[00:03] <GaryvdM> Hi igc
[00:04] <igc> hi jelmer, GaryvdM
[03:31]  * igc lunch
[08:53] <LarstiQ> ronny: create_branch_convenience returns the branch as a result
[08:53] <ronny> LarstiQ: then something is weird, i got none, let me check again what i did break
[08:55] <ronny> ok, pdb sucks
[08:55] <mwhudson> LarstiQ: i guess you're not here at uds?
[08:56] <LarstiQ> mwhudson: correctly guessed.
[08:56] <mwhudson> boo
[08:56] <LarstiQ> mwhudson: I will be at EuroPython fwiw.
[08:56] <LarstiQ> mwhudson: yeah :/
[08:56] <mwhudson> i won't, but oh well
[08:56] <LarstiQ> mwhudson: what would be the next intersection point?
[08:57] <ronny> LarstiQ: ok , pretty much found the issue - pdb wont print bzr branches when typing them in, wraping str/repr around kinda fixes it
[08:57]  * LarstiQ blinks at ronny 
[08:57] <LarstiQ> ronny: the print fine in ipython, I wonder what pdb is doing?
[08:58] <mwhudson> LarstiQ: coming to lca? :)
[08:58] <ronny> no idea, defered finding the issue
[08:58] <LarstiQ> mwhudson: eh, I'll take it into consideration, will have to look for travel sponsorship I guess :)
[08:58] <mwhudson> yeah, that's the problem
[08:58]  * igc dinner
[08:59] <LarstiQ> igc: eet smakelijk
[08:59] <mwhudson> i guess there will be other bzr sprints
[08:59] <ronny> LarstiQ: ok, pdb itself is pretty much broken, and nothing to easy fix
[09:01] <LarstiQ> mwhudson: yeah
[09:01] <LarstiQ> ronny: does ipdb fare any better?
[09:03] <ronny> LarstiQ: no idea, afair thats not in py.test
[09:05] <LarstiQ> ronny: something like --enter-pdb-on-error?
[09:05] <ronny> yes
[09:05] <LarstiQ> right, probably not then
[09:20] <hsn_> you guys are working on versioned properties aka editable commit messages?
[09:47] <jelmer> hsn_, I'm not aware of anybody actively working on it, but there are some vague plans for it
[09:53] <mwhudson> jelmer: could mrevell and i get you to write a little paragraph on why bzr-get kernel imports don't work yet?
[09:53] <mwhudson> (for the launchpad blog)
[09:56] <jelmer> mwhudson: sure
[09:56] <mwhudson> jelmer: thanks
[09:56] <mwhudson> jelmer: email it to me?
[09:59] <jelmer> igc, hi
[10:01] <jelmer> mwhudson, I hope evolution cooperates long enough for that to be possible :-)
[10:01] <mwhudson> jelmer: heh
[10:01] <mwhudson> jelmer: you can sms it to me if that doesn't work? :)
[11:09] <igc> hi jelmer
[11:09] <igc> hi mwhudson
[11:09] <mwhudson> hi ig
[11:09] <mwhudson> c
[11:10] <jelmer> igc: Thanks for the send review; I think I might've read more into your original review after seeing bzrlib.version_info
[11:11] <igc> jelmer: yeah - sorry for the multiple round trips
[11:12] <igc> jelmer: btw, is there an easy way for me to benchmark your bencode revision stuff?
[11:12] <igc> I could fastimport OOo again but I was wondering if there was a quicker script for just converting the revisions?
[11:15] <jelmer> igc, No worries
[11:21] <lifeless> igc: to microbenchmark the revisions, use the serializer interface directly
[11:22] <mwhudson> beuno:
[11:31] <awilkins> jelmer: Would a fastimport that produces a bzr-svn compatible branch be feasible
[11:31] <igc> hi lifeless, sure
[11:31] <awilkins> jelmer: The bzr-svn behaviour of holding a whole revision in memory requires some heavy iron for some repositories
[11:32] <poolie> hello
[11:32] <igc> awilkins: in theory, yes
[11:32] <igc> it's doesn't yet though
[11:33] <igc> awilkins: I'm pretty sure the fast-import code is designed though so that the rev-ids could be bzr-svn compatible (if we added some way of enabling that)
[11:33] <igc> hi poolie
[11:33]  * igc bbl - time to get the kids to bed :-)
[11:34] <jelmer> awilkins, It shouldn't actually keep the contents of the revision in history
[13:55] <igc> night all
[14:01] <poolie> night igc
[14:44] <ronny> hi
[14:44] <ronny> is there any simple way to get the workingtree of a branch in case them sharing the same path?
