=== jamesh__ is now known as jamesh [03:59] hello, I did: bzr tag -r 77 datahub-release-0.7 now how do I push this to launchpad? [03:59] Do I have to recommit and push? "No new revisions to push." [04:02] lszyba1: no, the push worked [04:03] lszyba1: the message is confusing (there's a bug open about that), but the tag will have been pushed. [04:04] lszyba1: $ bzr tags -d lp:datahub [04:04] datahub-release-0.7 77 [04:05] lszyba1: so you can see the tag is there :) [04:06] (bug 164450 appears the bug number) [04:06] Launchpad bug 164450 in bzr "should show a message when pushing(overwriting?) tags" [Low,Confirmed] https://launchpad.net/bugs/164450 [04:16] thanks, is there a command to list tags bzr lstags ?? [04:16] just 'bzr tags' [04:16] bzr tags ..found it [04:17] thanks,,updated my manual..http://lucasmanual.com/mywiki/Bazaar [04:20] anybody knows if launchpad shows tags anywhere? [04:24] Not that I know of. [04:32] jelmer: ping [04:33] jfroy|work, pong [04:33] jelmer: OK, so I thought I understood, but bzr-svn (or me) just did something I'm not sure I understand. [04:34] Basically, what are you supposed to do again when bzr pull from the subversion branch tells you the branches have diverged? I did a merge, commit, push operation. [04:35] And although that mostly worked, bzr-svn committed, as far as I can tell, the same set of changes that someone else had committed via svn (which caused the branches to diverge in the first place). [04:35] Is that expected, or did I do something wrong? [04:35] jfroy|work, yes - see the bzr-svn FAQ [04:36] you may want to use bzr rebase rather than merge [04:36] Ah right, I remember that... [04:37] Hum, I see, I forgot I wasn't working with a Subversion checkout, but rather a branch [04:37] (since committing on a checkout is broken in 1.9) [04:37] A checkout would not have normally allowed by commit which caused the branch to diverge. [04:37] * jfroy|work is unconfused [04:37] *allowed my [04:38] The amazing thing is, no conflicts were generated. [04:38] It's as if the same set of changes were applied twice without a problem. [04:39] yes, that's correct - the revision that was moved to the right hand side is no longer on mainline [04:40] indeed, it got moved to the right-side of the merge commit. [05:47] uhoh. "Device not ready" while trying yo read the kernel is no nice greeting in the morning [06:00] spiv, still around? [06:01] poolie: yeah [06:03] i'm going to finish up in a bit, about 5:30 or 6, and go out [06:03] jelmer: if I have the latest stable branch of bzr-svn installed I can't branch from http://bzr.arbash-meinel.com/branches/bzr/1.10-dev/revision_stream [06:04] mwh reported a new bug with those checks i put in for stacked branches [06:05] i'm just thinking about whether to get on to that now or keep going with john's merges [06:05] probably the second is better, as it's already work in progress [06:06] and then i can take an uinterrupted go at 302968 [06:06] how did you go? [06:06] 302968? [06:06] correction bug 302698 [06:06] Launchpad bug 302698 in bzr ""BzrCheckError: Internal check failed: Newly created pack file has delta references to items not in its repository:" on pushing a stacked branch" [High,Confirmed] https://launchpad.net/bugs/302698 [06:07] I saw mwh's bug, it seems to still be noticing the problem a bit too late. [06:07] yeah [06:07] Sooner than the next push, at least :) [06:07] right [06:07] i'm happy with the check there [06:07] and it should be fairly possible to just work backwards and see what should be done [06:08] It's weird that the autopack's _check_references trips over, but the push itself didn't. [06:08] Oh, unless an earlier pack that's part of the autopack has the dud record(s). [06:09] If that's the case then it might just be a matter of waiting until a report of that error comes in that isn't from an autopack. [06:10] I agree that at the moment doing work that will make the rc1 tomorrow better is probably the best bet. [06:10] indeed, but what would that be? :) [06:12] Heh. [06:12] i'll look at your tweak for 'avoid