/srv/irclogs.ubuntu.com/2009/10/25/#bzr.txt

syncrondithanks lifeless00:06
lifelessnyu: I would use branches00:08
lifelessnyu: and when finishing the branch, merge it to trunk and commit00:09
nyuok00:14
syncrondiIf I've made an init on my server, do I also have to branch it before I can branch on my workstation?00:57
syncrondiAnyone know how I can remove a bzr project/repo?01:13
lifelesssyncrondi: not sure I get the question01:43
syncrondilifeless: I just deleted the .bzr/ and I hope that removes any projects created with init01:13
syncrondiBut I can't seem to connect to my new project01:14
syncrondiI get bzr: ERROR: Not a branch: "C:/Documents and Settings/syncrondi/My Documents/My Webs/project/bzr:ssh:/me@project.org/home/project/htdocs/".01:14
syncrondiI'm just running the same command as I was before to connect to a server on another project : >bzr branch bzr:ssh://me@project.org/home/project/htdocs01:16
spivsyncrondi: you want "bzr+ssh" not bzr:ssh01:16
syncrondiah..01:17
syncrondiI just noticed that myself01:17
syncrondixD01:17
syncrondiI hate being pesty, but it seems as though I'm just unable to get this right. When I try branching, I get bzr: ERROR: Connection closed: Unexpected end of message. Please check connectiv01:33
syncrondiity and permissions, and report a bug if problems persist.01:33
syncrondiI've got my key imported, host is accepted and now I get this :-\ Any ideas?01:34
lifelesssyncrondi: check your bzr.log (bzr --version shows where that is)01:45
lifelessI suspect you don't have bzr on the server, or some such thing01:45
syncrondilifeless: I'm pretty sure I apt-getted bzr.. What should I be looking for on the log?02:06
offby1.oO("apt-got"?)02:35
lifelesssyncrondi: if you're on linux, ~/.bzr.log02:37
MTecknology!logs02:37
ubottuOfficial channel logs can be found at http://irclogs.ubuntu.com/ - For LoCo channels, http://logs.ubuntu-eu.org/freenode/02:37
lifelessMTecknology: ?02:39
MTecknologylifeless: I just needed the link and this channel was the closest w/ ubottu in it - sorry for noise02:40
lifelessheh02:40
lifelessyou can msg ubottu i think02:40
* kfogel is away: zzz02:49
syncrondilifeless: well, my tail has this http://pastie.org/66861802:55
lifelesssyncrondi: you've clearly run other commands ;)02:57
lifelesssyncrondi: is that on the client or the server?02:57
syncrondilifeless: I thought you were asking for the server log. That's what I pasted and I was trying to init my project03:21
syncrondimy server is linux, client windows03:21
lifelesssyncrondi: I was asking for the client log03:33
lifelesssyncrondi: its going to be someting windows specific I suspect03:34
syncrondiok lemme get that03:34
syncrondihttp://pastie.org/66864303:49
lifelessso that means its not getting a connection via ssh to bzr on the server03:50
lifelessyou might try BZR_SSH=paramiko03:50
lifelessI believe our windows devs think that works better03:50
syncrondiI can connect to my other server (that wasn't set up by me) just fine from that client, lifeless03:55
lifelessok03:56
syncrondithat's why i thought it was something on the server side03:56
lifelesswell, the sudden disconnect error like that means we didn't get a handshake03:56
lifelessif bzr is installed on the server03:56
syncrondiit is03:56
lifelessthen its almost certainly client ssh issue03:56
lifelesse.g. a new host key03:56
lifelessplink doesn't prompt when its used in a subprocess03:57
syncrondithat could be the problem03:57
syncrondithe pw for my key and user are different, so perhaps I messed that up03:58
=== wgrant_ is now known as wgrant
MTecknologyHow can I uncommit a revision?07:09
MTecknologyI did bzr uncommit until I got back to where I wanted to be; then bzr push07:09
MTecknologybut it's still at r507:10
bob2you'll need to push --overwrite, I think07:10
bob2(back up remote side first)07:10
MTecknologythanks07:10
Peng_bob2: There's not really any need to back up. If you screw it up, you can always fix it.07:11
