[00:05] is it usual to use bzr commit --fixes=id more than one time for the same bug? [00:24] * jelmer sleep [00:47] I'm trying to upgrade Twisted's bzr mirror to 2a [00:48] it seems to be stuck halfway. would I be better off just creating a new repo and branching into it and/or using svn-import than to do bzr upgrade? [00:57] I found upgrade to not be all that fast [00:57] I haven't noticed stuck though. [00:58] this is on our dinky little VM [00:58] I think it took several minutes or so on a 20M repo with about 5000 revisions [00:58] (on a basic linux system, not a vm) [00:58] dash: doing the import again with bzr-svn will probably be just as fast [00:59] mwhudson: yeah ok [01:10] does anyone know how to toggle bzr-builddeb between behaving like debuild -S -sd and debuild -S -sa? [01:19] I want to upgrade lp:tortoisebzr to 2a format [01:19] I read upgrade guide. http://doc.bazaar-vcs.org/bzr.dev/en/upgrade-guide/index.html#migrating-branches-on-launchpad [01:20] The guide says "On Launchpad, unset the current trunk from being the development focus." [01:20] I don't know how to unset "development focus" [01:21] Clear the Branch text box at https://launchpad.net/PROJECT/DEVFOCUSSERIES/+edi [01:21] +edit, that is. [01:22] Wow! [01:23] lp:tortoisebzr is now 2a! [01:23] Thank you! [01:26] ok so i've got a 2a edition of trunk now [01:27] if i had permissions i'd put it on launchpad [01:32] whoever here is Ian: I'm almost done with the 10.6 package. It's a "core" package (no frills, basic set of plug-ins). [01:33] jfroy: Which Ian are you looking for? [01:33] Ian Clatworthy [01:33] jfroy: That's igc, unless I'm making some horrible mistake. [01:36] Hi [01:37] I've got a unshelving question. [01:37] $ bzr unshelve 1 [01:37] bzr: ERROR: No such file: None [01:38] I've read that it was a known bug, and I was wondering if anyone knows how to manually recover my shelved files. [01:47] Does anyone know about how to manually recover shelved files? [01:48] I guess I could read the shelving source code. [01:57] how do I display the contents of a single file from a past revision? [01:58] I don't actually want to bring that old file back, just want to see what was there [01:58] 'bzr cat -r' [01:58] damn, how did i not see that [01:58] thanks dash [02:01] also, i have a branch i'm currently working in that was branched off from the trunk, i haven't done anything in the trunk since, what should I do to essentially make the current branch the trunk? [02:01] should I push? or should I merge? [02:01] or perhaps switch and pull? [02:02] (my working dir is a lightweight checkout) [02:02] it points to the experimental branch that i'd like to merge/pull/push back into the trunk branch [02:03] and apologies if this is a stupid quetsion [02:03] *question [02:04] itistoday: i'd just merge it into the trunk [02:05] if you merge, 'bzr log' for trunk will show a single revision where you merged your branch, with the changes in the branch as subrevisions [02:05] if you just push it to trunk, bzr log will show all your commits to the branch just as if you'd made them to trunk. [02:06] dash: thanks, i guess that is useful to see [02:19] Does anyone have an idea about how to manually unshelve? [02:22] I have lost some important changes to code in a shelve that I need to recover. Ideas? [02:23] I don't think I've seen the issue you describe [02:23] have you searched the bug tracker for it? [02:24] Yes, I found this: https://bugs.launchpad.net/bzr/+bug/319790 [02:24] Launchpad bug 319790 in bzr "bzr unshelve crashes losing all changes" [High,Fix released] [02:25] It mentions manually recovering the changes, but not how to do it. [02:25] hman: what version of bzr are you on? [02:26] 1.13.1 [02:26] When was the fix applied? [02:27] hman: hmm, looks like the bug report doesn't say. Are you sure you're actually hitting the same bug (did you delete any files recently)? [02:28] There was a deleted file in the changeset. [02:28] I shelved on 1.3.1 and just upgraded to 1.13.1 but I get the same error. [02:29] I see the shelve-1 file but I don't know how to manually get to my changes. [02:32] I don't know what format that file is actually in, but last time I peeked at it it wasn't very human-readable. [02:35] Yeah. I'll keep digging around. [02:46] are the C extensions necessary -- in other words, is bzr written so that it can be run solely on