keithy_hmm I wont to install config manager on my laptop but I dont have gcc/develper tools loaded00:29
timClicksam attempting bzr pull --overwrite and am receiving01:29
timClicksbzr: ERROR: Cannot lock LockDir(lp-64843152:///~flavour/sahana/sahanapy-trunk/.bzr/branchlock): Transport operation not possible: readonly transport01:29
timClicksis there anything I can do on my end? it looks like something to do with launchpad01:29
fullermdYou're doing that in a checkout of the branch on LP, to which you don't have write access, so writing to it fails.01:30
versatiletechI'm having issues with getting my linux user permissions right for someone on the outside to bzr push bzr+ssh:// in? Can someone point me to the right direction?01:33
fullermdWell, what's wrong with 'em?01:34
fullermdThey need write into the branch, and its respository.01:35
mkanatIn pre_change_branch_tip, what's the right way to get the email address of the committer? By parsing new_revid, or is there some better way?01:35
versatiletechfullermd: I guess the ssh user needs to be in the same group as the bzr folder's group correct?01:36
fullermdWell, that's one way to do it.  Another would be with world permissions.  Another possibility would be stepping outside the normal *nix permission system with some sort of ACL, if your system supports that.01:37
fullermdBut group is the normal solution, yes.01:37
versatiletechright now the .bzr's group is proftp01:38
versatiletechdoes the user's primary group need to be proftp?01:38
versatiletechor can there secondary group be proftp?01:38
fullermdNo, they just need to be in it.01:38
versatiletechlet me try that01:38
fullermdIf you're on a system with SysV filesystem semantics, you probably want g+s on some of the directories.01:39
keithy_is there a way to checkout such that you get just the files no scm behavuior01:50
keithy_i.e checkout --lightweight01:50
keithy_but more a deploy01:50
fullermdYou mean like export?01:50
fullermd(of course, with no scm behavior, you can't do things like "update" either)01:51
MTecknologybzr: ERROR: KnitPackRepository('lp-64843152:///~scripters/scripting/trunk/.bzr/repository')03:55
MTecknologyis not compatible with CHKInventoryRepository('lp-64843152:///~mtecknology/scripting/server-scripts/.bzr/repository')03:55
MTecknologyThat's odd.. any ideas?03:55
fullermdOne side is a pack-0.92 or the like, the other is 2a.03:59
fullermdSpecifically, yours is 2a, and you're trying to move revs from that into the ~scripters one which isn't possible.04:00
MTecknologyfullermd: any way to make that work?04:03
fullermdNo.  Rich-root revs aren't compatible with poor-root repos.04:03
fullermdSo your choices are (a) Upgrade the far side, or (b) Redo changes in a poor repo.04:04
MTecknologyfullermd: I'm trying to push to a branch that doesn't exist04:04
versatiletechassuming I do bzr remove MYCONFIGFILE.ext --keep, then bzr commit && bzr push-and-update. If someone else does a bzr pull, will their MYCONFIGGILE.ext be erased?04:05
fullermdMmm.  The error could be coming from trying to stack, I guess.04:05
fullermdversatiletech: Yes.04:05
versatiletechany other options?04:06
fullermdNot in-bzr, no.  A file's either extant in a rev or not.04:06
versatiletechI guess what I could do is tell everyone to backup their MYCONFIGFILE.ext, then I bzr remove MYCONFIGFILE.ext --keep from the central repo. Then bzr ignore MYCONFIGFILE.ext.04:09
versatiletechare bzr ignore's local?04:09
versatiletechor if somebody bzr pull's they'll get the new ignore that was committed to the central repo?04:10
fullermdDepends on whther the ignore file is versioned.04:10
fullermdWhich it presumably is, if you're using the 'bzr ignore' command.04:10
wgrant(by default it is)04:10
versatiletechyeah I noticed that I couldn't ignore a versioned files04:11
versatiletechWarning: the following files are version controlled and match your ignore pattern:04:11
versatiletechThese files will continue to be version controlled unless you 'bzr remove' them.04:11
fullermdOh, you can put it in just fine.04:11
fullermdbzr ignore just warns because the command doesn't DO anything.04:11
versatiletechso basically it doesn't truly ignore the file unless I do bzr remove04:12
versatiletechin other words if I make a change to database.php after doing bzr ignore to it, it will still be committed since it's versioned04:13
