[01:39] morning all [01:39] * fullermd waves at igc. [01:39] Good morning. [01:39] hi fullermd! [01:40] It's so peaceful 'round here while everyone else is off summitting ;) [02:10] What does an uncommit actually do? I had a pull get aborted last night and repository is in inconsistent state. [02:11] Returns true for has_revision(X) but doesn't actually contain that revision. [02:11] I'm wondering if I should try and salvage the repository or not. [02:12] uncommit doesn't do anything to the repo, just the branch. [02:15] Peng_: Unfortunate for me. Maybe I can cheat and try to add the revision again, it's not in a consistent state so it might work. [02:29] hi, can somebody help me figure out whats wrong with my copy of Bazaar? [02:29] Tim-7967: what's up? [02:29] when I type bzr init I get this: http://pastebin.com/HVtDw2aZ [02:29] obviously its a problem with Python on OSX [02:30] but I've tried installing Bazaar through the DMG and MacPorts... and neither seem to work and produce the same error [02:30] ok, so you are using OSX? [02:30] yes, 10.6 Snow Leopard [02:30] and how did you install bzr? [02:31] originally through the Disk Image available at the website [02:31] then I tried it through MacPorts after that error showed up and had the same results [02:31] * thumper knows nothing about OSX [02:31] ok [02:31] those errors are saying that some plugins aren't working [02:31] think of screwed up BSD, and then you have something to reference too :P [02:31] but that won't stop bzr working [02:32] I'll just be noisy [02:32] the plugins should be installed in ~/.bazaar/plugins [02:32] if they are local [02:32] are they local or system plugins? [02:32] lemme check [02:32] jbowtie: What format repo is this? It should not be possible for it to become inconsistent. [02:32] bzr plugins -v [02:33] jbowtie: Hey, could you take a backup before you do anything to it, in case someone wants to debug it? [02:33] I would think it would be system as its refering to the Library folder [02:33] Yes, they are system plugins [02:34] which version of bzr are you using? [02:34] bzr 2.1.1 [02:35] hmm... [02:35] did you install the pipeline and rebase plugins separately? [02:35] no [02:36] they *should* have been installed automatically [02:36] I selected install all when i was installing Bazaar [02:36] Peng_: Will see, might have too much internal data in it to pass around though. [02:37] it looks then like there is a bug in the installer as the versions it is installing aren't compatable [02:37] an email to the bazaar mailing list may be in order [02:37] I don't know who looks after the OSX bits [02:37] or a bug in Launchpad [02:37] Peng_: It's a 2a repo, but it was converting from a TFS repo; probably a bug in my TFS code caused the abort. But would have thought revision would not have committed in that case. [02:39] ok, trying out the beta one and seeing if it makes any differance [02:40] ok [02:40] "*** Bazaar has encountered an internal error. This probably indicates a [02:40] bug in Bazaar. You can help us fix it by filing a bug report at [02:40] https://bugs.launchpad.net/bzr/+filebug [02:40] including this traceback and a description of the problem." [02:40] maybe not :P [02:42] * Tim-7967 casually ignores the additional warnings and downgrades to stable release branch [02:53] -_- [02:53] I've had more problems with bzr then git, cvs, svn and several other package managers combined :P [02:53] jbowtie: OK, import bugs could create a wildly invalid repo, so that could be it.. [02:54] I'll wait a little while for some bugs with bzr and OSX 10.6 to be fixed before using it [02:58] :/ [02:59] did somebody just bring the launchpad site offline? [02:59] "Firefox can't find the server at launchpad.net." [03:01] Tim-7967: WFM [03:01] WFM? [03:01] Works For Me [03:02] it must have been my DNS [03:02] I just flushed it and it was fine :P [03:02] * Tim-7967 curses OpenNIC and its tendancy to screw up DNS every 3 hours [03:02] Then why use it? [03:03] because its awesome [03:03] * Tim-7967 also points to cloak [03:03] Heh. [03:04] * Tim-7967 exclaims! [03:04] I've had my project for ~1day... working on uploading stuffs [03:04] and somebody already forked it 0_o [03:04] https://launchpad.net/nyx [03:06] what do you mean someone forked it? [03:07] Tim-7967: oh you have startcommander-x? [03:07] *StarCommanderX :P [03:07] I need a new name for it anyways [03:07] I used to call it starbot [03:07] and did it in PHP [03:07] it always screwed up though === dabomb69 is now known as Chazz [05:10] Holy crap. The user experience for bazaar in Windows is terrible... [05:15] rockstar: Really? I've had excellent experiences with it. [05:16] rockstar: Anything specifically bothering