python code? and if so how much slower is the performance typically? [02:48] rsvp: the c extensions are not necessary [02:48] i'm not sure what the performance impact is these days [02:48] the packages in the ppa have the extensions built, though === fenris__ is now known as ejat [02:52] rsvp: they are sufficiently unnecessary that there have been a few releases accidentally omitting their source from the source tarball, but it's not like they're hard to build either. [02:52] ("a few" being "at least two" but I forgot the actual number) [02:54] so going the source code route (not ppa packaging) one should make those C extensions, else there is some (significant?) performance hit. [02:55] Compiling the C extensions is as simple as a "make", so why not do it? [02:55] exactly [02:55] But you certainly can get by without them. It just won't be as fun. [02:55] also, why were you not using the ppa again? [02:55] BTW, with the 2a disk format, the C extensions will use less disk space. [02:55] (Line-based vs. byte-based deltas.) [02:55] ohh, I forgot about that. [02:56] Still totally compatible, of course. [02:56] * mzz is now mostly done cleaning up and upgrading all the branches randomly spread over his hds [02:57] I hope I never decide to do that. [02:57] well, it was not just the format upgrade that made this something I wanted to do. [02:57] also, there weren't *that* many. Although there was some embarassing ancient code in there. [02:58] hey guys: i'm finding the instructions on how to publish a repo over http, a bit confusing: 0) there is no bzr-smart.py script readily available 1) it doesn't look that bzr likes multiple repos from the same root, and 2) it doesn't look that there's a landing page [02:58] what instructions? [02:59] these instructions: http://doc.bazaar-vcs.org/bzr.dev/en/user-guide/index.html [02:59] mzz, thanks for your response yesterday by the way... I wanted to understand the comparison with hg (without C extensions) ... [03:00] project2501a: what section? [03:00] mzz: under 'mod_pythong' [03:00] d-oh. freudian slip [03:00] mzz: under 'mod_python' [03:00] that tong tong tong [03:00] rsvp: mercurial has c extensions, although I don't know if they're also optional [03:01] mzz: They are, but unless something changed while i was away, it's a bit of a pain to use the Python versions. [03:01] project2501a: afaik mod_python isn't the best choice for serving stuff. Is the manual recommending it over other options? [03:02] project2501a: The smart server script is given in the docs. [03:02] no, it's not. but the manual is not offering much guidance or solutions [03:02] project2501a: What do you mean about "multiple repos from the same root"? [03:02] like hg or git does: gives you a root dir, ie, http://marsel.is/repos/git and you can have multiple repos under that url [03:03] each independant of the other. plus it serves as a landing page [03:03] mzz, I think the early hg versions required c extensions. [03:03] project2501a: the script is given below (under "configuring bazaar", subsection "mod_python") [03:03] rsvp: That's correct. [03:03] project2501a: but I'd use fcgi. and afaik you can just expose more than one smartserver instance that way (although I don't think it automatically creates a landing page) [03:04] yes, i saw it. that's why i said "there is no readily available script" [03:04] project2501a: there are plugins (loggerhead mainly) that serve html to humans. [03:04] project2501a: You can have as many branches and repos as you want... [03:05] Aside from bug 348308, which broke bzr+http with shared repos.... But that's not the same thing. And it can be worked around anyway. [03:05] Launchpad bug 348308 in bzr "Smart server jail breaks bzr+http with shared repos" [High,Triaged] https://launchpad.net/bugs/348308 [03:05] maybe i should go back and read some more [03:07] maybe i should take it with the bzr maintainer in debian, cuz apparently neither the bzr nor the bzrtools package includes bzr-smart.fcgi [03:08] I'm thinking of switching over to mercurial because I get the impression the python(.org) community is going that direction (under Guido's nudge wink wink say no more) -- any good counter-arguments? [03:08] don't be a sheep [03:08] rsvp: get a brain [03:08] project2501a: notice that script is also included in the manual, and mostly consists of configuration [03:08] project2501a: