UbuntuIRC / 2012 /04 /03 /#bzr.txt
niansa
Initial commit
4aa5fce
[00:00] <dakira> mgrandi: argh.. I think I see what happened
[00:00] <mgrandi> oh?
[00:01] <mgrandi> did addpatch add a bunch of files?
[00:02] <dakira> bzr lp-propose automaticalle selected lp:ubuntu/compiz as the branch to merge with. That is the branch where the development takes place. lp:compiz is where the merge should happen (it only contains the debian directory)
[00:03] <mgrandi> ah yeah
[00:03] <mgrandi> it did seem like there were a bunch of files that were present and got added
[00:04] <mgrandi> because the branches diverged
[00:04] <mgrandi> so what is lp:compiz then if dev happens at ubuntu/compiz
[00:04] <poolie> upstream development ought to be in lp:compiz
[00:04] <poolie> well
[00:04] <poolie> that ought to be a mirror, i suppose
[00:05] <mgrandi> ah makes sense
[00:05] <dakira> https://code.launchpad.net/~compiz/compiz/ubuntu
[00:05] <mgrandi> so yeah, just change the merge proposal dest and you should be good
[00:05] <dakira> this branch is listed in debian/control under the Vcs-Bzr tag
[00:06] <mgrandi> hmm
[00:07] <mgrandi> im not too familiar with how lp-propose works
[00:07] <mgrandi> but poolie can help hehe!
[00:07] <mgrandi> i have to go, gl
[00:13] <dakira> poolie: so what do I have to enter as the target branch if I want the above to be the target branch?
[00:15] <dakira> When I try to propose my branch to be merged back into lp:compiz I get "This branch is not mergeable into lp:compiz."
[00:29] <dakira> okay.. this looks better https://code.launchpad.net/~mniess/ubuntu/precise/compiz/fix-screenshot/+merge/100539
[07:03] <vila> hi all
[08:23] <mgz> morning
[08:24] <lifeless> ok look, its wgz renamed :P
[08:27] <fullermd> He's not renamed, he's just on the other side of the world from you, so he looks upside-down.
=== quicksil1er is now known as quicksilver
[08:37] <mgz> we are certainly each other's evil twin
[08:49] <fullermd> That's just the sort of thing an evil twin would WANT us to believe.
[09:37] <lifeless> spiv: https://code.launchpad.net/~spiv/ubuntu/lucid/paramiko/address-families-579530-lucid/+merge/25297
[09:37] <lifeless> spiv: perhaps that needs a 'delete' button clicked on it or something. I imagine its still in your activereviews list
[10:34] <jelmer> mgz, vila: are we hanging out now?
[10:34] <vila> jelmer: waiting for poolie ?
[10:35] <mgz> jelmer: that's the general plan
[10:37] <poolie> hi all
[10:37] <jelmer> hi..
[10:37] <vila> h...
[10:37] <vila> hi poolie
[10:40] <poolie_> hi
[10:40] <poolie_> am i here?
[10:40] <jelmer> poolie_: hi
[10:57] <jml> how can I eliminate this warning? 2012-04-03 10:57:12,217 - udd.scripts.mass_import - WARNING - Error accessing max threads file /srv/package-import.canonical.com/new/max_threads: [Errno 2] No such file or directory: u'/srv/package-import.canonical.com/new/max_threads'
[11:46] <poolie_> jml, hi, does the file exist?
[11:47] <jml> poolie_: no. when I create it empty the error goes away. useful error huh?
[12:11] <poolie_> :/
[12:11] <poolie_> it's a bit hardcoded
[12:11] <poolie_> whack james
=== yofel_ is now known as yofel
=== Quintasan_ is now known as Quintasan
[14:59] <ErwinM> Hi, I am looking for the automv plugin, but it seems gone
[15:00] <ErwinM> http://bzr.oxygene.sk/bzr-plugins/automv/ <-- empty
[15:00] <jelmer> ErwinM: it's now in bzr core I think
[15:00] <jelmer> ErwinM: "bzr mv --auto"
[15:00] <ErwinM> Aha, thanks
[15:01] <ErwinM> Yes, that worked just fine. :-)
[15:02] <jam> jml: echo 5 > /..../max_threads ?
[15:03] <jam> I believe it does a live check to allow you to turn up and down the number of running threads
[15:03] <jam> (udd uses it to lower throughput when LP is under high load)
[15:03] <jam> I think warning is probably useful for UDD, since it indicates a configuration error.
[15:04] <jam> Not as useful for your project, though.
[15:04] <jam> anyway, EOD here, have a good evening to all.
[15:04] <jml> jam: it runs when the file is empty
[15:04] <jml> jam: so it's a bogus warning.
[15:04] <jam> jml: well, I'm sure there is a default num threads.