fullermdThe only thing that 'ignored' means is that 'add' won't add it when it walks across it, and 'status' won't show it as 'unknown'.04:13
fullermdIt doesn't affect anything else.04:13
versatiletechk I got it now04:14
MTecknologyfullermd: any idea about this? http://paste.ubuntu.com/362361/04:27
MTecknology2 servers give me that; 2 servers work correctly04:27
fullermdIt means you don't have your public key setup right.04:28
MTecknologynone of the root users on any server have ssh keys published to launchpad04:28
MTecknologyI only want them to pull down the content - should need authentication..04:29
fullermdIf they've been lp-login'd, they'll always need authentication.04:29
MTecknologyI didn't think I did.. if they did how can I undo that?04:30
MTecknologythere's no /root/.bzr*04:30
fullermdGoing through sudo and using your original user's ~?04:30
MTecknologyfullermd: thanks :D04:32
* fullermd is a cheenius 8-}04:32
MTecknologyya, you are04:33
ChrisMorganI've got bzr-gtk installed on my Ubuntu system but Nautilus integration isn't happening; the emblems are available but overlaying of files isn't happening.06:02
ChrisMorganAny idea why or what I should do to get it working?06:02
johnfhas 2.0.4 not hit pqm?08:21
fullermdPQM committed "Release 2.0.4 final" on Thursday...08:24
johnfmy bad. Wasn't looking at the 2.0 branch08:24
pooliehello all08:33
poolieigc, hi?08:35
=== sven is now known as Guest12460
james_wgood morning sprinters09:29
pooliehello james_w09:47
poolieare you home now?09:48
james_wno, I'm still in NZ09:54
hydratenaturallyi did something stupid...10:01
ChrisMorganDoes anyone know how I can get nautilus integration via bzr-gtk working?10:02
hydratenaturallyusing a hosted svn solution, i checked out the repo using bzr-svn, made some bzr --local committ's, went to make a big upstream committ, and when i did a bzr update it removed all my locally committed changes10:02
ChrisMorganhydratenaturally: I think that's one of the times for that famous saying, "Whoops."10:04
hydratenaturallywell, i'm just lacking the knowledge of bzr to go back and find my locally comitted changes10:04
hydratenaturallyi imagine a few commands here and there saves my life10:04
hydratenaturallyusing debian or ubuntu by chance chris ? https://bugs.launchpad.net/ubuntu/+source/bzr-gtk/+bug/28798810:05
ubottuUbuntu bug 287988 in bzr-gtk "Bzr Nautilus integration (Tortoise like)" [Undecided,Fix released]10:05
fullermdUpdate didn't remove your locally committed changes, it pivoted them to be a pending merge.10:05
hydratenaturallymm ok10:06
hydratenaturallysigh so where is this pending merge10:06
fullermdCheck status.10:06
ChrisMorganhydratenaturally: ubuntu, amd6410:07
hydratenaturallystatus says nothing10:07
fullermdThen you did a revert or some such since the update.10:07
ChrisMorganhydratenaturally: the thing is it's not working.  It's meant to be in there (they've even got screenshots of it) but the emblems are not being overlaid as they should10:07
ChrisMorganhydratenaturally: `bzr log`?10:08
fullermdLog by itself won't help after the pivot, the revs aren't in that branch's history anymore.10:08
ChrisMorganYou may want just the last N revisions though, add -l20 for the last 2010:08
hydratenaturallyugh i promise i didn't revert, here is a log of what happened http://pastie.org/79327610:08
ChrisMorganOoh, another chris!  ;-)10:09
hydratenaturallyyeah, is there a way of reviewing only your --local committ's ?10:10
fullermdWell, then either (a) there weren't any local commits in that branch, or (b) some weird bug exists hidden in bzr that nobody else has ever seen.10:10
* fullermd isn't better on (b).10:10
fullermdOr betting, either.10:10
fullermdYou can use 'heads' (in bzrtools) to see if there are any dead heads in the repo.  Lost local commits would show up there.10:11
hydratenaturallyhistory | grep ci | grep local shows me making local committs, i maybe dellusional...10:11
fullermdOr you may be making them in different checkouts.10:12
hydratenaturallychecked that10:12
hydratenaturallyheads doesn't say much, just something about my last upstream committ10:13
fullermdYou probably need heads --dead-only10:14
fullermdI don't think it lists dead heads by default.10:14
hydratenaturallystill nothing10:14
fullermdIs the repo shared with other branches/checkouts?10:15
hydratenaturallynah, very vanilla10:15
hydratenaturallyremotely hosted on beanstalk10:15
hydratenaturallydid initial import with svn10:15
fullermdNot the svn repo, your local bzr repo.10:15