GaryvdMMTecknology: If other people allready have that revision, you may want to do bzr revert -r -2 bzr commit07:11
MTecknologyooh; it works nice too07:11
MTecknologythe other guys are sleeping atm07:12
GaryvdMMTecknology: Otherwise they will need to do pull --overwrite07:12
MTecknologythanks :)07:13
MTecknologyI like how uncommit kept my local files the same07:13
MTecknologythat worked out very pretty :D07:13
MTecknologyawesome stopping point for tonight07:17
MTecknologyafter using cvs, bzr makes me so so so happy07:21
MTecknologyafter dealing with svn servers, bzr makes me so so so happy07:22
lifelesswe're glad :)07:22
wgrantbzr-svn means I don't have to deal with svn when it's mandated at uni.07:23
wgrantSo no svn for me.07:23
MTecknologywgrant: I didn't know that existed.. happy day07:37
MTecknologywgrant: drupal.org might move from cvs to svn07:37
MTecknologyso some more saps will be making bzr work over svn jsut to make the rest of us happy07:38
Kamping_Kaiserany step up is a good step *g*07:38
MTecknologysvn is much better than cvs, but using bzr over svn... nice07:39
* Kamping_Kaiser wonders how easy it will be to backport bzr-git from sid07:40
MTecknologyI should try bzr-git.. I've been playing with the ubuntu kernel some07:42
fullermdMTecknology: By the way, you may recall we had a discussion a few months ago about mainline and revision numbering...07:42
MTecknologybut git isn't too bad07:42
MTecknologyya?07:42
MTecknologyI remember that07:42
Kamping_Kaiserit'll involve a new bzr, so its probably going to be annoying to backport. mutter. *puts on todo list*07:42
fullermdMTecknology: I stole the transcript of that out of my logs the other day and posted it up, to save having to type it again next time the question came up.  Hope you don't mind.07:42
MTecknologyfnope07:42
MTecknologynope*07:43
wgrantKamping_Kaiser: What are you running these days? Still gNewSense?07:43
MTecknologyhope it helps others too07:43
MTecknologyI should be sleeping......07:43
fullermdThat's what you said then too, as I recall   :p07:43
Kamping_Kaiserwgrant: yeah, except the version based on Debian stable, not Ubuntu07:43
MTecknologyI'll get ~5hr if I'm sleeping within the next 17min07:43
Kamping_Kaiserwgrant: gday and good evening, by the way.07:44
wgrantKamping_Kaiser: Indeed. Evening!07:44
Kamping_Kaiser:)07:44
MTecknologyfullermd: you have others ask the same thing?07:46
fullermdIt comes up from time to time.  It's one of those things that's Obvious(tm) once you grok it, but kinda hard to see beforehand.07:47
fullermdAnd it's not necessarily obvious that it's important to understand either, until you run into it headon.07:47
MTecknologywhere's teh link?07:48
fullermdhttp://bazaar-vcs.org/MatthewFuller/AboutMainline07:48
MTecknologyfullermd: you should s/USER/user/07:49
MTecknologyjust a thought07:49
MTecknologyThanks for the bookmark07:50
MTecknologyI think I'll read it again07:50
fullermdI swear, someday I'll actually get time to distill that sorta stuff down into an actual doc...07:52
fullermdI expect to get to it shortly after my pony arrives...07:52
Kamping_Kaiserhehe07:52
MTecknologylol07:52
MTecknologyif nothing else, the reader can have a laugh07:53
* Kamping_Kaiser goes to see how fullermd 's discussion of mainline fits with the documentation07:56
MTecknologyKamping_Kaiser: I learned A LOT that day07:57
MTecknologyI was too tired to retain it the first read though. It was almost the afternoon07:57
MTecknologythat's when I started work at 02:00; I quit. Now I desperately need a new job07:59
MTecknologyIf I don't get one, idk about next tuition07:59
fullermdI specialize in blasting information at people too tired to retain it; that way I get to seem brilliant without all the hard work of actually being so  ;)07:59
Kamping_Kaiserhehe07:59
MTecknologythis has been a while....08:09
MTecknologysending a process to the background - never do that intentionally08:09
=== AfC1 is now known as AfC
lifelessPeng_: testtools is in 2a now10:26
Peng_Eh eh?10:30