[14:44] <jelmer> they share a container
[14:44] <jelmer> You can use something like b.bzrdir.open_workingtree()
[14:45] <ronny> jelmer: oh, intresting, that should do
[14:45] <jelmer> ronny, it'll raise bzrlib.errors.NoWorkingTree if there's no working tree present
[14:47] <ronny> jelmer: the vast differences in terms of separation betwen repo/branches/workdirs and the different sets of syncronisation are still making headaches for me
[14:50] <ronny> i end up with the issue of lacking a coherent way to create light/heavy checkouts from local/remote locations for all of the vcs's
[15:14] <awilkins> ronny: If it's any help, I'm using the following ; a .repo folder with the repository in, a `list` file, and then I can synch to the svn stuff by doing  ; for B in `cat list` ; bzr pull -d $B/trunk svn+http://server.com/svn/$B/trunk ; done
[15:15] <awilkins> ronny: Then I branch each trunk so it has a sibling, local
[15:15] <awilkins> And take lightweight checkouts of those local branches
[15:15] <ronny> awilkins: what do you think the context is?
[15:15] <awilkins> Context?
[15:16] <awilkins> ronny: Your comment above at 1450
[15:17] <ronny> awilkins: but im not talking about bzr/bzr-svn
[15:17] <awilkins> ronny: Fair enough
[15:29] <mrevell> hey jelmer, were you able to look at that paragraph for the Launchpad blog post?
[15:30] <beuno> mwhudson, ping
[15:30] <jelmer> mrevell, Sorry, I forgot about it. Michael just reminded me about it and I'm working on it now
[15:30] <mrevell> ah thanks :)
[15:36] <bialix> jelmer: hi
[15:36] <jelmer> Hi bialix
[15:36] <bialix> may I ask about some strange bzr-svn behavior?
[15:36] <jelmer> bialix: sure, np
[15:37] <bialix> I've tried to catch you yesterday
[15:37] <bialix> I've branched svn repo from http://projects.serverzen.com/pm/p/cluemapper/browser/ClueBzrServer/trunk
[15:37] <bialix> I can see the log of the branch
[15:37] <bialix> but when I'm trying to see diff I've got: NoSuchRevision: KnitPackRepository('file:///C:/Temp/ClueBzrServer/trunk/.bzr/repository/') has no revision rocky@serverzen.com-20090226184017-m3cemyrz487uxhg3
[15:37] <bialix> this svn repo contains bzr branches inside
[15:38] <bialix> or used bzr-svn to interact with
[15:38] <bialix> I see many ghost merges
[15:38] <bialix> does I branch trunk in wrong way?
[15:39] <bialix> I'm using bzr 1.14 (for Windows) with bzr-svn 0.5.4
[15:39] <bialix> jelmer: ^
[15:40] <jelmer> bialix, That URL doesn't actually work for me, are you sure it contains a Subversion repository?
[15:41] <bialix> I've used this command: bzr get https://dev.serverzen.com/svn/cluemapper/ClueBzrServer/trunk
[15:41] <bialix> the first URL is Trac site URL
[15:43] <jelmer> bialix, I think this may be a bug that's fixed in 0.6.1
[15:43] <bialix> entire svn repo root at https://dev.serverzen.com/svn/cluemapper/
[15:43] <jelmer> bialix, At least, branching that URL works fine here
[15:44] <bialix> branching is OK
[15:44] <bialix> try to run bzr diff -c8
[15:44] <bialix> in your copy
[15:44] <bialix> of that branch
[15:44] <bialix> does it work for you?
[15:44] <jelmer> bialix, that breaks here as well and is a bug in Bazaar dealing with diffing ghosts
[15:45] <jelmer> there's an open bug report about it I think
[15:45] <LarstiQ> bialix: does branching the bzr branch work?
[15:45] <LarstiQ> bialix: and or running `bzr reconcile`?
[15:45] <bialix> but... it seems this ghost revisions were created with bzr and then pushed back to svn repo
[15:45] <bialix> LarstiQ: hi
[15:46] <LarstiQ> bialix: if they bork, I think I know which bug it is
[15:46] <LarstiQ> a rather elusive one :/
[15:46] <bialix> LarstiQ: what do you mean by: "does branching the bzr branch work?"
