[00:03] but if a replay fail due to a conflict there is no replay continue like rebase [00:04] visik7, yes, you can use "bzr rebase-continue" [00:04] ah, no [00:04] nevermind [00:04] :\ [00:04] you can commit and run "bzr replay" again though from the point where it aborted [00:05] nevermind seems it's working [00:05] ok I think is ok [00:06] now I need to clean it up branching again ? [00:06] poolie: I'd like to rename the (about to be released) diffinternals plugin to crosscheck. Is that ok with you? [00:06] igc, that sounds better to me [00:07] poolie: cool. me too. [00:08] visik7, yeah, branch it to a different location [00:08] visik7, (replay wasn't really designed for this sort of thing) === mw______ is now known as mw____________ === mw____________ is now known as mw [00:27] jelmer: I see anyway it worked so, thanks for the help and hope soon we will have something like git-filter-branch [00:33] visik7, cool === kiko is now known as kiko-zzz [00:38] can anyone give me a helping hand.. bit confused as used to svn and cvs(spits)....... [00:38] good night thank you again jelmer [00:39] p_masho, hi [00:41] hi jelmer.. late at night here in wales.. and got this zim on my ubuntu box - http://bazaar.launchpad.net/%7Epardus-cpan/zim/pyzim/ [00:41] so I been hacking away... and have already modified and not creaed a branch... [00:42] I'm looking for 1) - equivelent of svn copy foo.x bar.x ? [00:43] p_masho, copying a branch or a file/directory ? [00:43] p_masho, to copy a branch, use "bzr branch

" [00:43] to copy a file, simple copy it using cp and then add it [00:44] Ok. advice.. am used to svn.. so be gentle.. please [00:44] I go this branch bzr branch lp:~pardus-cpan/zim/pyzim = /home/zim/pyzim on my machine [00:45] so I hacked that a bit and got a diff file..... [00:45] p_masho, what are you trying to copy? a file? [00:45] What I want to know is how I can create a new "version" on my machine copying all the mods I got [00:46] and then restore the master branch to tha release from launchpad... [00:46] sorrry if this sounds confusing... [00:47] Suppose the question really is.. first .. [00:47] sorry, phone.. [00:47] How do I branch off the mods I already done and restore the "main/HEAD" to main repos [00:48] p_masho: to make a new branch, 'bzr branch' [00:48] p_masho: e.g. 'bzr branch . ../new-branch' [00:48] p_masho: if you haven't committed your changes, do so first [00:49] p_masho: I don't know what you mean about the main repos, you haven't altered the mainline afaict [00:49] thats the problem (as sv user) I made changes to head... [00:50] you have? [01:01] * Sidnei is away: I'm busy === mw is now known as mw|out [01:23] * igc lunch, etc. - bbl [01:34] poolie1, hi [02:37] hi jelmer [02:37] spiv, thanks for that note! [02:49] poolie1: you're welcome :) [03:14] * spiv -> lunch [03:16] spiv, still here? [03:17] to build on that mail, i'd like to get a page like the smart push analysis, that for each scenario says what operations we _should_ be doing [03:20] in other words: we have a list of what is done at the moment and we have some vectors we can apply to them but i'd like to see where the expected endpoint is too [03:25] Yay. "bzr heads" + ack + sed + bash == easy upgrade from subtrees to rich-roots. [03:25] "bzr upgrade' would've been even easier, of course. Whinewhine. [03:26] since lifeless doesn't appear to be here... [03:26] Peng_: "Patches welcome" :-) [03:26] * Peng_ cries [03:27] Wait, I was upgrading a copy of bzr-svn, so I can blame you for putting me in this situation in the first place! :P [03:27] * Peng_ hides [03:27] Peng_, recursive-upgrade? [03:27] That was actually pretty painless. :) [03:27] beuno: ? [03:27] I know jelmer doesn't want you to know about it [03:27] but he wrote a plugin [03:27] http://people.samba.org/bzr/jelmer/bzr-recursive-upgrade/trunk [03:27] Oh? === abentley1 is now known as abentley [03:28] Peng, subtrees is more than just a version string [03:29] Peng, the inventory format is also different afaik [03:29] jelmer: So? I don't care. I just want to be able to convert my repos easily. I can live with it burning some extra CPU. [03:30] * Peng_ gets grumpy when thinking about this. :P [03:30] Peng_: I wasn't sure what you were doing exactly, but just editing .bzr/repository-format isn't sufficient [03:30] jelmer: Yeah, I know. I didn't do that. [03:30] Ok, I'll shut up then :-) [03:31] I used "bzr heads" to grab all of the heads, created a new repo and a branch, and used "bzr pull --overwrite -r revid:..." to copy all of the revisions into it. [03:33] And...let's see...frozenset(bzrlib.branch.open_containing('.')