[15:04] <jam> But not having the file would be a misconfiguration for UDD
[15:04] <jml> pha
[15:05] <jam> jml: I can see your point, it is just a "how generic is this code" issue. It isn't super-duper useful for UDD
[15:05] <jam> other than if we set a value and it doesn't take
[15:05] <jam> a warning helps understand if we accidentally have the file owned as "root" for example
=== lool- is now known as lool
=== deryck is now known as deryck[lunch]
[16:51] <Merwin_> Hey, I've got would like to make patch using diff of a revision range A..B but excluding the rev XXX, how can I do this ?
[16:51] <Merwin_> I would like*
[16:52] <mgz> I would... make a new branch -r B, merge -rXXX...(XXX-1), and diff -rA..
[16:53] <Merwin_> Hum good idea, I'll try, thank you
[16:59] <Merwin_> Erf, I tried to remove a colo-branch with bzr rmbranch and know my depo is unusable...
[17:00] <Merwin_> I had no pushed commits...
[17:00] <Merwin_> bzr: ERROR: No repository present: "file:///home/thibaut/OpenERP%206/OpenAssur/addons-opas/"
[17:00] <Merwin_> Seriously, I lost everything ?
[17:01] <mgz> hm, and I'm not sure the fix for bug 922953 has been landed on 2.5
[17:01] <ubot5> Launchpad bug 922953 in Bazaar "rmbranch should refuse removing active branch" [Medium,Fix released] https://launchpad.net/bugs/922953
[17:02] <Merwin_> I used bzr rmbranch mycolobranch
[17:02] <Merwin_> How can I fix that mgz ? I had work I didn't pushed, is it lost ?
[17:02] <mgz> welcome to the git way of working, I guess...
[17:02] <mgz> am checking to see if pre-fix there was a way of recovering or not
[17:03] <Merwin_> I don't understand why it removed my trunk branch wheras I told him to remove a colo branch Oo
[17:04] <mgz> see also bug 920653
[17:04] <ubot5> Launchpad bug 920653 in Bazaar ""bzr rmbranch" doesn't handle colocated branches" [Medium,Fix released] https://launchpad.net/bugs/920653
[17:05] <Merwin_> ...
[17:06] <Merwin_> Here, I don't even have a branch, brz branches prints nothing...
[17:06] <mgz> I did recover from doing various daft things with colo, but don't recall which issue was which
[17:06] <mgz> jelmer: any suggestions?
=== deryck[lunch] is now known as deryck
[17:08] <Merwin_> or vila maybe ?
[17:09] <Merwin_> Please tell me I'll be able to push my last 2 hours work I just commited before doing this :D
[17:11] <mgz> there's always the filesystem recovery tool route
[17:12] <Merwin_> At least it didn't removed the files...
[17:13] <mgz> just the repo and branch metadata, not the tree, right?
[17:13] <mgz> comma after repo.
[17:13] <Merwin_> Yep I guess, I've got a .bzr dir and my files are still there
[17:14] <Merwin_> So, I've got to call my boss est tell him why I can't push the things we were supposed to put in production tonight (I was making a patch for this)
[17:14] <mgz> but no repository/ or branch/ inside, right?
[17:14] <Merwin_> branch branches branch-format branch-lock checkout README
[17:14] <Merwin_> In .bzr
[17:16] <mgz> Merwin_: so, you have your actual changes, so this isn't a complete disaster
[17:16] <mgz> can either branch from previous location, copy tree across and commit for now
[17:17] <mgz> or.. it was a fresh project? recreate and copy tree, add and commit
[17:17] <Merwin_> I can't commit
[17:17] <mgz> in that branch.
[17:17] <mgz> you can make a new one.
[17:17] <Merwin_> Hum
[17:17] <mgz> which unblocks you for tonight.
[17:17] <Merwin_> I try
[17:18] <Merwin_> Don't seems to work: bzr branch actual_dir new_dir
[17:18] <Merwin_> Same error: bzr: ERROR: No repository present: "file:///home/thibaut/OpenERP%206/OpenAssur/addons-opas/"
[17:18] <mgz> Merwin_: so, did you create this from scratch 2 hours ago, or branch from an origin?
[17:19] <Merwin_> Nope, it's a long time branch I used for weeks
[17:19] <mgz> but you have no other version of the same base project anywhere?
[17:19] <Merwin_> Yes, on our dev server
[17:20] <Merwin_> Where I regulary push
[17:20] <mgz> okay, so branch from that.
[17:20] <Merwin_> But I won't have the last commit I made in my branch (the one I did't pushed yet)
[17:20] <mgz> then copy in your changes from today, and do one big commit.
[17:21] <Merwin_> Ok
[17:21] <mgz> you lose any little commits you did along the way, but the changes are still there.
[17:21] <mgz> and +affectsmetoo that bug I linked :)
[17:21] <Merwin_> That's enought for tonight
[17:22] <Merwin_> I did it thank you. It's a hudge repo, it's downloading I keep you informed :)
[17:23] <mgz> :)
[17:30] <Merwin_> mgz, while it's downloading, could you tell me how I'm supposed to remove a colo-branch ?
[17:30] <mgz> that would be the way, it's just buggy in 2.5.0
[17:31] <mgz> it should be roughly equivalent to removing the dir .bzr/branches/NAME though
[17:31] <Merwin_> ok...