hydratenaturallyno local branching or checkouts10:16
fullermdThen there aren't any commits in the repo that aren't in the history in svn.10:16
hydratenaturallybzr log shows none10:17
hydratenaturallyactually they're off by 110:19
=== poolie changed the topic of #bzr to: Bazaar version control | try https://answers.launchpad.net/bzr for more help | http://irclogs.ubuntu.com/ | Patch pilot: (volunteers wanted?) | bzr 2.1.0rc1 and 2.0.4 have gone gold
ChrisMorganWell, I got bzr/nautilus integration working.  sudo ln -s /usr/share/pyshared/bzrlib/plugins/gtk/nautilus-bzr.py /usr/lib/nautilus/extensions-2.0/python/ and then restart nautilus10:22
hydratenaturallyman this sucks....10:27
hydratenaturallyum where does bzr store committs with --local ?10:30
fullermdThe same place it stores any other revisions.10:30
hydratenaturallyhow would i go about pursing that, to say look for one of these deleted files10:30
fullermdAvailable evidence suggests there aren't any there, which means (since it takes heroic measures to remove things from a repo) there never were any.10:32
fullermdIf they don't show up in heads --all, they're not heads.  The only way they could not be heads would be to be in the ancestry of a branch in the repo.10:32
fullermdSince you say there's only the one [pseudo-]branch (that checkout), it means either the revs aren't there (and so never were), or they're in the ancestry of that checkout (which is the same as the svn repo's).10:33
fullermdOf course, it's _possible_ there's some bug somewhere along the line that's preventing them from showing up, and also prevented them from being left as pending merges in that update.10:34
fullermdBut it would be the first hint I've heard of such, after all the features have been around a long time, and I wouldn't know where to start looking for it.10:35
hydratenaturallyyeah i doubt it's a bug, probably something stupid i did10:35
fullermdWell, start looking at the assumptions.10:35
fullermdCheck out info -v, for instance.10:36
fullermdIf it's not using a shared repo, there's no way some other branch could be interfering with the heads detection.  If it is, see what other branches are in it.  That can eliminate one question, one way or another.10:36
fullermdSee how many revs it lists being in the branch history, and how many are in the repository.  Being from svn, it's probably a linear history, so maybe they should match up?  Not sure how strong an indicator that might be.10:37
hydratenaturallyk odd, bzr info -v shows repository: 36 revisions10:37
hydratenaturallybut bzr log only shows me 30 revisions10:37
fullermdWhat's the first line of info say?10:37
fullermd'k, take your right hand, and hold it about 2 feet away from your left cheek.10:38
hydratenaturallyyeah i'm stupid i get it10:38
fullermdThen look at how 'heads' lists two heads, with one of them flagged as the tip of the branch (checkout), and the other listed as dead.10:39
fullermdWhat version of bzr are you running, BTW?  rich-root-pack suggests pre-2.0; an upgrade may be in order (not that it has any bearing on this issue; just a general query)10:40
hydratenaturallythis bug seems interesting10:41
ubottuUbuntu bug 395514 in bzr "bzr update with local updates and commits does an unexpected merge (dup-of: 113809)" [High,Confirmed]10:41
ubottuUbuntu bug 113809 in bzr "update performs two merges" [High,Confirmed]10:41
hydratenaturallydebian 2.0.3-110:41
hydratenaturallybzr-svn 1.0.0-110:41
hydratenaturallydebian = bzr10:41
fullermdHm.  Maybe Debian makes local changes, or you had an older version when you made the checkout.10:41
fullermdSo that dead head is presumably the head of your local commits.10:42
hydratenaturallyok, makes sense10:42
hydratenaturallymy bzr knowledge is lacking on how to merge those changes into my "working" head10:42
fullermdSo you can use 'merge' to setup a merge of them.10:42
fullermd(of course, a lot of the merge info won't get into svn; unavoidable.  But the changes will, crumbed into one rev)10:43
fullermd`bzr merge -rchris@camus-20100125091843-zdk7x8gxpdeq8s4l .`10:43
fullermd(the trailing '.' is important)10:43
fullermdThen the usual check/commit.10:43
hydratenaturallyah ok didn't see revision-id there in heads -all10:44
hydratenaturallychris@camus:~/projects/fashionbone/bzr-beanstalk$ bzr merge -rchris@camus-20100125091843-zdk7x8gxpdeq8s4l .10:45
