=== vednis is now known as mars === jordan__ is now known as jordan === echo-are` is now known as echo-area === idnaria is now known as idnar === echo-are` is now known as echo-area [08:47] morning [08:49] hey [09:08] jelmer: around ? [09:17] smspillaz: hi [09:18] jelmer: hey :) [09:18] smspillaz: what's up? [09:19] jelmer: so I was looking at getting bzr join working for git imported branches, although it seems to fail when adding the new items to the inventory, since git imports set the file_id of the entry in the inventory to the name of the file (which can clash with items in the new inventory easily) [09:19] so I was wondering - is it safe to modify bzr-join to modify the file_ids of imported entries to be, eg, prefixed with the branch we're joining ? [09:20] (sorry if that sentence was a bit long, been poking around in pdb a bit) [09:22] here's an example in case you need one: http://paste.ubuntu.com/988614/ [09:23] smspillaz: it should be safe to do that, but it will mostly defeat the point of using bzr join at all [09:23] jelmer: hmmm, how so ? the history would be kept right ? [09:23] smspillaz: the revision history will be kept, but the files will appear to have been removed and readded [09:23] ah, that's a problem [09:24] jelmer: do you think there might be any other way around the name conflicts in the inventory then ? [09:24] smspillaz: I guess you could use something else for importing from git (fastimport?) [09:24] smspillaz: which will generate non-deterministic file ids [09:26] jelmer: right, although I'm already imported from git [09:28] smspillaz: yeah, you would have to reimport in that case [09:29] jelmer: so I guess the next question is, is it possible to export to git while keeping the history (not flattened ideally) ;-) [09:30] smspillaz: using fastexport/fastimport? I'm not sure how well that works, to be honest. [09:31] mmm [09:31] ok, I'll give a couple of things a try then [09:31] thanks :) [09:31] smspillaz: it might be useful to have a look at git-bzr-ng [09:31] IIRC it allows that sort of thing (but from the git UI) [09:31] mm [09:31] and it uses fastimport/fastexport internally === jam1 is now known as jam [12:54] jelmer: do you recall what in 2.5.0 conflicted with launchpad when you tried to update it in their codebase? [12:55] mgz: well, launchpad has to be updated to avoid the methods deprecated in bzr 2.5 [12:55] mgz: doing that broke a bunch of things [12:57] for parallel testing reliability, they could do with the fix for bug 396819 [12:57] Launchpad bug 396819 in loggerhead "AttributeError: _factory - lazy imports are not threadsafe" [Critical,Triaged] https://launchpad.net/bugs/396819 [12:58] which is in 2.5.0 but not current launchpad bzr-2.5.0dev2_r6152 [12:58] see bug 998040 [12:58] Launchpad bug 998040 in Launchpad itself "lp.codehosting.tests.test_bzrutils.TestGetBranchStackedOnURL.testGetBranchStackedOnUrlNotStacked(RemoteBranchFormat-v2) fails rarely/intermittently in parallel tests" [High,Triaged] https://launchpad.net/bugs/998040 [12:59] maybe some of our maintenance time would be well spent working through the things preventing that upgrade? [13:00] mgz: I guess it would be better to just cherrypick that fix? [13:02] perhaps? I don't think I have a good sense of the comparative amount of work involved. [13:03] using an actual release rather than a random dev revision seems like a good thing to do regardless, [13:03] but if it's really disruptive maybe not. === zyga_ is now known as zyga === yofel_ is now known as yofel [14:35] jelmer: I was told to paste you this http://paste.ubuntu.com/988992/ to see if you had any idea why its throwing an error.. [14:42] cjohnston: hi [14:42] cjohnston: at a quick glance it seems like python-dev isn't installed [14:47] * jelmer EODs [14:48] that seems to be it.. thanks jelmer === zyga_ is now known as zyga [15:57] hello i am working in a team so we get a centralized server, but i want to configure to push from team needs an approve, when the push don't auto commit changes [16:02] so i need to implement a workflow with gatekeeper [16:02] but is not clear for me [16:02] how to user make a request to merge [16:03] that's really a social question [16:04] asking and giving a reference to the branch is the basic requirement [16:05] if the gatekeeper is a person, yelling across the office would work [16:05] for a script, you generally want to send email, or something along those lines [16:05] click a button on a webpage, whatever [16:06] there are various existing tools like pqm and tarmac that do this and much more with launchpad [16:08] mgz: so pqm work with bze send ? [16:09] and yes gatekeeper is a person [16:17] if the gatekeeper is going to be a person, just tell everyone to put branches they want merged (with or post review) somewhere he can access them [16:26] with or post review ? [16:26] mgz can you help me explaining it or give a doc link ? [16:36] ovnicraft: http://doc.bazaar.canonical.com/beta/en/user-guide/using_gatekeepers.html [16:38] well i understand how it works but i don't get it how implement it [16:38] what's to implement with a human? [16:39] so gatekeeper will make merges from users branches ? or users can try to merge ? and gatekeeper is notified to reviewed === zyga_ is now known as zyga-afk [16:58] ovnicraft: the gatekeeper does the merges (on his side of the gate), submitters publish their branches wherever they want as long as the gatekeeper can read them (on the other side of the gate) [16:59] ok and users can use bzr send to notify the gatekeeper a patch === samkottl_ is now known as samkottler === zyga_ is now known as zyga === zyga__ is now known as zyga === zyga is now known as _zyga [23:19] hello i am working with pqm but i can't installed so it can't find pqm python library, i branch code from LP and ./autogen.sh;make; sudo make install [23:20] How do I see a branch as it was in a previous revision? I want to see a tree that looks like revision 51. In git, I'd do "git checkout 51". What's the bzr equivalent? "bzr switch -r51" crashes on me. [23:24] rbasak: bzr revert -r51 [23:24] rbasak: or perhaps 'bzr up -r51' (which updates the branch too) [23:25] jelmer: both of those commands crash too :-( [23:30] there is any way to implement my own pqm manager [23:30] i need to use merge proposal workflow (gatekeeper) and but lp-send works with LP [23:31] and code review app [23:39] rbasak: how do they crash? [23:39] ovnicraft: perhaps tarmac? [23:39] really pqm doc is poor [23:39] in fact there is no exists [23:40] ovnicraft: pqm is pretty hard to set up [23:41] yes miy friend [23:41] have you experience with tarmac to read your feed back ? [23:41] I have experience with both pqm and tarmac [23:42] tarmac is a lot easier to set up; I'd recommend it if you're using launchpad [23:45] well i am working in my own server [23:46] i want to implement my own server with merge proposal workflow [23:50] with bzr send, iget this errror [23:50] bzr: ERROR: Public branch "sftp://USER@192.168.1.119/home/gnuthinkserver/proyectos/nomina_61/addons/" lacks revision "ovnicraft@gmail.com-20120515180023-rgtcfe3a2fcibyq8". [23:50] i can't understand the problem? [23:50] ovnicraft: bzr send won't work with tarmac - tarmac picks up merge proposals from launchpad only, not from bzr send [23:51] sorry, I can't help right now - about to get some sleep [23:51] yes i know i just working with bzr send [23:51] ok thanks