[00:44] I assume that the "left parent" is basically the mainline. [00:44] does it have a fixed merge_depth=0 ? [01:06] nevermind. figured it out. stupid question. [09:23] is it possible to create a "middleweight" checkout that has some, but not all, of the changesets? [09:25] Stacked branches might store a bit more data locally. [09:25] might or do? [09:26] I want to implement/support a "gradual" fetch of history where I "fill in" historical data based on things like network availability [09:31] Oh, interesting. How will you avoid locking problems? (I mean, the repo has to be write-locked while it's doing this, so it gets in the way of anyone trying to use it.) [09:31] aha https://blueprints.launchpad.net/bzr/+spec/shallow-checkouts [09:32] mindlace: You might want to start with stacked branches, since they support only having partial history locally. Then code something up that gradually fetches more. [09:32] yah, just not clear yet on how to cause "gradual fetch" of more [09:34] No idea! [09:34] Like Isaid, if you want it to be some sort of background process, locking will be an issue. [09:40] yeah… well, i'm not tremendously worried about that for the moment, as it wouldn't be working on 'active' repos ... [09:40] the idea is to auto-fetch repos 'nearby' - reported via zeroconf, for example [09:40] to support ad-hoc collaboration [09:41] mindlace: well, there is bzr share and bzr browse already via bzr-avahi [09:41] for precisely adhoc stuff [09:42] Eh. [09:43] well, isn't that fabulous, i love it when things already exist. Thanks, lifeless! [09:44] basically I'm trying to build a non-programmer project management tool / CRM with VCS inside … right now bazaar is the winner. (Started with git, moved to checking out mercurial, now I'm at bzr) [09:44] only bzr actually has a supported, documented api (bzrlib) and supports non-complete checkouts... [09:45] cool [09:45] I'm off, ciao. [09:45] Why do you want shallow checkouts? [09:45] lifeless: See ya. :) [09:45] "See ya"? I never say that. [09:46] Well, obviously one of you is lying. [09:46] hm… well, I want to implement "search" across all repos who are using this cms-thingy … and have results of search be auto-fetched [09:47] I had the idea that search results may return historical results but I guess I can simplify things by skipping that for the moment [09:48] (historical results - to fetch, then, you would need that historical revision -> present) [09:48] I guess I also want fragmentary checkouts ;P [09:55] hm… seems like sub-tree kind-of gets there... [10:20] oh hey, lifeless was robert collins [10:21] oh well guess i'll catch him when he's aboot [10:22] mindlace: I suspect you'll find he still is robert collins :) [10:22] just figured that out. [14:05] Hi ! [14:05] I'm getting a error like that: [14:05] bzr: ERROR: Revision {walter.php@gmail.com-20091019135859-x986fv0x73323qpy} not present in "KnitPackRepository('file:///home/walter/repositories/whissip-dev/.bzr/repository/')". [14:05] I tried to bzr check, but it don't works [14:05] can I just remove that entry from bzr index? [14:06] withouth it, evertything is ok [14:07] waltercruz: when do you get that error? [14:07] hum.. a lot of operations :( can't remember 100% [14:07] I upgraded my repository with upgrade -rich root.. sent it to launchapd [14:07] and when I tried to merge another branch.. I got this error [14:08] waltercruz: I mean, what operation prints that error exactly? [14:08] bzr log [14:08] but [14:08] bzr log -r-1 [14:08] ------------------------------------------------------------ [14:08] revno: 1 [merge] [14:08] committer: Walter Cruz [14:08] branch nick: whissip-dev [14:08] timestamp: Sun 2009-11-01 11:44:04 -0200 [14:08] message: [14:08] whissip merge [14:08] ------------------------------------------------------------ [14:08] Use --include-merges or -n0 to see merged revisions. [14:08] works [14:09] looks like I have a entry in the index... but no commit with that ID. [14:09] I suppose ;) [14:10] waltercruz: so "bzr log" immediately errors out saying that it doesn't have that particular