hydratenaturallybzr: ERROR: No namespace registered for string: u'chris@camus-20100125091843-zdk7x8gxpdeq8s4l'10:45
hydratenaturallyjust wtf10:45
pooliehydratenaturally: you need -r revid:christ@.... i think10:45
fullermdOK, now I think you're lying about your bzr version too   :p10:45
hydratenaturallychris@camus:~/projects/fashionbone/bzr-beanstalk$ bzr version10:46
hydratenaturallyBazaar (bzr) 2.0.310:46
fullermdOK, maybe not.  I thought the DWIM was in 2.0....10:46
fullermdBut it looks like 2.1.  So what poolie said.10:46
hydratenaturallycan't get the right syntax10:51
hydratenaturallyvice versa10:51
fullermdNo, -rrevid:chris@camus-20100125091843-zdk7x8gxpdeq8s4l10:51
fullermdrevid: tells bzr it's getting a revision id, then the string.10:52
hydratenaturallyah it worked, awesome10:53
hydratenaturallythanks for all the help guys, you save me oh so much pain10:53
fullermdDon't forget to commit the merge, before you forget it's there and mix new work in with it.10:53
hydratenaturallyheh, course, i just rsync'd the whole dir to my raid1 "just in case"10:54
hydratenaturallyi'm so overtired at this point that....10:54
=== poolie changed the topic of #bzr to: Bazaar version control | try https://answers.launchpad.net/bzr for more help | http://irclogs.ubuntu.com/ | Patch pilot: igc | bzr 2.1.0rc1 and 2.0.4 have gone gold
hydratenaturallyfullermd: thanks again11:02
hydratenaturallyi was about to take 2 xanax and have a drink11:02
hydratenaturallyi hear they call it the "jackson"11:02
* fullermd . o O ( Isn't that pretty much the same thing as using SVN anyway? ) O o .11:03
hydratenaturallyi should tack on 10/hour to my rate to use subversion11:04
hydratenaturallyand 25/hour for cvs11:04
fullermdHow much for Clearcase?11:05
hydratenaturallyoh jesus11:06
hydratenaturallyfree healthcare11:06
doctormoWhat's the best way to diagnose bzrlib being used in a thread... I can't decide if it's just taking a while to get a transport, or if it's stalled dead.11:28
doctormoI think it's the get transports that are freezing out11:35
doctormoMan bzrlib is way worse than launchpadlib. at least lp is ok so long as your lp object is in the same thread. bzr just dies if is so much as sees threading being imported.11:35
doctormoWorse, it dies by freezing without errors. locking up, charming.11:36
Raimis there a short URL spec to refer to the parent/push branch from the command line?11:42
Kinnisonwhat are you trying to achieve?11:46
fullermd:parent, :push11:48
Kinnisonare those urlspecs?12:09
lifelessdoctormo: this is patently false as we use threading ourselves12:10
fullermdNot urlspecs, no.  Builtin aliases.12:10
lifelessdoctormo: do you have a ssce ?12:10
lifelesssorry, sscce12:11
doctormolifeless: No, just rotten luck.12:35
lifelessif you have a deadlock, gdb can be great at debugging it12:37
lifelessyou may have found a python bug, for instance.12:37
bialixheya bzr!12:37
doctormolifeless: Possible, I'll have to look into it using gdb.12:39
doctormoTime for bed now though...12:41
doctormothanks for your help12:41
epscherncan i set up tortoise bazaar in a way so someone with bzr 1.3.1 can access it? (think i had no voice when i asked before)13:10
blenderkidhi. I did a mistake with bzr revert. I reverted all files instead of just one file. now the other files are lost. Is there a way to revert a revert?13:13
beunoblenderkid, bzr makes a backup of the files when reverting13:14
beunotry an ls -la in the dir13:15
blenderkida file with a ~1~ ?13:15
blenderkidahh good *puh*13:15
blenderkidDo I just need to rename it?13:16
blenderkidthank you so much :)13:16
MTecknologyblenderkid: I think that little feature has probably been able to save countless hair pullings13:16
epschernhm, can i somehow downgrade my repository format?13:24
fullermdDepending on the exact pair of formats.  'down' and 'up' are arbitrary directions.13:25
epschernwell, the server has only 1.3.1 and sysadmin refused to upgrade... but our windows machines have newest tortoise bazaar13:28
epscherni found now i can set the format on init... but don't want to lose all my commits so far13:28
fullermdYou can 'update' to rich-root-pack, which 1.3.1 can read.13:28
fullermdPresumably, your current stuff is in 2a, which you can't move to a non-rich-root format.  But r-r-p will take you back to 1.0.13:29
bialixvila, beuno: here?13:31
beunobialix, hey13:32
bialixq about upload13:32
bialixbeuno: any ideas what's wrong: http://pastebin.com/d5aaaadc013:33