[15:46] <LarstiQ> bialix: 'bzr branch svn:/// foo; bzr branch foo test'
[15:46] <bialix> yes, this works
[15:46] <LarstiQ> bah :)
[15:47]  * LarstiQ hoped to have found a way to reproduce a bug of his
[15:47] <bialix> try yourself. there is only 15 revisions
[15:47] <jelmer> LarstiQ, it's a known bug, where bzr diff tries to get the file mtime of a ghost text
[15:47] <jelmer> LarstiQ, different from the one you and Ted Gould have been hitting
[15:47] <bialix> jelmer: but why there is ghost?
[15:47] <LarstiQ> jelmer: k
[15:49] <bialix> jelmer: run bzr log --show-ids
[15:49] <jelmer> bialix, There are ghosts because there have been pushes into svn with just the mainline, not the merged revisions.
[15:50] <bialix> it looks like these revisions have pushed using bzr-svn
[15:50] <jelmer> bialix, Sure
[15:50] <poolie> lifeless: are you home now?
[15:51] <bialix> perhaps I don;t understand how bzr-svn push works
[15:54] <jelmer> bialix: So, if you use "bzr push" from bzr-svn and have "push_merged_revisions = False" it only pushes the revisions on the mainline into the branch you're pushing into
[15:54] <bialix> jelmer: branching entire project with command: bzr get https://dev.serverzen.com/svn/cluemapper/ClueBzrServer
[15:54] <jelmer> bialix, so the right hand side revisions end up as ghosts if you later try to get them out again
[15:54] <bialix> this one does not have ghosts
[15:54] <jelmer> bialix: This is all expected behaviour
[15:54] <jelmer> bialix, except that Bazaar doesn't handle the ghost case very well in diff
[15:54] <jelmer> but that would also happen in other situations in which we end up with ghosts
[15:55] <jelmer> it's nothing to do with bzr-svn
[15:55] <bialix> jelmer: I think there is something strange
[15:55] <bialix> these revisions ARE in the svn repo
[15:56] <bialix> I've just get entire project with trunk-branches-tags and now I don't have this error
[15:56] <bialix> it looks for me that when I get only trunk bzr-svn does not try to get corresponding merged revisions from branches/
[15:57] <bialix> looks like bzr-svn round-trip breaks somewhere in the middle
[15:57] <jelmer> bialix, in that case there's two bugs here, one in bzr and one in bzr-svn
[15:58] <jelmer> bzr-svn missing revisions during fetch
[15:58] <bialix> maybe
[15:58] <bialix> you're the master of bzr-svn, not me
[15:59] <bialix> I hope you understand what's going on better than me
[16:00] <bialix> maybe it's another use case for colocated branches
[16:01] <jelmer> How are colocated branches related to this?
[16:03] <bialix> pull all related branches from svn repo at once, maybe
[16:04] <bialix> GaryvdM around?
[16:04] <GaryvdM> Hi bialix
[16:04] <bialix> hi Gary
[16:04] <bialix> I found that qlog does not show ghost merged revisions in revision properties. Is it right?
[16:05] <GaryvdM> Done the log -> diff changes. Just doing some testing before I push.
[16:05] <jelmer> bialix, pull is already supposed to pull in all merged revisions
[16:06] <GaryvdM> bialix: I have never found a way to create a branch with ghosts, so I've never been able to test how it behaves.
[16:07] <bialix> GaryvdM: try this: bzr get https://dev.serverzen.com/svn/cluemapper/ClueBzrServer/trunk
[16:07] <bialix> and then compare output of bzr log --show-ids and qlog
[16:07]  * GaryvdM installs bzr-svn
[16:07] <bialix> jelmer: so it's a bug in bzr-svn then
[16:08] <awilkins> bialix: Is cluemapper a SNOMED-CT product?
[16:08] <bialix> GaryvdM: I can push to lp
[16:08] <bialix> awilkins: I don't know, saw the link from Gary yesterday
[16:08] <GaryvdM> don't worry - allready branching.