[0].repository.all_revision_ids()) to verify it went well. :) [03:34] lifeless: can you point me to a mail about dev3 making some loggerhead pages much faster? [03:39] poolie: no, there isn't one [03:39] poolie: there is a branch with a merge request in LP that makes lh compatible with bb-core [03:40] poolie: and its --development4 that works for lh, fwiw [03:41] lifeless, about that, I didn't quite understand what you meant [03:41] it breaks with the current bzr [03:41] what is it that I'm missing? [03:41] beuno: works for me [03:41] * beuno tests again [03:42] lifeless, hrm, it seems to as well now for me... [03:42] I'll run the tests and merge [03:43] (back) [03:45] beuno: :P [03:45] lifeless: but it has been measured to make some pages faster? the diff display, i'd presume? [03:45] lifeless, so I had a b0rked copy of bzr? [03:45] poolie: unscientifically yes [03:46] poolie: revision display for instance [03:47] lifeless, merged. Thanks for the patch [03:48] 101 branches owned by 24 people and 1 team, 75 commits by 12 people in the last month [03:48] ^ Loggerhead [03:48] * beuno dances [03:48] :) [03:51] beuno: no probs [03:51] Hm, tempting to write a bazaar branch browser django app, but laziness D: [03:56] reuse loggerhead? [03:57] I should try that. [03:58] * beuno thinks "loggerheadlib" and looks at thumper [04:00] why me? [04:00] beuno: JFDI [04:00] thumper, sure sure, and world peace, and all that [04:00] BUT [04:01] there are some interesting opportunities to work on that [04:01] like re-writing the log FILE bit which leaks billions of litres of memory [04:04] maybe during UDS [04:04] lifeless, poolie, will you guys be at UDS? [04:06] ah! you are! [04:07] * beuno starts ploting on how to take advantage of their presence [04:07] Memory being measured in terms of three dimensional volume. Neat. [04:12] AfC: if you squeeze memory enough, you can get liquid out [04:14] silicon-juice? === jamesh_ is now known as jamesh === sdboyer|class is now known as monkeyy === monkeyy is now known as sdboyer [04:59] Hm..what do btree indexes do for Loggerhead? [04:59] they should make it faster === abadger19991 is now known as abadger1999 [05:27] lifeless: do you know if brisbane-core with vila's changes is now passing all tests, or not quite? [05:28] poolie: its either close or all passing [05:28] which isn't really helpful, sorry [05:28] it's approximately helpful :) [05:29] he made all the intertree ones pass; I don't know about the entire test set though === sdboyer is now known as sdboyer|goober [05:30] me either [05:47] No, it's not all passing. [05:53] Um, there were 6 errors and 5 failures total, in test_merge, test_pack_repository and test_repository. === sdboyer|goober is now known as sdboyer [06:05] Peng_: thanks! [06:05] you should fix them :) [06:42] What would be the use of upgrading bzr.dev to the 1.6 format? [06:42] Does doing "bzr branch --stacked" require the source branch to support stacking? === sdboyer is now known as sdboyer|vurk [06:54] Peng_: No, but default stacking policies, which Launchpad uses, require that. [07:00] Oh [07:00] hi alla [07:00] ?? Hi all :) [07:01] ...What are default stacking policies, exactly? "bzr branch" will stack by default? [07:01] poolie: I think it's "quite close" if merged with a recent bzr.dev, but the failing ones are still a bit hard to fix (I can dig to more precise if you will) [07:02] poolie: I think it's "quite close" if merged with a recent bzr.dev, but the failing ones are still a bit hard to fix (I can dig to be more precise if you will) [07:02] Peng_: bzr push will. [07:02] Peng_: tbh, I haven't tried