* beuno looks13:33
beunobialix, no, there seems to be an error with the path somehow13:35
epschernfullermd: i tried it but it failed: http://paste.debian.net/57608/13:36
beunodoes the branch have anything funky?13:36
bialixI guess no13:36
beunobialix, maybe you could try trunk?13:37
bialixtrunk of the bzr-upload?13:37
bialixok, I'll try13:37
fullermdepschern: Mmph.  Stupid upgrade.13:38
fullermdepschern: You can work around it by 'init'ing a new branch in rich-root-pack format, then 'pull'ing from your 2a branch.13:38
bialixbeuno: upload trunk works for me13:52
bialixbeuno, vila: can you release it as 1.0.1?13:53
bialixcurrent version bundled into windows installer seems to be broken13:53
bialixand it seems you know about it13:54
beunobialix, sure, I'll chat with vila about releasing soon13:54
epschernfullermd: thanks a lot, that worked13:59
epschernnow i still hope cent-os or whatever the server has installed will upgrade their bzr version at some point so there won't be the deprecation message each time :)14:00
fullermdWell, you can push and pull back and forth.  So you could use 2a in your branch on the machine with non-ancient bzr version, and just use r-r-p on the old box.14:01
epscherni see14:02
epschernmight do that once i'm more used to it (just had to switch over from git)14:02
jelmerpoolie, lifeless, vila, jam: http://wiki.bazaar.canonical.com/BzrForeignBranches/Infrastructure14:14
vilabialix: bzr-upload is supposed to talk to a server where *bzr* is not installed, trying to make it talk to a bzr *smart* server is... interesting but they have opposite aims: bzr-upload want to provide a remote file system abstraction, the smart server try to get away from the remote file system abstraction14:28
bialixыщ Ш ырщгдв гыу ыаез штыеуфв,14:28
bialixso I should use sftp instead?14:28
bialixhi vila14:29
vilabialix: sftp sound better14:29
vilabialix: hi :D14:29
bialixok, I understand you about abstraction14:30
bialixmaybe you need to puth this remark to upload's help14:30
=== salgado is now known as salgado-lunch
lifelesspoolie: http://kitenet.net/~joey/blog/entry/a_clean_BTS_is_a_sign_of_a_sick_mind/15:18
lifelesspoolie: http://bethesignal.org/blog/2009/12/31/this-is-not-a-new-years-resolution/15:30
jelmerpoolie, https://code.edge.launchpad.net/~jelmer/tribunal/view-subunit/+merge/1800615:43
radoe_Where should I report broken links in the online version of the user guide?16:01
rubbsradoe_: make a bug report on lp, or mail the mailing list16:03
=== radoe_ is now known as radoe
radoerubbs: report it simply against bzr or is there something  better suited for docs?16:05
rubbsah, report against bzr, but put 'doc' in the tags list at the bottom. Someone should take it up soon.16:05
radoerubbs: thx.16:06
=== salgado-lunch is now known as salgado
micahgwith bzr-svn do I have to worry about branches not being from the same source when merging?16:25
=== beuno is now known as beuno-lunch
micahgwith bzr-svn do I have to worry about branches not being from the same source when merging?17:27
maxbmicahg: The question is a little vague, could you clarify?17:28
micahgI want to have a production and a devel branch in svn and import them both locally with bzr and push to svn with bzr svn17:28
micahgI was wondering if they need a common ancestor to merge17:28
micahgin bzr17:28
micahgmaxb: is that any clearer?17:31
maxbThey do need a common ancestor, but bzr-svn should deterministically import the common history i.e. it is _supposed_ to just work.17:31
micahgmaxb: if I branch the dev code in bzr and then bzr-svn push to the prod branch, will that take care of the common ancestry17:34
maxbI think that bzr-svn will use various bzr* svn properties to record this such that bzr-svn on another computer can do the right thing17:35
jelmeryes, it will17:36
pooliejam: https://bugs.launchpad.net/bugs/51241818:01
ubottuUbuntu bug 512418 in launchpad-code "no "diff coming soon" message on new mp?" [Undecided,New]18:01
=== beuno-lunch is now known as beuno
malibuHi there.. can anyone tell me how to see previous file versions in bazaar explorer?20:08
malibuI'm selecting the file in the tree pane but can't figure out how to see the versions20:11
rubbsmalibu: ok, this may not be the only way, but I found a way to do it.20:15
malibucool, thanks20:16
rubbsclick on the browse button and it opens up a "file browser" You can then put a revision in the upper right corner20:16