I don't think including it makes sense [03:08] sheeple are annoying [03:08] rsvp: do you hack on python itself? [03:08] SamB_irssi: reddit rules [03:09] rsvp: i've used hg, main reason i stick with bzr is that I think the way it handles branches and merges is better [03:09] mzz: i'm a sysadmin mate; i code, but my bread and butter is sysadmin. inclusion would make my life easier. happy sysadmin == happy office [03:09] unhappy sysadmin == wrath of root. [03:09] project2501a: I guess it could live under /usr/share/doc/bzr/examples or something [03:10] it could. it doesn't [03:10] that's what I meant [03:10] <3 [03:11] my point is that just running it doesn't make any sense, you have to edit roughly half of the roughly a dozen lines in there. So sticking it in /usr/bin or the like would just be weird. [03:12] Mine is 130 lines long. How did that happen? >.> [03:12] rsvp: neither mercurial nor bzr is likely to go away anytime soon, so what projects you don't hack on use isn't important [03:12] Peng_: serves too many repos? [03:12] from my humble BOFH point of view is that svn, hg and git took me 40 minutes each to set up, and that's because i was lazy. and bzr is currently taking the piss [03:13] hmm [03:13] i understand what you are saying, by the way [03:13] project2501a: iiuc it's just a wsgi app, and there are multiple ways to actually expose those to the net. I guess the manual doesn't go into as much detail as it could about how to do that. [03:14] i'm not trying to argue against you, i'm just mounting a retort [03:14] mzz: There isn't per-repo stuff. It's just customizations: mostly logging, working around that bug, and something else. [03:14] mzz: that is correct mate. [03:14] (and I'm assuming just plain "bzr serve" doesn't suit your needs) [03:14] no [03:14] that would mean more ports open [03:14] bad karma, sysadmin gets less sleep [03:14] nod [03:14] so there's a bit of a documentation hole there. [03:15] What's wrong with opening a couple ports? [03:15] the size of a small moon. [03:15] You have the exact same exposure either way. [03:15] Well, I suppose it would take some bot attacker longer to find a bzr+http server. [03:15] But they both do the same thing. [03:16] Anyway, what were we talking about? [03:16] no you don't. much much easier to use Simple Event Correlation with one port and tell SEC to ban an ip for 40 minutes when the IP is takign the piss [03:16] Peng: reddit [03:16] that's what we're taking about [03:16] specifically /r/jailbait/ [03:18] mzz, an important factor is the format of the (most popular/useful/institutional) repositories -- if hg or bzr had plugins which makes the interchange convenient then there's no question, however, that might be incompatibilities which could be annoying down the road. [03:18] rsvp: fastimport is supposed to help with that [03:19] it's 2009 [03:19] rsvp: although I once actually used that and hit some bug on removal of files, so I know what you mean [03:19] you're going to have to use more than one revision control system, anyway [03:19] what bob said [03:19] these days SCMs tend to be a trend anyway [03:19] but yeah, just make sure you at least pick a dvcs that isn't totally obscure and you'll usually be ok. [03:20] no monotone [03:20] no? [03:21] factors include which way is guido pissing, if joel got laid last night, and if paul graham wrote another essay portraying hackers as painters [03:21] project2501a: but ... nobody would read it [03:21] oh and if the stock for va linux is still plummeting or not. ESR gets frisky about that [03:21] SamB_irssi: lol :) [03:21] they would not realize it was new [03:21] and believe that they had already read it [03:22] i tell you there are people who are total fan boys of graham [03:22] they would re-read it and suck his dick in the process [03:22] ... okay [03:22] and then declare that he's a genious can we have more funding from y-combinator? [03:23] * SamB_XP_ amuses himself by following his own newly-minted blog [03:23] you have a blag? [03:24] http://naesten.blogspot.com/ -- just one post, started the thing because I'm sick and tired of having to fish for lambdacats when I want them ... [03:24] SamB_irssi: yo dog, i'm gonna let you finish, but i just want you to know that blogspot sucks [03:24] project2501a: yeah, sure [03:25] I guess I knew that [03:25] but if I used something else, I would have