[16:09] <bialix> http://projects.serverzen.com/pm/p/cluemapper
[16:09] <bialix> awilkins: ^
[16:09]  * awilkins looks
[16:09] <awilkins> Ok then, no it isn't
[16:09] <awilkins> Sorry, we have this thing called CLUE
[16:10] <awilkins> And mapping is a logical thing to do with the stuff it manages
[16:10] <jelmer> bialix, trying to verify that now
[16:31]  * bialix waves
[16:34] <jelmer> GaryvdM, bzr revert -rtag:bzr-svn-0.5.4
[16:40] <mwhudson> james_w: jml is looking for you
[16:40] <james_w> mwhudson: room 12
[16:40] <mwhudson> jml has now run away, damn him
[16:40] <james_w> damn him indeed
[17:14] <poolie> estella damm
[17:14] <GaryvdM> pygi: http://sourceforge.net/project/showfiles.php?group_id=121075&package_id=132255
[17:17] <LarstiQ> poolie: ah, found the barcelona beers? :)
[17:18] <pygi> poolie: the icons are working :p
[17:22] <hsn_> what command is used for setting parent branch?
[17:24] <poolie> pygi: yay
[17:32] <LarstiQ> hsn_: `bzr pull --remember parent/url`
[17:35] <hsn_> LarstiQ: thanks
[18:09] <trondn> Hi folks :)
[18:11] <trondn> is there any speed improvements planned for the "bzr log -v" command??? currently it takes 7 minutes on a branch from lp:drizzle on my 2.4ghz quad core box (with 4gb ram)...
[18:13] <poolie> trondn: it should be much faster in the development-rich-root format
[18:13] <trondn> poolie: can I convert my repository to this format?
[18:14] <trondn> the reason I ask is because I need it for OpenGrok (see http://src.opensolaris.org/source if you want to see it in action)... I need to get the complete list of files in each revision along with the commit message...
[18:15] <poolie> trondn: yes, but there are two caveats: this format is still in beta for a 2.0 release later, so there will be bugs
[18:15] <poolie> and the conversion takes some time
[18:15] <poolie> like hours, for a large branch
[18:15] <poolie> also, there are some log improvements coming up soon
[18:15] <poolie> even for other formats, so just using 1.15 may help
[18:15] <trondn> poolie: I'm running 1.15 already :)
[18:16] <trondn> great to know!
[18:16] <trondn> Right now this is the biggest problem I have with bazaar :-)
[18:17] <poolie> trondn: so this is just a readonly mirror?
[18:17] <poolie> then i'd definitely recommend upgrading
[18:17] <poolie> you can still pull from lp:maria
[18:17] <poolie> i mean lp:drizzle
[18:17] <poolie> and we'd like to hear your feedback
[18:17] <trondn> poolie: Yeah, the browsing will be from a readonly mirror :-)
[18:19] <trondn> poolie: I am currently using it (opengrok) to index mercurial repositories with over 76 gb disk footprint on src.opensolaris.org. I would like to have the same search options on the source I am working on from launchpad (btw. I love the complete integration of everything on lp!! nice work :) )
[18:20] <poolie> thanks!
[18:58] <schmichael> how do i get diff (preferably colorized somehow) output in vim when i'm writing a commit message?
[18:58] <fullermd> Well, the colorizing would be vim's turf.
[18:59] <fullermd> But try ci --show-diff
[18:59] <schmichael> fullermd: cool thanks, anyway to make --show-diff the default behavior of ci?
[19:00] <fullermd> I have it aliased.
[19:00] <schmichael> seems setting filetype=diff or filetype=<lang> in vim pretties it up a bit, wonder how to make that automatic
[19:01] <fullermd> au BufEnter bzr_log.* <whatever>
[19:01] <schmichael> fullermd: i owe you a beer ;-)
[19:02] <schmichael> thanks
[19:02]  * fullermd has his uses   8-}
[19:34] <rindolf> Hi all.
[19:34] <rindolf> What is the equivalent of svn export for http://bzr.savannah.gnu.org/r/gnash/trunk ?
[19:34] <rindolf> I want a snapshot of the sources. Don't care about the history.
[19:37] <schmichael> rindolf: bzr export dest http://bzr.savannah.gnu.org/r/gnash/trunk
[19:38] <rindolf> schmichael: thanks!
[20:05] <rindolf> schmichael: is bzr export like that faster than doing bzr branch?
[20:08] <LarstiQ> rindolf: I haven't looked at it, but I can imagine a worst case in which it's slower than bzr branch
[20:08] <LarstiQ> rindolf: so, I wouldn't place faith in statements about it's speed without backing it up with some evidence
[20:09] <rindolf> LarstiQ: ah.