lifelessPeng_: you're subscribed to lp:testtools; there isn't a list, but it just got upgraded to 2a10:30
lifelessI presume you're subscribed for some reason ;)10:30
Peng_Ah. I was wondering why you had singled me out.10:31
Peng_lifeless: Cool. I'll upgrade. :)10:31
Peng_Oh, the branch has moved too10:33
Peng_Working tree formats that support content filtering have to disable some optimizations, right? Do they even do that when content filtering isn't being used?10:37
lifelessPeng_: hardlinking, and yes, and there is a bug open I believe10:45
Peng_Ah. I don't use hardlinking, so I guess that's okay.10:45
bialixjelmer: ping12:22
=== eix is now known as Goundy
jelmerbialix: pong15:58
vadi2Does bzr eclipse work with launchpad?16:25
Peng_What would be Launchpad-specific about bzr-eclipse's workingness?16:26
vadi2I think I read something about it not working with certain types of repositories. Anyway it's stuck uploading to a branch at 0% for a good while16:26
vadi2Was wondering if anyone knew something16:26
Peng_Oh. Well, I don't,16:27
vadi2cancelling didn't work either, so my whole eclipse is frozen on it16:29
vadi2@_@16:29
jelmerbialix: pong18:05
bialixhi jelmer18:05
bialixI've encounter bug in bzr-svn with russian characters18:05
jelmerbialix: that's been fixed :-)18:05
bialixtoday?18:05
jelmer5 minutes ago18:06
bialixyou're fast!18:06
bialixI'll try it18:06
bialixjelmer: in which branch?18:07
bialixall branches art https://code.launchpad.net/bzr-svn is at least 5 days old18:08
bialixjelmer: btw, how's your subvertpy-based fast-export-from-svn?18:09
bialixjelmer: are you still here?18:14
jelmerbialix: yeah, sorry18:14
jelmerbialix: it's in the main branch (launchpad only has mirrors)18:14
jelmerbialix: http://people.samba.org/bzr/jelmer/bzr-svn/1.018:14
bialixaha, thanks18:15
jelmerbialix: the subvertpy-based fast-export-from-svn should work18:15
jelmerbialix: but it's only in the subverpty development branch18:15
bialixI'm interested in this, I want to glue together 2 fast-import streams18:16
bialixand then produce integrated branch18:16
jelmerbialix: you can rebase as well18:17
jelmeror import to bzr and then fastexport18:17
bialixno, I think I can't18:17
bialixI have CVS repo and forked svn repo18:17
bialixthey are unrelated, because svn repo is not converted from CVS18:18
bialixso I need to figure out how to graft svn on CVS history18:18
bialixI have success with cvs2bzr though18:18
bialixnow I think I need to get fast-import stream from svn repo and then join them somehow18:19
bialixI'll try svn-import, bzr fast-export18:19
bialixyour bzr-svn branch is big18:21
bialix3200 revisions... still branching18:21
jelmeryes18:21
bialixincredible18:22
bialixjelmer: does your 1.0 branch is compatible with bzr 2.0.1?18:28
bialixor I need cherrypick?18:28
jelmerbialix: it's compatible with 2.0.118:28
bialixgood18:29
* bialix trying18:29
bialixjelmer: it working, many thanks18:33
bialixjelmer: after svn-import I have 2 branches with common history and 2 more branches w/o common history. is it normal?18:35
jelmerbialix: it depends on what happened in svn18:36
jelmerif there was common history in svn there will be common history in bzr18:36
bialixok, I'll aske the author of that repo18:37
bialixjelmer: are you interested in the selftest result of bzr-svn on windows?18:56
bialixthere is many unicode-related failed tests18:56
=== asac_ is now known as asac
jelmerbialix: please file a bug19:02
bialixand attach test.log there?19:02
jelmeryeah19:03
bialixok19:05
bialixjelmer: done: https://bugs.launchpad.net/bzr-svn/+bug/46061319:23
ubottuLaunchpad bug 460613 in bzr-svn "bzr-svn selftest failures on windows" [Undecided,New]19:23
igcmorning21:07
jelmer'moin igc21:12
igchi jelmer21:12
jelmerigc: this is probably a configuration issue on my end, but I wanted to try out bzr-explorer the other day and it seems to segfault when I try to start it21:19
jelmer"bzr qlog" segfaults in a similar way21:19
igcjelmer: which version of qbzr as you running?21:20
