[00:25] lifeless: looms look pretty cool to me, thanks again. [00:25] lifeless: any direction you want to point me in to do a bit of hacking to start to understand the mechanics? [00:27] I think the only difference with what I was looking at was the order of the parents in the merge commits that result from up-thread. [00:27] and obviously that you have a working implementation :) [04:24] Hi! Is there a way for a repository to specify its submit_to address, such that `bzr send` could send mail to a default address for a pulled repository? [04:31] [04:47] The "HistoryOfBazaar" wiki page says: "In early 2008, Bazaar became GNU Bazaar when release 1.2 was accepted as part of the GNU project." is there any news or more details available ? [05:05] baijum: Not yet. I was expecting some kind of announcement, but it's just kind of leaked out gradually. [05:06] cjb: Not yet. [05:33] Wait, so that's true? [05:33] * bob2 watches bzr climb up to 1gb of ram to check out ikarus.dev [05:34] with a healthy 3MB of data in the branch so far [05:39] So, what's the point of being a GNU project? Canonical can provide all of the hardware bzr needs, and I doubt they'd want to give up any control over the project... [05:39] marketing [05:39] Advertising? [05:39] Heh. [05:39] homestarrunner.com cool website [05:39] heh [05:40] So, the big thing for shareware in the 1990s was to be get a Download.com badge, and now FOSS projects are going to go after GNU badges? [05:40] my local repo is messed up, is there a way, other then do do a bzr pull of upstream to just suck down all of upstream and overwrite and conflicts in my local repo? [05:41] Basic: Conflicts? Like, you've committed things in the branch, or your repo is corrupt? [05:41] like I forgot to commit stuff in local (desktop), hacked code in my laptop, pushed the changes from laptop to upstream, now want local/desktop to be in-sync [05:42] bzr merge gives me 138 conflicts and I know upstream and laptop are the changes I want to keep [05:42] now hit 1.5GB, nice [05:43] Basic: Why can't you bzr pull from upstream? [05:43] trying [05:44] Basic: "bzr help pull" says "If you want to forget your local changes and just update your branch to match the remote one, use pull --overwrite." [05:44] s/ / / [05:44] excellent, exactly what I want thanks [05:44] with giant red flashing lights "you may lose data" [05:45] Heh, right. [05:45] Basic: You can use "bzr missing" to see if you have anything important in your version of the branch. [05:46] Oogh, why do I feel sick? [06:18] jelmer: I'm being lazy and not doing research here, but the BzrPlugins page lists bzr-svn as beta. Is it still? [06:41] bob2: 'bzr heads' can recover commits, at leat until a gc. [06:41] 16:33 * bob2 watches bzr climb up to 1gb of ram to check out ikarus.dev [06:41] woops, didn't mean to paste :| [06:42] lifeless: By gc you mean the revisions being removed from the repo? Does that ever happen automatically? [06:42] Peng: not presently [06:42] Right. [06:43] Ok, I'm going to go take that nap I meant to take an hour ago. Bye. :) [06:53] Peng: it's at 0.4.x iirc [06:53] Peng: i need to try it... [06:54] Peng: recently saw a bug fixed where it made a remote svn repo unusable by *anyone* on commit fwiw [06:54] (not all commits) [06:55] * jeremyb needs to sleep as well [07:39] lifeless: ah, neato [09:59] jelmer: I'm in the Debian room ATM. [11:16] Odd_Bloke: ping [11:17] jelmer: Pong. [11:18] Odd_Bloke: still wandering around on FOSDEM? [11:18] jelmer: Yeah, in the Debian room ATM. [11:19] Odd_Bloke: ah, cool [11:19] me too [11:19] LarstiQ is the guy behind the big camera btw [11:20] Yeah, I ran in to him last night. [11:20] His hair was less blue at that juncture. [11:21] jelmer: I'm kinda opposite the door, wearing a red T-shirt using a Thinkpad. [11:21] A T60, if that helps narrow it down. :p [11:24] is there anyone about that understands the smart server code? I'm trying to fix a bug in it [11:25] johnf: Asking your question makes it more likely that people will respond. :) [11:25] Odd_Bloke: ok let me go paste some stuff somewhere [11:25] Odd_Bloke: ah, I think I know who you are then :-) [11:26] * jelmer is sitting close to the door with a Thinkpad X60 and wearing a black debian sweater [11:27] http://pastie.caboo.se/156526. Basically a change was made which meant bzr+https stopped working. I've fixed that but now it looks like _remote_is_at_least_1_2 needs to be set somewhere. [11:28] I think just setting it to true like is