullermd> igc: BTW, keywords plugin is making selftest fail, looking for now-renamed workingtree_implementations.
[15:43] <jelmer____> "bzr send --format=git" works for the most part, main thing missing is support for sending more than one revision since that implies sending multiple attachments :-)
[15:47]  * fullermd doesn't understand how the RevisionSpec_branch tests are passing   :|
[16:21] <ronny> yo
[16:22] <ronny> is there anz convention/notion to figure the "default" branch of a shared repository
[16:23] <ronny> ie something semilar to trunk in svn, master in git and default in hg
[16:23] <fullermd> 'trunk' is probably the more common, but I don't think there's any particular community pressure one way or another.
[16:24] <ronny> hmk, so i go for try trunk, try the repo itself, then fail
[16:43] <fullermd> Well, I think that's more than blown my bzr-time for this week...
[16:48] <jelmer____> fullermd, good stuff
[16:49] <jelmer____> fullermd: I think this is a very useful change (haven't looked at the code yet though)
[16:49] <fullermd> Yeah,  I got grumpy waiting for somebody else to do it for the past 3 years, and really didn't want to touch the Real Work I should have done instead.
[16:50] <fullermd> Potent combination.
[16:50] <Luke-Jr> jelmer____: please re-review my patch ;)
[16:51] <jelmer____> Luke-Jr, I don't see any new merge reviews..
[17:59] <LarstiQ> lifeless: good, I'd personally prefer fixing bugs first so that we can upgrade in place, I agree it is a far preferable workflow
[17:59] <jetole> hey guys, I just copied the wrong file over a repo file minutes after I checked it in to launch pad, how can I tell bzr to revert it back to what launchpad has on my system?
[18:00] <LarstiQ> jetole: `bzr revert`?
[18:00]  * jetole reads it over (although sounds right from what I remember from SVN)
[18:01] <LarstiQ> if that's not what you're looking for, I didn't understand your description of the situation correctly
[18:04] <jetole> no it was
[18:04] <jetole> it worked perfectly
[18:04] <jetole> thanks LarstiQ
[18:29] <Moof> hi
[18:29] <Moof> I'm trying to get to grips with something here.
[18:31] <Moof> I'm goign to be developing various projects on my machine. They'll all be stored in directories in remote servers. On my local machine, can I get away with creating one repository, and storing multiple branches form different servers in it?
[18:32] <NEBAP> it's possible to init a repository on a ftp server
[18:32] <NEBAP> is it also possible to recursively add folders and files to it?
[18:35] <NEBAP> directly on the ftp server
[18:41] <NEBAP> no ideas?
[18:41] <NEBAP> or is it not possible?
[19:00] <NEBAP> hmm
[19:00] <NEBAP> no ideas how I can add the files remotely?
[19:15] <NEBAP> no one online?
[19:37] <mzz> NEBAP: I think there's a plugin for that somewhere
[19:38] <NEBAP> I've used the init to initialize a standalone branch on the ftp server which worked without problems
[19:38] <NEBAP> then I've used checkout to create a local branch
[19:38] <NEBAP> and added the necessary files there
[19:39] <NEBAP> after changing some files I've commited the bind branch to the ftp server
[19:39] <NEBAP> which also worked without problems
[19:39] <NEBAP> but the files on the server are not changed
[19:39] <NEBAP> what is wrong?
[19:40] <NEBAP> the history is up to date
[19:40] <NEBAP> but the files arent
[19:40] <NEBAP> I've just used "bzr update" on the bound branch
[22:27]  * LarstiQ discovers bugmail is a task he can do on low energy
[22:27] <LarstiQ> reading/triaging
[22:39] <jelmer____> LarstiQ!