rubbsthat should change it to the revision and the files should then open as that revision...20:17
malibuok mine opens up with WT:, which I suppose stands for working tree..20:17
malibuok very cool.. thanks!20:18
rubbsyou can replace that with any revision spec. So like to get to the one just before you can put "-1" in20:18
malibuStill I wonder if there is a way to get a list of every revision for a file..20:19
rubbsoh you mean like annotate?20:19
malibuwhat is annotate?20:19
malibuSorry I'm a bit of a newB20:20
rubbsannotate will open the file and give you the revision that each line was changed last on20:22
rubbsor do you mean just the lastest revision that file was changed?20:22
malibuWell what I would like to see is a list of revisions that the file was changed on20:23
NfNitLoop'bzr log <filename>'.   Not sure how to do that in the GUI you're using.20:23
rubbsso if fileA was changed on revision 2, 5, and 7 then you want to see 2, 5, 720:24
NfNitLoopmaybe right-click on the file?  IS there a "log" option in the context menu?20:24
* NfNitLoop will brb, fixing IRC daemon.20:24
rubbsmalibu: ok, easy.20:24
malibuRight clicking on the file does not do anything, unfortunately20:24
rubbsmalibu: go browse then right click on file and hit "show log"20:25
malibuahh.. must be in browse20:25
malibuI was trying to right click on the tree node for the file in the main window20:25
malibuI get error  bzr+ssh://.... is not a local path20:26
rubbsyeah, I have done that before. The reason he doesn't do that is the work that is needed in the browse window can be slow, so he didn't us it in that main view, because it could make the whole thing slow20:26
rubbsthis way if you need some of that history viewing, you just hit browse, but don't have all the slow-downs by default20:26
rubbsthe manage button opens an actaul file browser to the working tree20:27
malibuOK well I'm not sure if it will work for me..20:30
malibuI get an error20:30
=== salgado is now known as salgado-afk
KilrooHmm. Does a bound stacked branch actually even mean anything?22:25
thumperKilroo: I don't think so22:32
thumperKilroo: although I guess you could have a light weight checkout of a stacked branch22:33
KilrooI'm guessing the command I left running at work won't accomplish much then...22:33
thumperKilroo: but I don't think you can commit to that due to a known bug22:33
KilrooI'm trying to find a workaround for something22:33
thumperwhich something?22:33
thumperand where are you using stacking?22:33
KilrooWell, what I have set up (albeit not thoroughly tested) for one of the subversion repositories I need to work with is a treeless repository with one branch bound to the subversion repo, feature branches as needed, and a lightweight checkout elsewhere that I can switch around among them. Nothing stacked there, at least not so far.22:35
KilrooUnfortunately I can't do that with the another subversion repo I need to work with because revision 30 makes bzr run out of memory.22:36
* bialix looking for igc22:37
KilrooThe command I left running when I departed from work today (which makes less and less sense to me the more I think about it) was a branch --bind --stacked from an svn checkout of the repository in question to a branch in a treeless repository.22:39
mwhudsonbound and stacked are basically orthogonal22:40
bialixKilroo: looks very funny22:40
KilrooI think what I was trying to get out of it almost makes sense...but not quite...22:41
bialixmwhudson: does our beloved core devs sprinting again?22:42
mwhudsonbialix: yes, in france this time22:42
bialixI suppose igc is not there22:43
mwhudsonbialix: right22:44
mwhudsonhe got the ok from his dr to travel, but decided not to go this time22:45
KilrooWhat the heck would that do if it did make sense, come to think of it? Make a local branch from the subversion repo that would require a connection to it to do anything, have to be rebased to it any time it changed, and automatically commit to subversion when I committed locally?22:45
KilrooThis is making my head hurt. I think I'm being ridiculous.22:45
bialixmwhudson: thanks for the info. I have some q for him about explorer and translations. will hang around some time tonight22:46
KilrooStupid revision where the whole blasted site including graphics and pdfs and logs and probably freaking archive files gets added all at once.22:46
mwhudsonbialix: yeah, i think he'll be on in a bit22:46
* bialix nods22:47