igcjelmer: it may me a PyQt thing in karmic which I beliebe bialix or garyvdm recently fixed21:20
jelmerr100621:20
igcs/me/be/21:20
igcbelieve21:21
jelmershould I update my version of qbzr or my version of pyqt?21:21
jelmerNever mind, updating my version of qbzr fixed it - thanks21:22
igcjelmer: I'd try the latest qbzr21:22
igcjelmer: you on karmic now?21:22
jelmerigc: I'm on Debian sid/experimental21:24
jelmerigc: but that should have mostly the same (and newer) packages as karmic21:24
igcjlemer: right. I think pyqt 4.6 introduced some changes which triggered some issues21:24
igcjelmer: when explorer runs, try Help > About to see the version of PyQt you're running21:25
jelmerigc: Yeah, I'm indeed running 4.621:25
rubbsHello all. I'm doing a presentation for my local Linux Users Group on DVCS's and I had a few questions about some things I had trouble finding in the documentation.21:53
jelmerhi rubbs21:53
rubbsFirst, I am a little fuzzy on what packing does. My guess is that it makes the storage of the meta data more efficient, but I'm unsure as to how that works.21:54
rubbsHello Jelmer.21:54
rubbsSecond, if one were to track binary files, and a small change was made (say one line in a .odt file) does the storage double, or can it keep a diff stored similar to text files?21:56
rubbsThe answers don't have to be bzr specific and any info on the differences about these questions in relation to git and hg would be greatly appreciated as well.21:58
lifelessrubbs: packing in bzr < 2.0 just reduces latency21:58
lifelessthe db format consists of a number of files on disk21:58
lifelesspacking coalesces these files21:58
rubbsah thank you! That makes sense21:59
rubbsdoes this mean it's safe to delete the packs in the "obsolete packs" directory?21:59
lifelessin the 2a format it will recompress as well, as the compression algorithm isn't as fixed as it was21:59
lifelessyes22:00
lifelessthat directory is used because there isn't such a thing as 'write barriers' on e.g. FTP/SFTP servers22:00
lifelessand fsync() on ext3 is appallingly slow22:00
rubbsOk that makes sense to me then.22:00
rubbsthank you22:01
rubbsbtw, I'd really like to thank the developers of bzr. I love it. I use it at work all the time.22:02
lifelessnow, for binary files22:03
lifelessodt files are zipped22:03
lifelessso the size of a commit of a small change in the doc will be a large [size of .zip] delta22:04
lifelesssadly.22:04
lifelesshowever, we'd like to allow unpacking and deltaing the text22:04
rubbsThat's what I thought, but I wanted to make sure before I said so in my presentation.22:04
rubbsThe reason I asked about the binaries, was I'm trying to find reasons why a normal Linux user (non-developer and non-admin) would like to know about bzr/git/hg. I'm having trouble coming up with usage ideas for them (if there is any) now that binaries are for sure not saving space for each version, I'm running out of ideas.22:08
lifelessso, space isn't a reason22:09
lifelessbut organisation22:09
lifelessbackups22:09
lifelesslogging22:09
lifeless'what did I do yesterday'22:09
lifelessetc22:09
lifelessa VCS lets you rollback time, in a way that *some* apps allow, but not all - but the VCS lets you do it for all22:09
rubbsThat's a good point. I'll put that in for sure22:11
rubbsThank you for all your answers. I really appreciate it.22:12
nyurubbs: it also helps with coordination22:12
nyuwhen different people want to modify the same code at the same time22:12
rubbsI also had thought of the fact that you could push a "history" of any file to the next maintianer of whatever project you may be working on (even word docs)22:13
rubbsgot to go eat, but I'll be back. Thanks again.22:14
MethsIn a pull or update what does a * next to a file mean?22:32
lifelessexecute bit changed22:33
MethsAh, okay.  Thanks.22:33
igcbbl23:02
jelmerLOL23:33
jelmerhaving both hg with hg-bzr and bzr with bzr-hg loaded is a bad idea :-)23:33

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