to decide what! [03:25] you can use a blog inside a blog so you can blog while you blog [03:25] Use WordPress! Constant security problems make life fun! [03:25] yeah, php! it's not a language, it's a brand new estate for it's creators! [03:26] project2501a: wouldn't putting a toilet in the blog make more sense? [03:26] that's putting a toilet in the blog so you can defecate while you blog [03:26] SamB_XP_: i take it you're not up to date with your memes [03:26] I was thinking of words more like "shit" and "crap" [03:26] that's slashdot mate [03:26] project2501a: I can't remember all the ones I've even heard of [03:27] I accidentally the memes [03:27] all of them? [03:27] and I don't read slashdot very often [03:27] you're not missing anything [03:27] project2501a: I them all [03:27] taco is vested, cowboy neal likes young boys [03:28] chuck norris doesn't like where this is headed [03:28] talk to jack cuz the hand aint hearing it [03:28] bob2, it a real hassle with multiple version controls -- I'd go with the one that can read the others the most without hassle. [03:29] wow, big surprise, rsvp is recommending bzr in #bzr ;-) [03:29] or hg [03:29] voila [03:29] hg can read other VCS branches? [03:29] rsvp: do you use vi or emacs? [03:29] what other VCS's branches can it read ? [03:29] vim [03:30] SamB_XP_: svn afaik, but i'm just taking the piss here [03:30] yeah, ok, time for sleep [03:30] so [03:30] no landing page [03:30] no easy way to publish repos over http [03:30] project2501a: okay, that's about what I thought I might have read [03:31] no multiple repos under a single root [03:31] no easy way to publish repos over http ... what ??? [03:31] it's not harder than with darcs! [03:31] sorry, it's kinda late here and my brain is slowing down [03:32] SamB_XP_: no easy way to publish multiple repos under a single root over http [03:32] ^-- better? [03:32] project2501a: sure there is ... you just push them into subdirectories of a directory in your www tree ... [03:33] There's even a plugin to push a whole directory of branches, I think. [03:33] project2501a: bzr push bzr+ssh://webhost/srv/www.site/b1 [03:33] etc [03:33] SamB_XP_: that doesn't make them independent repos [03:33] project2501a, (but you see text is text, to both vi and emacs -- and one chooses the unix newlines in vim because ms can handle that but NOT vice versa -- so always go with the set that encompasses the other) [03:33] project2501a: what do you mean? [03:33] project2501a: what do you mean 'independent' [03:33] anyway, i apologize. as i said it's late over here. i'm gonna go take a nap and i'll be back in the morning [03:33] I mean, you just stick the directories in a directory [03:33] brain no worky worky [03:34] oh, you feel like I felt this morning maybe ? [03:34] have you been feeling like you should stop being a sysadmin doing admin spotting [03:34] ? [03:34] well, no, but there's a reason I posted that lambdacat today [03:35] pissing your last in a miserable newsgroup, nothing more than an embarrassment to the f*cked up lusers gates spawned to replace the computer literate? [03:35] heh [03:35] import import_beer; import more_women; [03:36] project2501a: could you tone down the language please, we like a friendly channel [03:36] * SamB_XP_ generally tries to do his pissing into the toilet [03:36] oh, the p****g part [03:36] that's just brittish idiom [03:37] could mean wasting [03:37] anyway, long live the GPL, see you tomorrow morning where i will have had more sleep [03:40] no, down with the GPL and up with SPLs! [03:45] the other interesting thing between mercurial and bzr... if one uses Ubuntu, bzr shows up in updates in the week of the release, whereas for hg you can probably hang on to your stale version for a longtime. [03:45] rsvp: thats because we push updates to Ubuntu; the hg community could do that if they wanted [03:46] how often does hg get a new release ? [03:47] its not time based [03:47] lifeless, I understand, that's the "institutional" aspect of releases. [03:47] I meant approximately [03:48] I wasn't meaning to imply time-based-ness [03:48] SamB_XP_: couple of times a year, more or less [03:48] and Ubuntu of course uses bzr, not hg, internally. [03:49] http://arcanux.org/lambdacats/tail-recursion.jpg nice [03:49] there is that [03:49] but its not why releases get to ubuntu :P [03:50] lifeless: well, it might