mwhudsonbialix: oh22:48
mwhudsonbialix: except it's australia day today22:48
mwhudsonso he's probably not going to be around :/22:48
bialixaustralia go forward!22:49
* bialix likes to think somewhere is summer right now, while outside his home is -25 celsius degrees22:50
Kamping_Kaiserwoo, public holidays :)23:12
NfNitLoopJust had a loooong discussion w/ coworkers about branching/merging/SVNfail.23:16
NfNitLoop(rather intelligent) coworker was asking how large companies handle large numbers of branches.23:23
NfNitLoopI said hopefully not via svn. :p23:23
KilrooYou know what's worse than svn?23:28
Kilroosvn where instead of having a repo for the project and, say, a production branch and a development branch...you have two repos. One for production and the other for development. With everything at the repository root so you can't branch from it either.23:29
KilrooAnd at least one revision in the repository that's so big bzr runs out of memory trying to look at it, so you can't even get local branches working.23:30
NfNitLoopoh god, yeah, you mentioned the prod/dev repos the other day.23:33
NfNitLoopThat's some special kind of fail.23:33
NfNitLoopWe actually do things mostly sanely around here.23:33
NfNitLoopwe've got a 4.0 branch (old)   a 5.0 branch (current stable) and a 6.0 (dev) branch.23:34
NfNitLoopWe do merges forward (->) most of the time.23:34
KilrooThe one thing to be said in its favor is that we have the revision history and who committed what, which we didn't have with plain ftp.23:34
NfNitLoopbut every now and then someone does something that makes svn merging go all wonky.23:34
NfNitLooplike, not merge from the root of the project.23:35
NfNitLoopOr manually revert a change.23:35
KilrooUnfortunately the people who set it up apparently didn't think beyond stacking that on top of the ftp approach.23:35
KilrooAnd I seem to be the only person who ever uses commit messages.23:35
NfNitLoopI'll say it again -- look for new work. :p23:35
KilrooI -want- to fix it.23:36
KilrooI think I actually have them convinced that it's worth it to let me take a break from developing new features to work on improving HOW we develop. Unfortunately I think I'm going to have to finish the two months-overdue database projects in my queue first.23:37
KilrooAnd because of the braindead way we have svn set up on the things that are using it now, switching to using it in a reasonable way is going to be as much of a hassle as switching to it in the first place was.23:39
KilrooAnd since I can't pull in the existing history using bzr because of the giant revisions, we may have to lose all the existing history instead of just half of it.23:39
NfNitLoopWhy would you have to lose history?23:40
NfNitLoopwould you be ditching SVN?23:41
KilrooI wish...no, there's two things to deal with.23:41
KilrooOne, with development and production having thus far been separate repositories, I'm not sure there is any sane way to convert them into branches that can be treated like branches without pretty much ditching the history from one of them.23:43
KilrooSpecifically, make production into the trunk, branch from it for development, paste development over top of it and commit all of those changes as "fixing this harebrained separate repositories business."23:44
KilrooThe other problem is that the linux administrator seems to think we can migrate everything without either his intervention or root access to either the old repositories or where the new ones are going.23:45
NfNitLoopshould be able to as long as you have write access.23:45
KilrooAs far as I've been able to determine, svn doesn't seem provide any means for performing the migration with those limitations, though.23:46
NfNitLoopthough you might need him to shut down the svnserv or whatever it's called, if that's how you're accessing it.23:46
NfNitLoopThis isn't #svn. :p23:46
KilrooAt least, nothing I tried seemed to be working. However, branching using bzr-svn and then pushing to the new repository worked a treat.23:46
mkanatKilroo: So it's fastimport that's running out of memory, here, that you're talking about?23:46
Kilroomkanat: No, it's bzr-svn.23:47
mkanatKilroo: Oh. Have you tried bzr-fastimport?23:47
Kilroomkanat: currently I can't make any use of a solution that does not leave me capable of committing back to the current svn repository.23:48
KilrooYeah, fun stuff.23:49

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!