branch. [07:04] Oh. [07:05] ...http://bazaar-vcs.org/bzr/ needs that? [07:08] * spiv finishes for the day [07:12] abentley1: I don't have time to give your remoterepo stacking fix a proper review atm, but at a glance it looks good [07:12] abentley1: if it hasn't already had a review by the time I start tomorrow I'll certainly review it then. [07:12] spiv: Cool === abentley1 is now known as abentley [07:16] hi abentley! [07:16] poolie: hi! [07:21] ciao [07:22] I'm on leave now ;), I may drop by and chat when playing with pet projects, but I'll be focusing on relaxing and zoning out mainly :) [07:32] lifeless: have fun. [07:54] lifeless: I know I'm saying this just after you left, but what are you planning to do about group_size in bzr-search? [08:03] Peng_: right now? nothin [08:04] abentley: tanks [08:05] Heh, okay. [08:29] night all [08:31] Good night. [08:32] Ehhh? [08:33] I think I may have just figured out the cause of my weird Loggerhead issues: a simple typo. That's what I get for not reading the error messages well? [08:35] Oh, no, the typo is new, I think. [08:40] jml: right, one less TUIT; testresources now sorts properly [08:40] Hm...what's the right way to get a patch into Loggerhead? Usually I just bug someone on IRC. :P [08:41] push a branch to lp [08:41] file a merge request in lp [08:41] :) [08:48] man i wish i could just click a button to land that branch [08:48] Hi. :) [08:48] See? Bugging people on IRC is more efficient. :P [08:53] Peng_: merged, ta [08:53] and ooh, the merge proposal is automatically marked merged :) [08:54] Thanks. :) [08:55] Peng_: do you want to join ~loggerhead-team? [08:58] mwhudson: Meaning I could push to the trunk without needing to go through you? [08:58] Or just to have a neat badge? [08:58] the former [08:59] (i don't think ~loggerhead-team has a badge...) [08:59] No badge? That's no fun. :P [08:59] heh [08:59] i guess we'd also want you to commit to being happy to assign (c) Canonical on your work [09:01] I dunno. I don't contribute enough code to be a major review burden. If I ever contributed something large, I'd want it to be reviewed anyway, and my small things aren't really high-priority. [09:09] mwhudson: I wouldn't tie the things together; surely a committment to stay under the paragraph limit when pushing (until a cc assignment is done) is enough [09:09] mwhudson: can always uncommit after all [09:10] lifeless: true === jszakmeister|awa is now known as jszakmeister [09:17] Honestly, I dislike copyright assignments, but I never contribute anything large enough to be worth caring about the rights of anyway. :P [09:17] (That doesn't mean I'd be unwilling to sign one.) [09:19] Then I get to the part where I'm not even an adult, so I don't know if I legally *can* sign one, and then I decide to put it off and go watch TV. :\ [09:22] Peng_: adult is a state of mind [09:24] So when I turn 18, I still won't be an adult? Damn. :( [09:33] i don't _think_ there's an age thing with copyright is there? maybe there is [09:35] I don't think so either, but copyright assignments are legal document thingies, right? so there might be one there. [09:36] hmm === jszakmeister is now known as jszakmeister|awa === andreas__ is now known as ahasenack === weigon__ is now known as weigon === doko_ is now known as doko [12:45] lifeless: FYI, I let a "bzr index" of bzr.dev with a group_size of 50 run to completion. It took 54.5 minutes. :D But it peaked at only about 115 MB of RAM. :) [13:37] lifeless: Also FYI, group_size = 100 took 33:00.01 and used 122 MB of RAM. === sdboyer|vurk is now known as sdboyer|class [13:59] hey... i understand what the push and pull branches that a remembered are, but what is the "submit" branch? [13:59] rocky: For "bzr send". [14:00] so the first time i pull in merge info from a particular source branch, it gets remembered as the "submit" branch? === Mario_ is now known as pygi === mw|out is now known as mw === bac` is now known as bac [15:51] bzrlib/_dirstate_helpers_c.c:7704: warning: dereferencing type-punned pointer will break strict-aliasing rules [15:51] shame on bzrlib [15:51] (one of many, grabbed