be if you didn't push them so hard [03:50] * rsvp is LOL allowed in here? [03:50] SamB_XP_: in the past, they've trickled down from Debian [03:50] rsvp: after looking at a specialized type of lolcat ? yes! [03:50] lifeless: ah, sure [03:50] SamB_XP_: and thats still the process, just we get in and own it [03:51] ... so you're blaming this on jelmer then ? [03:51] blaming ? [03:51] Isn't hg trying to release more frequently now? [03:51] whatever word you want to use [03:52] sure thing Christopher Robin :) [04:20] bah, i have a problem. pushing ends with the message ConnectionReset, unexpected end of message. in bzrlib\smart\message.pyo:286 [04:20] the remote branch then has the lock set and loggerhead fails to display it [04:34] darn it [04:34] zsquare should know to wait longer than that at 11:30 EST! [05:13] is what i'm told true? you can make a checkout of the last x amount of revisions of a project and work with that? [05:21] SamB_XP_: 'specially on a weekend. [06:11] meoblast001: kindof, who is saying that [06:11] some people in #gamedev [06:12] well, did they give an example command line? [06:12] no [06:12] i saw you in there [06:12] just popped in for a sec [06:12] see if I recognised folk [06:12] so i suppose you know it's raining over here now :P === abentley1 is now known as abentley === abentley1 is now known as abentley === khmarbaise_ is now known as khmarbaise [10:31] say I have a a local branch with rich-root support and an upstream shared-repo which doesn't. How can I push my branch to that repo? [10:32] The branch was originally a branch of an svn tree which I'm not going to use any longer === johnf1 is now known as johnf [10:32] upgrade the the shared repo [10:33] lifeless: :) Yeah trying to avoid that. I'd have to bug everyone else using it. Which I suppose I could do [10:34] make a a new repo at the subdir you want to push to then [10:34] bzr init-repo URL; bzr push URL [11:08] A repo can exist within a repo? [12:47] helloooo [12:47] good morning [12:48] i'm trying to start loggerhead manually so i can test it. i get this error: bzrlib.errors.NotBranchError: Not a branch: "/repos/bzr/.bzr/branch/". any clues please? [12:52] project2501a: can you put your loggerhead.conf on a pastie somewhere? [12:53] sure mate [12:53] give me 2 clicks [12:54] johnf: http://pastebin.com/m7f0dd30b [12:56] what do you have in /srv/repos? [12:57] there's no such path in my machine mate [12:57] there's /repos/bzr [12:57] ahh sorry that's what I means [12:57] jibber:/repos/bzr# ls -la [12:57] total 12 [12:57] drwxrwxr-x 3 bzr www-data 4096 2009-09-27 10:18 . [12:57] drwxr-xr-x 6 root root 4096 2009-09-22 01:01 .. [12:57] drwxrwxr-x 4 bzr www-data 4096 2009-09-27 10:18 .bzr [12:58] did i init the repo wrong? [12:58] hmm does the repo have anything in it yet? [12:59] no, it's blank [12:59] i'll import code later [12:59] so there are no branches then [13:00] i'm setting up the infrastructure right now so i will be spanky-comfy later and not sweat about it [13:00] project2501a: which is what loggerhead was complaining about [13:00] yeah, no branches. [13:00] LarstiQ: yup. [13:00] its rather unsurprising to get 'not branch error' with no branches. [13:00] well, shouldn't loggerhead account for a brand-new repo? [13:01] and say "hey, there's nothing to serve here" [13:01] "empty page" [13:01] or something like that [13:01] if you run bzr serve --http, I think thats what it will do [13:01] LarstiQ: so, you're saying, just check in something? [13:01] loggerhead.conf is a different beast [13:01] lifeless: well yeah, i'm trying to make it all pretty :) [13:02] so, just check in something? [13:02] can i make just a silly branch without committing anything? [13:03] 'bzr init' [13:03] ' [13:04] ooh [13:04] i used bzr-initrepo something [13:04] my bad :) [13:04] thank you :) [13:04] init-repo makes a repository [13:04] repositories are an optional container for branches [13:05] it looks like i'll have to read up [13:06] my general advice for folk getting started with bzr [and contemplating large deployments etc] is [13:06] 'play with it'. put something small in, fresh import, no history, and fiddle round [13:06] very true mate [13:06] get used to the UI and how to push pull and merge with other branches. [13:06] i has ideas :) [13:07] yay, loggerhead now seems to work [13:08] now if i can just solve that damn [client 