you? [05:16] Well, it IS on Windows... [05:19] jbowtie, lots of things. I'll file bugs and whinge to jam in the morning. [07:25] jam: http://pastebin.com/CZBzX4Lq [07:28] parthm: http://pastebin.com/eLuhbYR2 [07:29] "bzr info nosmart+lp:bzr" is shorter [07:29] parthm: 2a MySQL? :D [07:32] Peng_: Yeah :) ... I wanted one for testing [07:33] Peng_: I couldn't upgrade it locally so I push it into +junk and launchpad did it nicely. [07:39] Pah, kids these days. I did it locally before LP supported that. [07:42] Not literally locally. I got a VPS with 8 gigs of RAM... [07:49] Peng_: Nice :) [08:03] branch/pull themselves weren't that bad, but "bzr check" used like 5 gigs. [08:03] Fun times. [08:38] Peng_: yeah, check definitely needs some work.. [08:38] lifeless: let me know when you have time to look at foreign stuff :-) [08:40] jelmer: come on by, the waters warm [08:41] Version control slash fics? [08:45] lifeless: *nod* [08:46] Peng_: Isn't that what rebase is for? :p [08:53] lifeless: https://code.edge.launchpad.net/~spiv/ubuntu/lucid/paramiko/address-families-579530-lucid/+merge/25297 :P [09:06] spiv: your next step is to ping james_w or another main committer [09:10] lifeless: ta [09:21] jam: ooh, we can delete loggerhead.apps.config now i think [09:21] this makes me happy, because that is _horrible_ code === nlisgo_ is now known as nlisgo [09:36] mwhudson: because it was used only in start-loggerhead? [09:36] jam: right [09:36] It is still in the test suite, but we can fix that :) [09:36] oh [09:36] well yeah [09:36] I'm trying to clean it up now, because there were a lot of places that weren't passing 'cachepath' which is no longer optional [09:36] and its being a little bit tricky [09:36] but I'll get there [09:37] (we have a temp dir to put the work in, just having some difficulty cleaning up after each test) [09:37] mwhudson: but yes, only one reference to Root is in the test suite that I can find [09:38] mwhudson: oh, and that ref actually isn't ... referenced :) [09:38] 'loggerhead.tests.test_simple.test_config_root" doesn't seem connected to anything [09:38] \o/ for delete paths [09:41] jam: BTW, I haven't looked into this much at all, but I occasionally see a KeyError in get_revno. [09:41] Haven't pulled the latest code -- maybe it changes things. [09:42] the *latest* code is currently broken, but I'm finishing up some edges [09:42] Heh. [09:42] I've been working on it with mwhudson this week [09:42] I just merged history-db code into loggerhead so it isn't an external dependency [09:42] and some other stuff like that [09:42] I'm without email at the moment, which makes it hard to keep up with things, but ping me on IRC when you have something. [09:43] k [09:43] Thanks. :) [09:44] Haha, I just noticed this on the line above the KeyError: # TODO: Should probably handle KeyError? [09:53] jelmer: https://lists.canonical.com/archives/bazaar/2009q2/059263.html [09:55] imported from CVS, how to add more changes? We use cvs2svn to import a project from CVS. Unfortunately, som people continued in CVS, and others used bazaar. === nlisgo_ is now known as nlisgo [10:09] abently I get the following when running `bzr help commands` today : ValueError: No help message set for amanica: interesting; it means a proactive code check is catching a plugin without help on a command [10:15] please file a bug on bzr-pipeline [10:15] preferrably with a patch :) [10:15] * fullermd filed that on -keywords last night. === vila is now known as mgz [11:26] Hi. I have a post-commit.sh designed for svn. It uses svnlook (e.g. svnlook log -r2 only returns the commit message). Is there a plugin for BZR that provides the same functionality? [11:27] hopeseekr: What is your post-commit script actually doing? [11:27] it's only 3 lines [11:28] http://pastie.org/960073 [11:29] i guess the format i want it in is Author RevNo "Commit Message" "changed files" [11:32] actually, i have no idea how to make a bzr post-commit hook, so this may all be moot [11:32] its done via python [11:32] you write it as a small script in ~/.bazaar/plugins [11:33] there are some examples [11:33] o great; i absolutely suck @ python [11:33] i remember tyring to fix a *simple* bug in bzrweb; no go [11:40] Peng_: so lp:~jameinel/loggerhead/history_db should be usable, and no longer depend on having bzr-history-db installed. I still have some more cleanup, and I know at least one view is broken. But I'm getting there. [11:53] OK! [11:59] jam: https://code.edge.launchpad.net/~parthm/bzr/538868-message-for-heavy-checkout/+merge/24483 [12:30] imported from CVS, how to add more changes? We use cvs2svn to import a project from CVS. Unfortunately, som people continued in CVS, and others used bazaar. [12:34] cvs2svn