done in bzrlib/smart/client.py is correct but I'm not sure of the right place to do it [11:28] the transport/http/_urllib doesn't feel like it would be the right place to do it [11:31] johnf: you may want to bring this up on the list [11:31] I don't think there are any smart server ackers about [11:31] that was my next move :) [11:32] *hackers === Odd_Blok1 is now known as Odd_Bloke [11:42] johnf: one moment [11:44] johnf: I thought this came up on the list the other day, but I can't find it now. [11:44] johnf: are you using latest bzr.dev? [11:45] james_w: opps too late. Just emailed the list with some potential fixes. Stupidly didn't check archives. I'm testing with latest bzr.dev [11:45] johnf: no loss, I just thought there was a tentative patch you could apply, but as I can't find it a mail to the list is best. [11:47] james_w: The patch I just email works. First part I know is right. Second part does the job but I'm not sure that its appropriate. As I learned while debugging it today there is a lot of codebase involved in getting the smart server to do stuff! [11:50] johnf: I can't confirm I'm afraid, but you should get a response either way tomorrow. === Odd_Blok1 is now known as Odd_Bloke === mw__ is now known as mw === mw is now known as __mw === __mw is now known as mw [16:39] someone with sf.net shell acount with installed bzr here? I need bzr path [18:01] New bug: #195133 in bzr-loom "Please notify when a thread becomes empty" [Undecided,New] https://launchpad.net/bugs/195133 [18:05] New bug: #195134 in bzr-loom "Lots of up-threads are tiring" [Undecided,New] https://launchpad.net/bugs/195134 === mtaylor is now known as mtaylor|away [20:18] does "bzr tag blah" tag the most recent revision committed or get applied to the next commit? [20:19] Janzert: the most recent committed revision [20:19] thanks [20:30] Can anyone help me with bazaar.launchpad.net? I'm trying to checkout some code but it's taking AGES to connect and then it gives me the error: bzr: ERROR: Connection error: Couldn't resolve host 'bazaar.launchpad.net' (-2, 'Name or service not known') [20:34] Hey all... can someone enlighten me why, when I try to push my changes back to my mainline, I get bzr: ERROR: No push location known or specified. [20:34] awmcclain: it doesn't know where to push? [20:35] jdong: How does that get set? I can pull fine [20:35] awmcclain: the first push has to have an explicit destination. Afterwards bzr will remember that destination. Also see push --remember [20:35] jdong: I see. So, the first time you have to specify. [20:35] awmcclain: right. [20:35] awmcclain: hte last paragraph in bzr help push explains it better than I can :) [20:36] jdong: Got it. [20:36] awmcclain: also, bzr info can show you the various remembered locations, if you forgot or need something to copy-paste from [20:38] jdong: Is there an easy way to convert a branch into a checkout w/o overwriting the files I've changed? [20:38] awmcclain: bzr bind [20:38] Done. [20:39] greart [20:39] awmcclain: any differences you've made will show up as a merge when you first bzr up [20:42] hrm === mwhudson_ is now known as mwhudson [21:28] shagbag: sounds like a dns error at your location [21:28] shagbag: what does 'host bazaar.launchpad.net' say ? [21:42] hi poolie [21:57] good morning lifeless [21:58] morning lifeless [21:58] morning poolie [22:28] lifeless: I might be being silly, but can you confirm that setting tree.branch.nick doesn't change the thread you are on? [22:28] I'm looking at test_up_thread_preserves_changes in blackbox.py [22:30] james_w: it changes the pointer at the current thread [22:30] james_w: in the TODO there is a note that this should be replaced by it renaming the current thread. === schierbeck is now known as __schierbeck__ [22:47] lifeless: thanks. [22:48] james_w: most of TODO can be made bugs now [22:48] james_w: for hopefully obvious reasons, using malone was a little tricky before the release. [22:53] lifeless: of course. [22:53] lifeless: you abandoned the rename to warps? [22:56] morning all [22:58] james_w: mixed minds [22:58] I think the accuracy of warp and weft reference will be beyond most folk; [22:58] that is not actually helpful [22:59] but most everyone knows what a thread is in fabric terms [23:00] OTOH warp and weft are a more detailed analogy with richer (and useful) implications [23:00] poolie: today I am doing ui for shallow branches [23:01] poolie: call now? [23:07] lifeless: I might spend some time sorting