[20:10] <LarstiQ> rindolf: exporting a revision tree is not a usecase the storage is optimized for
[20:10] <schmichael> rindolf: yeah, bzr repos are compressed so they might actually be signficantly smaller than the working copy if the project has little history and easily compressed files (i think ;) )
[20:10] <LarstiQ> rindolf: test it I say! :)
[20:11] <rindolf> LarstiQ: I'm trying to download a gnash snapshot.
[20:11] <rindolf> I gave up on bzr branch because it took so long.
[20:11] <rindolf> But bzr export downloaded 167408KB so far
[20:12] <schmichael> damn!
[20:25] <emmajane> beuno, ping!
[20:25] <emmajane> beuno, http://wiki.services.openoffice.org/wiki/Renaissance/Design_Proposals_for_%E2%80%9CAccessing_Functionality%E2%80%9D <--- usability testing for OOo
[22:09] <Peng_> mwhudson or beuno: ping
[22:09] <mwhudson> Peng_: hi, about to go to sleep; send email
[22:10] <Peng_> mwhudson: D'oh. One quick question? :D
[22:10] <Peng_> mwhudson: Anyway, good night. :)
[22:10] <mwhudson> Peng_: ok
[22:11] <Peng_> mwhudson: The "bzr serve" registry landed in bzr.dev. Do you want to keep the code to override cmd_serve for a few months?
[22:11] <Peng_> Or can we break compatibility now?
[22:12] <mwhudson> hmm
[22:12] <mwhudson> don't konw
[22:12] <mwhudson> ask beuno :)
[22:12]  * mwhudson zzzz
[22:13] <Peng_> mwhudson: Okay. :)
[22:14] <lifeless> Peng_: delete it; label the release as requiring a minimum bzr
[22:15] <Peng_> lifeless: Sure, but backwards compatibility is nice, and no extra effort.
[22:15] <lifeless> Peng_: you did ask :)
[22:17] <LarstiQ> imo, I'd raise the minimum bzr version after there is another release of bzr
[22:17] <Peng_> lifeless: :P
[22:17] <lifeless> LarstiQ: sure, I have no strong opinion
[22:18]  * LarstiQ nods
[22:19] <LarstiQ> lifeless: my condolences btw
[22:20] <lifeless> LarstiQ: thanks
[22:28] <Peng_> Oh, that's right. My condolences as well, lifeless.
[22:35] <lifeless> Peng_: thanks
[22:46] <cellofellow> how do I enable the bazaar nautilus integration? I have bzr-gtk installed.
[23:25] <robey> hey guys... my coworker is getting an OOPS for any "bzr branch lp:(anything)"
[23:26] <robey> and i do too... is launchpad down or something?
[23:26] <Peng_> robey: It's read-only for maintanence right now.
[23:26] <robey> yeah the website says that, but even "branch" fails
[23:27] <Peng_> I guess the XML-RPC server didn't take it well.
[23:27] <robey> i'm hoping there's something simple we're missing cuz this is his first impression of bzr after i've been promoting it for months, and he has the rage
[23:27] <robey> it looks completely broken :(
[23:27] <Peng_> Launchpad looks completely broken, not bzr.
[23:28] <robey> i doubt he cares about that distinction... oh well. :(  thanks
[23:28] <cellofellow> robey: too bad for your friend. Have any local branches he can branch?
[23:28] <dash> so give him a non-launchpad url
[23:28] <Peng_> Looks like bazaar.lp is down as well.
[23:29] <robey> it's something hosted on launchpad, and launchpad only seems to advertise the lp: names these days... i couldn't find any page that divulged the real url
[23:30] <cellofellow> robey: but maybe you could use `bzr serve` and he branch from your copy.
[23:30] <Peng_> Like I said, bazaar.lp is down anyway.
[23:30] <robey> i've never fetched that project
[23:30] <cellofellow> ah
[23:31] <robey> nevermind, i think the damage is done. thanks for helping tho.
[23:35] <Peng_> Oh, good, lp: URLs work again. Though bzr+ssh://bazaar.lp.net doesn't.
[23:36] <Peng_> ...Which makes sense if it's read-only, I guess...
[23:36] <cellofellow> too bad robey is gone
[23:39] <Peng_> robey's friend isn't very patient
[23:39] <cellofellow> :(