doesn't support incremental conversion. [12:35] I s'pose it's possible that it will output stuff similar enough that bzr fast-import can do incremental updates; I think it has some level of support for that. [12:35] But I wouldn't count on it. [12:37] fullermd: I was thinking I could do another import to new branch in bzr, and then copy changesets from the old one [12:37] Well, the new branch would be unrelated. Maybe rewrite could do something with that; I don't know. [12:40] Hi does anyone use Bazaar on Mac Snow Leopard? [12:41] and if so, how can I uninstall it if the uninstaller doesn't work on Snow Leopard? [12:48] fullermd: yes, metadata-wise it would be unrelated. I thought I could generate patches in one branch, and then apply them in the other [12:53] something like what a gatekeeper would do [12:54] Yeah. It's possible something in the rewrite plugin can help automate that. [13:04] Erm, no one? === nlisgo_ is now known as nlisgo === Chex- is now known as Chex [14:47] abentley: re problem with encoding of diff header again. from DiffText.diff_text method can launch either internal_diff or external_diff function. but external one has no path_encoding argument, so new code to pass the encoding down the layer is failing with external differ. will it be ok to just add path_encoding argument to external_diff function, although not really use it? [15:20] lifeless: re your blog post [15:26] poolie: yes ? [15:26] hi [15:26] just thinking about your useFixture thing [15:27] to me 'fixture' is the whole environment where a test runs [15:27] so it's a bit strange if you can assemble some of them [15:27] i don't know if this is the orthodox meaning [15:37] poolie: where are you right now? [15:38] bialix: cocobolo 3 on the ground floor [15:38] k [15:40] will be back abotu 5pm [15:40] lifeless: you should +1 my blog comments === nlisgo_ is now known as nlisgo === IslandUsurper is now known as IslandUsurperAFK === IslandUsurperAFK is now known as IslandUsurper === IslandUsurper is now known as IslandUsurperAFK === beuno is now known as beuno-lunch === nlisgo_ is now known as nlisgo === nlisgo_ is now known as nlisgo === radoe_ is now known as radoe === IslandUsurperAFK is now known as IslandUsurper === Pilky_ is now known as Pilky === beuno-lunch is now known as beuno === nlisgo_ is now known as nlisgo [18:18] Hi, I'm trying to see what changes a remote repository has [18:19] I seem to remember that 'bzr incomming location' was it [18:19] but that seems not to work [18:19] and searching the docs doesn't seem to give enything? [18:19] missing [18:19] ha [18:19] 'incoming' is hg I think. [18:19] confused by differences again [18:19] thanks fullermd [18:25] on a related question, is there a fast way to find the bzr branch root? [18:25] I'm currently using bzr info and then parse out the branch root: [18:25] but that is way too _slow_ to be in my prompt [18:27] Well, there's 'bzr root'. [18:27] But I suspect pretty much anything will be too slow to be in a prompt. [18:27] Well, unless you load bzrlib into your shell so there's no startup overhead ;) [18:28] fullermd: yeah, pretty much [18:28] and on osx I don't have bzr-service which loads it in the ram for me. [18:28] Or I'm missing something [18:35] * dwt thinks I could try storing the bzr source on a ramdisk [18:35] Well, I'd think that after running it once or twice, it would all be in cache anyway. [18:35] It's the python startup overhead that gets you. [18:37] yeah, I pretty much think so [18:37] I wonder how the mercurial guys get this [18:37] they must be bitten by the python startup overhead too [18:37] but somehow it's much less of an issue there [18:38] my prompt pretty much takes a second to build (bzr) and about one order of magnitude less time in mercurial [18:38] Well... a quick bit of find'ing and wc'ing says that my installed hg has around 17k lines of .py files, and bzrlib is 101k. [18:39] bzr's all in cache here, and 'bzr rocks' consistently takes ~110ms (and I don't exactly have a pokey box) [18:39] bzr help is about the same; 110-120. hg help is 30. [20:07] fullermd: they both take around 300ms here [20:07] (hg help and bzr help, that is) [20:08] Python startup pretty much sucks [20:09] I'm guessing my local python configuration is adding some overhead that you don't have, dwarfing the actual differences between bzr and hg === nlisgo_ is now known as nlisgo [23:42] hi. i would like to do bzr merge -c 15525,15526,15528,15532,15535 ../feature_pref_edition_block/ but it does not like me doing that :) [23:57] or is there an interactive merge tool where i pick hunk by hunk maybe? [23:58] chx_: you can only do one rev or range of revs at a time [23:59] chx_: use a bash for loop?