out the TODO and moving some stuff to Malone then. [23:07] patch sent anyway, that's all for tonight. [23:07] that would be good [23:07] thanks! [23:08] the code looks very clean from what I have seen, and the documentation is great as well, so thank you. [23:09] The only thing I am not that clear on is "record", but it seems that is a temporary solution according to the TODO. === jamesh_ is now known as jamesh [23:16] I wonder if there's already a bug in the system about bzr not tracking file permissions [23:16] (git does) [23:16] lamont: git doesn't [23:16] lamont: which permissions would you like tracking? [23:17] git diff 8e16fa6ff4f1bd3e42febd166f3e1074c6f87bdb..7fca3ff2f2ab0774d4a92c2bd40bfc7c4d596e4c [23:17] diff --git a/cron.d/cache-git-status b/cron.d/cache-git-status [23:17] old mode 100755 [23:17] new mode 100644 [23:17] james_w: I beg to differ. [23:18] although just tracking the last 12 bits would be plenty. [23:18] lamont: git doesn't track full modes; it just cheats for what it appears to record [23:18] lamont: see git-fastexport's stream format for instance [23:18] while bzr just plain ignores mode [23:18] bzr tracks exec [23:18] lamont: that just means executable, which bzr tracks as well. [23:18] ah, just not rw? [23:18] anyhow, for sysadmins see the discussion I'm having with elmo & co about a plugin [23:19] lamont: make it 750 and try again. [23:20] james_w: hrm.. thanks [23:21] lamont: score one for git...not :> [23:21] lamont: no problem. You can see etckeeper for some wrappers and stuff around git(?, maybe others) that tracks full (?) permissions for versioning /etc [23:21] It's easy to make it 750; you just have to change your umask ;) [23:22] lamont: but in general it is a headache to track them, so most systems don't. [23:22] right [23:22] james_w: the canonical sysadmins would love a etckeeper for bzr; I have a spec [23:24] lifeless: you mean a plugin, rather than using etckeeper? [23:24] yes [23:24] I doubt etckeeper is VCS specific, but a plugin may make for a better experience. [23:25] lifeless: I have the implementation [23:25] lifeless: see http://gitweb.samba.org/ [23:25] right, that really is all from me. Thanks all, and good night. [23:25] however, it hsn't been merged yet because bzr doesn't have a start-commit hook [23:26] so it's not totally functional, which is why it hasn't been merged yet [23:26] jelmer: could you be more specific; thats just a web page list [23:26] jelmer: also, I'm talking about not using etckeeper but doing a plugin [23:26] there's only one etckeeper repo there :-) [23:26] http://gitweb.samba.org/?p=jelmer/etckeeper.git;a=summary [23:27] it basically boils down to being a plugin since it requires a plugin that calls out to etckeeper [23:27] is it just me or that web ui ugly? [23:27] of git web? [23:27] I like it better than loggerhead [23:28] loggerhead is ugly too [23:28] in differnet ways [23:29] none of the web uis I've seen for vcs introspection feel nice [23:29] yeah, there aren't any really good ones yet [23:30] if anyone can come up with a nice design for loggerhead, i'm permanently volunteered to do the coding to make it happen, btw [23:31] New bug: #195245 in bzr-loom "'record' is too generic a name, and is possibly a conflict with another plugin" [Medium,In progress] https://launchpad.net/bugs/195245 [23:34] mwhudson: I think its basically too static and too slow [23:34] mwhudson: the visuals aren't the issue, the coarseness of the interface is. [23:34] lifeless: 'too static' ? [23:34] mwhudson: I'd like fast drill-downs, pivots between files/history/annotations [23:35] google maps for the vcs data [23:35] oh right yes [23:35] now theres an idea; if we built maps tiles from a branch we could literally point maps at it [23:35] well, i see the point but don't really have a clue how this would work well [23:36] (even out of a browser, truth be told: 'bzr viz' is pretty nice, but not completely there) [23:36] Is there a plugin which will allow you to commit selectively by deleting lines from your commit log "line below" section? [23:37] I believe so [23:37] selective commit I think its called [23:38] Could also use gcommit I suppose, but not as vimtastic :-) [23:58] jelmer: so when should I expect a patc to review for a commit start python hook ? [23:59] lifeless: when I find the time :-( [23:59] lifeless: My time for bzr is somewhat limited at the moment and the time I do have I spend on the plugins I'm maintainer for [23:59] you're coming next week ?