at random) === kiko-zzz is now known as kiko === ja1 is now known as jam [16:10] lamont: Shame on Pyrex. [16:11] ok [16:11] only 104 occurances of type-punned in the log :-( [16:11] lamont: There are generally fewer warnings with newer versions of Pyrex. [16:12] yeah, that's building 1.9 on hardy [16:12] But I don't remember ever seeing "type-punned". [16:12] google will tell you what gcc really means there. [16:12] "ZOMG this might not do what you want in some strange corner-case" [16:12] or something like that [16:14] it's a clever feature, not a bug (tm) [16:14] jdong: DENIED [16:33] the warning is completely valid [16:33] just compile -fno-strict-aliasing if you don't want it [16:56] Hola a todos, como les va? [16:56] tengo una duda, cuales son los comandos para usar bzr en forma distribuida [16:57] darkness51: ingles? [16:59] morning thumper_laptop [17:03] Kinnison: i am understand english, but i don't write fine [17:04] what command use form distribute work [17:05] darkness51: 'bzr push' can be used to upload your branch to another location. E.g. bzr push sftp://my-fileserver/home/me/projects/mything [17:08] what is the use of bzr pull?? [17:10] The 'pull' command lets you fetch changes from a remote branch into your local one. E.g. Pretend I have a project 'proj' and I work on my laptop and my desktop. I have a server and I keep a branch on there for synchronising. [17:10] I make a change on my desktop and push it to the server with 'bzr push' [17:10] Then on my laptop I can do 'bzr pull' to get the change onto my laptop [17:11] Naturally this requires you to tell it the location of the server first time. After that it will remember for itself. [17:13] ok, thank's Kinnison [17:14] tuna 17:13 ~/scratch % bzr branch lp:~ubuntu-core-dev/livecd-rootfs/trunk/ [17:14] You have not informed bzr of your launchpad login. If you are attempting a [17:14] write operation and it fails, run "bzr launchpad-login YOUR_ID" and try again. [17:14] can't it tell? :-P [17:15] Well, not with the brain of a tuna :p [17:15] My Launchpad login isn't the same as my local login. [17:18] Odd_Bloke, maybe you need to set it in .ssh/config [17:18] Host bazaar.launchpad.net [17:18] User Odd_Bloke [17:18] or whaeva' [17:18] Odd_Bloke: I meant, can't it tell if it's a write operation [17:18] elmo: Ah, I see. [17:18] Not easily, IIRC. :) === mark1 is now known as markh [17:19] not with the current layering [17:30] morning vila [17:30] morning #bzr [17:30] markh: what are you still doing up? :) [17:32] elmo: We can know that "bzr branch" is a readonly operation. I believe the problem is that people often do a readonly operation, and then shortly thereafter do a write operation which fails, but we have lost the info about the LP url they used to get the branch. [17:32] The most common one is: [17:32] "bzr co lp:foo" [17:32] cd foo [17:32] bzr commit -m "this will fail, because of something about HTTP" [17:33] jam: morning jam, evening jam (jetlag making me falling asleep again :-/) [17:34] vila: sorry to hear that. I do believe I had a harder time coming home than I did going to Sydney [17:34] I'm not sure if it is heading East [17:34] or if it is coming home to a baby with an ear-infection [17:35] my guess is the latter [17:35] Odd_Bloke: if you're using bzr 1.9, you don't need to configure ssh. At most, you need to launchpad-login again, but any lp: lookup will do that for you. [17:35] baby's ear-infections are bad for dad's sleep for sure :-/ [17:36] elmo: And no, it can't tell. We don't "open for write" or "open for read". We just open, and after that we're dealing with real URLs. [17:37] jetlag seems better than last week though, but I have a lot of sleep to catch up since I almost *not* sleep from Saturday morning to Monday morning (including the full day in Singapore as an added bonus for delayed flight :) [17:37] * elmo trollingly proposes rich-rich-root which records the original lp: url [17:37] abentley/jam: fair enough :-p [17:39] abentley: then again, that may change in the future. We are looking at having a Branch.open_or_create_locked() which would allow us to know that