[17:37] <mgz> not your fault, basically
[17:39] <Merwin_> mgz, merge -r 552.1.1...X => What should I put at X?
[17:39] <Merwin_> I want to undo the commit 552.1.1
[17:40] <mgz> 552
[17:40] <Merwin_> Looking at qlog, it seems that its 'ancestor' is 552.
[17:40] <Merwin_> Ok
[17:40] <mgz> it's easy in that case, .1 follows from the without-dot form always
[17:42] <Merwin_> Hum, my patch contains binary files that I want to have
[17:42] <Merwin_> It doesn't seems to handle binary files (OpenOffice .sxw, models for generating PDF)
[17:43] <Merwin_> I'll have to add them manually ?
[17:44] <mgz> you mean the textual patch only says "binary file added" but you want to be able to apply it and have the files added?
[17:44] <Merwin_> Yes, exactly :)
[17:45] <Merwin_> I don't have a lot I can push them via SFTP, it's not a problem, I'm just curious if there is a way to handle this
[17:46] <mgz> you can just use a different diff format I think.
=== CardinalFang is now known as qengho
[19:11] <mgrandi> hey jelmer if you are there, i remember you saying yesterday that you took out the fast export from x stuff from bzr-fastimport, does that include the bzr fast import (from svn dumpfile)?
[19:12] <jelmer> mgrandi: bzr fast improt never supported importing from a svn dump file
[19:12] <jelmer> mgrandi: bzr-svn supports importing from a dumpfile
[19:12] <mgrandi> hmm, does that get called by bzr fast-import?
[19:12] <jelmer> mgrandi: no, they're unrelated
[19:12] <mgrandi> just trying to understand the structure a bit
[19:13] <mgrandi> how did i import that svn dumpfile then
[19:13] <jelmer> mgrandi: bzr-fastimport only deals with the fastimport/fastexport format used by git-fastexport
[19:13] <jelmer> mgrandi: with bzr-svn I think? "bzr svn-import <dumpfile> <target>" ?
[19:14] <mgrandi> yeah
[19:14] <mgrandi> that seems like it, ok. I was going to say since import from dumpfile seemed important
[19:15] <jelmer> mgrandi: it doesn't read the dumpfile directly
[19:15] <jelmer> mgrandi: it creates a svn repo first from the dumpfile, and then uses that
[19:15] <mgrandi> yeah
[19:16] <mgrandi> i meant the general concept of importing from a dumpfile since svn doesn't seem to have fast-export
[19:16] <mgrandi> at least i think
[19:17] <jelmer> mgrandi: subvertpy has a 'subvertpy-fast-export' command that can generate a fast-export stream
[19:18] <mgrandi> ah ok.
[19:19] <mgrandi> also, before you dissapear again, i was wondering if there was anything to kinda 'verify' the integrity of a branch, or is it just bzr check
[19:20] <jelmer> mgrandi: just 'bzr check'
[19:21] <jelmer> mgrandi: generally, it's better to just talk to the svn repo directly rather than using a dumpfile
[19:21] <mgrandi> yeah that makes sense. let svn worry about the dump file.
[19:21] <mgrandi> but i was wondering cause im going to see if this bzr plugin for a IDE works, but it hasn't been updated in like a year so i have no idea if it still works
[19:32] <jelmer> mgrandi: the svn dumpfile is a distraction, it just adds two extra steps to the process
=== r0bby is now known as robbyoconnor
=== ajmitch_ is now known as ajmitch
[22:34] <poolie_> hi all
[22:37] <jelmer> 'morning poolie_
[22:37] <LeoNerd> Can I somehow "bzr ignore" files in a directory that don't have an extension (i.e. a "." in their name). These will be compiled binaries, but I don't want to ignore e.g. the .c source for them
[22:38] <mgrandi> should be able to with regular expressions
[22:39] <mgrandi> which i believe bzr uses for bzrignore
[22:39] <LeoNerd> Mhmm?
[22:40] <mgrandi> bzr ignore "RE:^#"
[22:40] <mgrandi> so to do a reg exp you put RE:
[22:40] <mgrandi> you can also look at 'bzr help ignore' for some more examples, you can also tell it to nOT ignore .c files
[22:40] <mgrandi> so, tell it to ignore everything in that directory, but not c files maybe?
[22:41] <LeoNerd> Oooh that might work too
[22:41] <bob2> or just ignore them all and add -f the source
[22:41] <LeoNerd> Nah, 'cause I'd still like to be reminded about new files in case I forget :)
[22:42] <wgz> LeoNerd: `bzr help ignore`
[22:43] <mgrandi> so add "somedirectory/*", and then "!somedirectory/*.c"
[22:43] <wgz> ...and I was scrolled up a little
[22:44] <mgrandi> time to go, laterz
[22:46] <wgz> can also do exactly "that don't have a dot in the name" with /[^.]*/
[22:46] <LeoNerd> That got it :)
[22:46] <LeoNerd> RE:examples/demo-[^.]+