revision? [14:10] yes [14:10] (bzr)walter@waltercruz:~/repositories/whissip-dev$ bzr log [14:10] bzr: ERROR: Revision {walter.php@gmail.com-20091019135859-x986fv0x73323qpy} not present in "KnitPackRepository('file:///home/walter/repositories/whissip-dev/.bzr/repository/')". [14:25] Wait, revno 1 is a merge? [14:26] Oh, nm, it presumably just be getting that from missing that rev as the parent... [14:26] Try log -r-1 --show-ids [14:26] 'll probably say that missing rev is its parent. [14:28] (bzr)walter@waltercruz:~/repositories/whissip-dev$ bzr log -r-1 --show-ids [14:28] ------------------------------------------------------------ [14:28] revno: 1 [merge] [14:28] revision-id: walter.php@gmail.com-20091101134404-9vahhw9duq5fjolg [14:28] parent: walter.php@gmail.com-20091019135859-x986fv0x73323qpy [14:28] parent: walter.php@gmail.com-20091031152550-jbf3xxslieifdpir [14:28] committer: Walter Cruz [14:28] branch nick: whissip-dev [14:29] timestamp: Sun 2009-11-01 11:44:04 -0200 [14:29] message: [14:29] whissip merge [14:29] ------------------------------------------------------------ [14:29] yes. [14:29] my repository screwd up somehow. [14:29] when I did a bzr check [14:29] and after that a bzr reconcile [14:30] I was with no revisions [14:30] bzr log showed nothing. [14:31] so I did a bzr merge 0..-1 theonethatIbranched.. after that, I stucked with that error in bzr log [14:33] The merge didn't cause the error, it just turned the existing screwup into a somewhat different screwup. [14:33] The problem existed the moment your tree base rev wasn't available from its branch's repo. [14:33] I don't know any way to make bzr do that, without trying to or poking around manually behind its back. Disk corruption of some sort, maybe. [14:34] hum... [14:34] You could try running 'check --repo', maybe, to bypass branch/tree checks and see if it turns up anything just in the repo. [14:34] strange. [14:34] will try! [15:05] KeyError: ('hacks.php-20090901021751-ck5pvqni080475dt-1', 'walter.php@gmail.com-20090916150844-ky37u7u3erm116a0') [15:05] bzr 2.1.0b1 on python 2.5.2 (Linux-2.6.30.5-linode20-i686-with-debian-5.0.3) [15:05] arguments: ['/home/walter/.virtualenvs/bzr/bin/bzr', 'check', '--repo'] [15:05] encoding: 'UTF-8', fsenc: 'UTF-8', lang: 'pt_BR.UTF-8' [15:05] plugins: [15:05] launchpad /home/walter/.virtualenvs/bzr/lib/python2.5/site-packages/bzr-2.1.0b1-py2.5-linux-i686.egg/bzrlib/plugins/launchpad [2.1.0b1] [15:05] netrc_credential_store /home/walter/.virtualenvs/bzr/lib/python2.5/site-packages/bzr-2.1.0b1-py2.5-linux-i686.egg/bzrlib/plugins/netrc_credential_store [2.1.0b1] [15:05] *** Bazaar has encountered an internal error. This probably indicates a [15:05] bug in Bazaar. You can help us fix it by filing a bug report at [15:05] https://bugs.launchpad.net/bzr/+filebug [15:05] including this traceback and a description of the problem. [15:06] (just part of traceback( [15:06] ) [19:16] It would be so nice to edit comments in history without uncommit [19:17] Specially when you find out that you have been calling combo box a drop-down list for so long time o.O [19:21] How to hack the bzr metadata to edit them? [20:29] morning [20:29] Good morning. :) [20:30] hi Peng [21:07] hi igc, up early! [21:21] hi lifeless. yup. Crazy day today [21:28] * igc breakfast [21:29] jelmer: bug 172360 [21:29] Launchpad bug 172360 in bzr "branching from rich-root to non-rich-root gives confusing errors" [Medium,Triaged] https://launchpad.net/bugs/172360 [21:29] jelmer: why did you change it from invalid? [22:18] lifeless: it was changed to incomplete without explanation, and the error message is still unclear [22:20] jelmer: it was changed to invalid after john said 'this is wont fix by design' [22:20] jelmer: I'm happy with it being open, but it needs an explanation as to what we're intending to do [22:27] lifeless: fair enough - I've commented [22:28] lifeless: fwiw it was changed to invalid before john said 'this is wont fix by design', [22:28] and john changed the importance to medium [22:31] oh right [23:01] Good morning. [23:38] bbl [23:48] hello spiv