we will be writing to it. [17:39] elmo: And yes, we've discussed recording the original lp: url rather than its redirection [17:40] vila: ouch on the extra day. I at least managed ~4 hrs of sleep on the plane, and I arrived home at 9pm, which is a great time to reset your internal clock and go straight to bed :) [17:44] jam: well, I did sleep well last night, so hopefully I will too this night and be ok tomorrow :) [17:44] So did you manage to get out of the Airport in Singapore, or was it just 24-hours of airport-chair pain? [17:55] hi folks, I'm getting the following error from config manager - "bzrlib.errors.TagsNotSupported: Tags not supported by BzrBranch5('file:///blah/blah/blah); you may be able to use bzr upgrade." using bzr 1.6.1 - should I "bzr upgrade" that directory? [17:55] mthaddon: It sounds like you have a remote source with a newer format than the local source. [17:56] And the remote source also has tags included. [17:56] (which we are failing to fetch) [17:56] I would recommend upgrading to whatever the remote format has. [17:56] (You can upgrade even further, but the safe bet is to just stay in step) [17:56] jam, how to I verify the remote format version - bzr info? [17:56] I'm not sure about how CM works, but perhaps just deleting the branch will recreate it at the right format. [17:57] bzr info URL should work, IIRC [17:57] thx [17:57] if it's a remote branch, just with http to get the proper format [17:57] bzr+ssh always says nonsense [17:59] You could use nosmart+bzr+ssh to get the real info. [18:05] bzr info bzr+ssh says "Standalone branch (format: unnamed)", bzr info nosmart+bzr+ssh says "Standalone branch (format: pack-0.92)" [18:06] mthaddon: Run "bzr upgrade" in your local branch then. [18:06] jam: well, the airport-chair pain was in Brisbane where they announced 1-hour delay... 7 times :-) [18:07] Peng_, that seems to have fixed it - thx [18:07] vila: what airline? [18:07] and yeah, ouch [18:07] if they told you in advance it was going to be 7 hours, you could have at least done something [18:07] I guess you at least weren't already *on* the plane [18:08] when we finally got to Singapore, the connecting flights were long gone so they transfered us to an hotel and re-book the same flights only 24h later [18:08] There was something about JetBlue in the states doing that [18:08] Quantas from Brisbane [18:08] http://gettingtomaybe.blogspot.com/2007/02/jet-blue-delay.html [18:08] Quote: News broke last week that passengers on Jet Blue flights were subjected to 10 hour delays inside the plane, while on the runway. [18:08] ouch, thats even worse than my regular train delays [18:08] It was a big deal, as they didn't even have food for the passengers, etc. [18:09] Since they are budget flights [18:09] The hotel were nice, but I was so sleep deprived already that I only half enjoyed the sight-seeing in Singapore :-) === mvo__ is now known as mvo [18:10] Quantas did pretty well, giving vouchers in Brisbane for refreshments and so on, the hotel was nice too (even if I ended up in a smoking *level*, ha ha the irony) [18:10] oops, now I did it. my proof for tomorrows algorithm class finally brake apart === mw is now known as mw|food === beuno_ is now known as beuno === fta_ is now known as fta [19:40] jam, What happened to your "This week in Bazaar" posts? [19:44] anyone experience with LP here [19:44] nm, i'll go to #launchpad [19:44] sorry === mw|food is now known as mw === thumper_laptop is now known as thumper [20:21] crap, I just realized I'm halfway through reimplementing Subversion in Python.. [20:24] jelmer: ai [20:24] A painful realization, to be sure. [20:25] 'course, you could always reimplement half of git in python to recover :p [20:25] or you could just fix bugs in bzr and make it fast === kiko is now known as kiko-afk === bac is now known as bac_afk [20:57] lifeless: sweet [20:58] lifeless: does it solve the TSP or what? [20:59] bzr alias upll="pull" [20:59] * LarstiQ has bpvl aliased [21:03] LarstiQ: to what? [21:04] thumper: bzr pull -v | less [21:04] jelmer: because writing libsvn bindings is too painful? [21:05] jelmer: how