93.97.20.215] client denied by server configuration: proxy:http://127.0.0.1:8080/, referer: http://marsel.is/ [13:11] now, we're getting along :) [13:11] lifeless: thanks mate [13:11] thanks everybody [13:18] YAY! [13:18] works [13:18] i love you guys :) [13:18] if you ever come by sheffield, UK, i'll buy you a pint [13:18] http://marsel.is/repos/bzr/ [13:19] I'm glad you're up and running [13:19] gnight everyone [14:28] any ideas about the cause of this? http://paste.ubuntu.com/279609/ and whether a bug already exists? I can't understand python tracebacks [14:29] this occurs when I'm trying to branch into a shared repository [14:33] I'll be afk for a bit but picking up any hilights [15:12] mdke: known bug; launchpad needs to upgrade its bzr; fix is to upgrade your local repo format to match the remote format, probably 2a [15:13] mdke: oh, hmm, LP is running 2.0.0 now. [15:13] mdke: best bet is to file a bug report on bzr [15:14] mdke: workaround will probably still be making your local format match the remote one [15:15] mdke: but off the top of my head now that LP is running 2.0 on the server this shouldn't happen anymore. [16:45] does "bzr log" always return a local log if i have a heavyweight checkout? [16:46] just need to do a sanity check [16:46] spiv: thanks, so just a bzr upgrade? [16:54] i was doing my final cleanup on my branch, and got this http://codepad.org/bUF52JMr [16:54] it looks like my original branch had 73 commits too, so i'm confused [16:57] spiv:I wonder if it is a downgrade I need, my local repo is already using 2a [17:10] spiv: bzr upgrade --1.9-rich-root has done the trick [17:10] spiv: thanks! [18:44] is there a easy to use bzr client for windows like tortoisesvn? [20:01] spiv: ping are you here? [20:04] This bug: https://bugs.launchpad.net/bzr/+bug/109143 is fixed in 2.0.0 ? or in later version? [20:04] Launchpad bug 109143 in bzr "hpss does not support ~ (tilde) for home dir access on bzr:// or bzr+ssh://" [High,Fix committed] [20:05] I have installed 2.0 on client and server pc and does not seem to work [20:06] reading the comments, I don't think it's in bzr.dev yet [20:08] LarstiQ: It seems that this branched is merged to 2.0 https://code.launchpad.net/~spiv/bzr/bzr-ssh-homedir-take-3 [20:13] sque: ah, you are right === BasicPRO is now known as BasicOSX [20:41] LarstiQ: I downloaded the source of 2.0 and the patches are not in there. I checked the lp:bzr/2.0 branch and no patches are there too. The patches are merged at the lp:bzr (the dev branch). So this bug will be fixed in 2.1 probably [20:41] damn. [21:06] Morning [21:06] When I try to download the latest version of launchpad, the bzr gives me this error [21:06] ./rocketfuel-setup: line 372: 19633 Segmentation fault bzr branch http://bazaar.launchpad.net/~launchpad-pqm/launchpad/devel/ $LP_TRUNK_NAME [21:07] Is there a way to install bzr 1.18 instead of 2.0 ? [21:13] Fly-Man-: Branch 1.18, compile the extensions, and bung the folder on your PATH [21:14] awilkins: No other way ? [21:14] because for some reason, it's stuck with either 1.16 or 2.0 [21:15] You could use a source tarball, I suppose [21:16] uhm, i'm more a apt-get person ;) [21:16] not a source tarball eater [21:17] Well, I bet you a cookie on a stick it's quicker than the fiddling with apt configuration to make it stick to a particular version [21:17] Hey guys, i'm getting a wierd bzr-buildpackage error. bzr: ERROR: Could not find changelog at "debian/changelog". [21:17] * awilkins downloads tarball [21:17] does anyone know what the deal with that is? [21:18] there's definitely a changelog.. [21:19] awilkins: Already found a ppa package that has 1.16 [21:19] thanks :) [21:19] * awilkins deletes tarball [22:27] moin [23:09] verterok: ping [23:19] lifeless: Do you know the process to get a package pulled into ubuntu during freeze. Namely bzrtools [23:20] requestsync --lp -e bzrtools [23:25] hmm is that the right process? -e Use this after FeatureFreeze for non-bug fix syncs [23:25] this is technically a bug fix right? [23:27] johnf: then no -e [23:27] will that give it the right priority to make sure it happens for kamic though? [23:29] orthogonal concerns [23:29] 1) get the bug filed, 2) NAG AN RM [23:31] hmm. Does someone else have time to do it. I have to run out the door and requestsync is core dumping on me right now [23:33] doing [23:49] thanks [23:49]