is the bzr-git work going? === Mario_ is now known as pygi [21:36] thumper: Pretty good; I'm working on two things at the moment - getting more familiar with git's pack format to be able to implement that for bzr-git and factoring out code that is shared by bzr-svn and bzr-git [21:40] jelmer: I found git's pack format to be pretty clear when I looked, how are you finding it? [21:44] james_w, Yeah, it's quite simple [21:46] btw, I was thinking of renaming the python-git code you originally wrote to "Dulwich" [21:46] (after the Monty Python sketch about a cocktail party there featuring a Mr and Mrs Git) [21:46] heh :-) [21:50] jelmer: have you talked with Jc2k about git and python? [21:50] jelmer: I think he did something... [21:50] jelmer: or perhaps he was working in C, I'm not entirely sure [21:50] thumper, Yeah, we talked about it a bit during the summer [21:50] * thumper nods [21:50] thumper, That was indeed in C - based on the upstream git code [21:51] ah ok [21:52] Unable to obtain lock chroot-14417232:///srv/bzr/ ... If you're sure that it's not being modified, use bzr break-lock chroot-14417232:///srv/bzr/.... $ bzr break-lock chroot-14417232:///srv/bzr/ ... bzr: ERROR: Unsupported protocol for url "chroot-14417232:///srv/bzr/... that's not terribly useful [21:52] how can I get it to break this lock? [21:52] eydaimon, I think the error message may be a bit clearer in newer versions of bzr [21:52] eydaimon, generally, you should be able to run "bzr break-lock" on the URL that was giving you problems [21:53] jelmer: Really? I thought it was still leaking underlying paths in the latest versions. [21:53] jelmer: but it says unsupported protocol for url [21:53] how do I deal with that? [21:53] eydaimon, You should be able to run "bzr break-lock" on the URL you were trying to pull from/push to/branch from/etc [21:54] ok [21:54] eydaimon: You ignore the URL it specifies. Use the normal URL for the branch. [21:54] abentley, Sorry, I must be mistaken then [21:54] ISTR some patch for this going in [21:55] jelmer: thanks, that fixed it [21:56] eydaimon: You're welcome. Hopefully the URL in that error message will get fixed at some point. [21:56] that's probably helpful :) === bac_afk is now known as bac [22:11] hello [22:11] I have commit A -> B locally, then pushed to the remote repository [22:12] I then uncommitted B and committed C [22:12] when I push it says of course that branches diverge [22:12] how can I force to push A -> C? [22:12] Lethalman, push --overwrite [22:12] push --overwrite [22:12] if you're sure [22:12] ah [22:12] yes i'm sure === beuno_ is now known as beuno [22:13] or merge from then push to [22:13] thanks very much [22:34] abentley, can you give me a rundown of the proper way to push a loom to launchpad? [22:39] rockstar: the loom must be in format 6. Then pushing it should work, but won't stack. [22:41] abentley, I apparently forgot to record before I pushed. So I deleted the pushed branch, recorded, and pushed again. [22:41] But when I pull from another system, the loom appears to only have one thread. [22:42] rockstar: I'm the wrong guy to answer this, because I never push looms on purpose. [22:42] abentley, :) [23:07] jelmer: do I need to set some things up to run bzr-svn's unit tests? [23:07] I'm getting failures across the board [23:07] jfroy|work, no, just run "bzr selftest --starting-with=bzrlib.plugins.svn" [23:12] jelmer: eh, it didn't even make it to the end :p [23:12] OSError, "too many files open" [23:13] I've never run the testsuite on Mac OS X, so I'm not sure how well it works there [23:13] Apparently it massively fails. [23:13] Hum, any known issues with Python 2.6? [23:14] oh, yeah, python2.6 still has issues [23:14] sqlite3 has slightly different behaviour, for example [23:14] right [23:14] I'll try to run the tests on 2.5 [23:15] I'll be back in ~45 minutes === sdboyer|class is now known as sdboyer [23:24] jam: you here and got a sec? === jamesh_ is now known as jamesh [23:45] jelmer: sent you a merge bundle as the first parter of a workaround for my situation [23:58] is it possible to use local svn repo with bzr-svn