extracting revision texts' and merge it [06:12] Landing already reviewed/proposed patches. [06:12] the patch makes sense to me [06:12] I'm doing that one atm, actually :) [06:12] oh ok [06:12] Just waiting for my push to finish then I'll hit pqm-submit (so only a few seconds) [06:13] then i'll look at 'retry 2/3', as no one has reviewed that yet [06:14] That'd be good. I took a glance at it but the diff seemed a bit odd, and merging bzr.dev gave some irritating criss-cross conflicts. [06:15] ("odd" as in "looks like it shows unrelated changes that are already in bzr.dev") [06:22] i might try that merge then [06:24] poolie: oh, and Mary would be happy if http://bundlebuggy.aaronbentley.com/project/bzr/request/%3C418c22640811181423hc9a2628y4b0bae4dd1f6e7fd%40mail.gmail.com%3E makes it into 1.10 [06:26] poolie: there's a newer patch that makes _get_nick into a public get_nick as suggested by the review, but that newer patch is stalled while waiting for tests to be written for that new public API. [06:26] poolie: so probably we should just merge the initial patch (with _get_nick) for 1.10 [06:27] +1 from me [06:27] poolie: I can do that if that sounds -- great :) [07:02] hi all [07:04] hey vila [07:34] ok, i've fixed up jam's retry patch and sent it to pqm, and now i'm signing off [07:34] pfeww, finally the fix for bug #277537 is ready to be used or should I run yet another mammoth 34h bzr check ? :-/ :-) :-\ [07:34] Launchpad bug 277537 in bzr "annotate says revision modified file, "bzr diff -c" widly contradicts" [High,Fix committed] https://launchpad.net/bugs/277537 [07:36] poolie: cool. I've got jam's pack re-order patch with PQM too, so I think we've merged all the important stuff he wanted to have merged. [07:37] * spiv goes for a swim [08:44] folks, bzr+ssh doesnt work with the setup installer for windows. So I decided to install python and paramiko and then bzr. Paramiko now works on its own. I try to install with the "simple" installer, but now bzr is not in the system path. And I can't find the exe. Where is it installed? === jszakmeister|awa is now known as jszakmeister [09:32] how is nested trees developing going? [09:47] <_mathrick> hey [09:48] <_mathrick> what is the best repo structure to have a dev branch, and another branch which is dev + local conf (DB/host settings in this case)? [09:49] <_mathrick> ideally, all changes would be pushed into dev unless I specifically say not to [09:49] <_mathrick> ie. I'd like to avoid manual sync as much as possible [10:07] poolie, spiv: hi. do you need any help getting the branches that are still causing a problem for online services? [10:07] SteveA: more details certainly wouldn't hurt [10:08] SteveA: is it easy to make a tarball of the offending branch(es) + plus instructions to reproduce the problem with the tarball? [10:13] <_mathrick> um, how can I export a shelf? [10:14] spiv: I think we only see it when talking to Launchpad [10:14] spiv: best thing is if I give you access to the private branches, I think [10:16] spiv: I just added you as a member of a team. You should be able to get the branches now. [11:05] jelmer: hey just discussing how to make --export-upstream work better for someone, and I think we may be able to solve your problem at the same time === _[PUPPETS]Gonzo is now known as [PUPPETS]Gonzo [11:29] Hey there [11:30] Can I get bzr branch (or bzr checkout) to not create a working directory but only to fetch the revisions? [11:31] I'm trying to setup a "mirror" repository on my laptop with branches bound to those on my server's repository. [11:31] hey Lo-lan-do [11:31] you can do it if you create a --no-trees repository [11:31] But even though the local repository is --no-trees, bzr branch creates working copies :-/ [11:32] oh [11:32] that's odd [11:35] (This is bzr 1.5-1.1 on Debian unstable, by the way) [11:35] But anyway, I'll do some shell magic with bzr reconfigure. [11:36] I'm pretty sure --no-trees should work [11:36] you might have found a bug [11:37] Lo-lan-do: thanks for working on loggerhead for alioth [11:37] You're welcome. You'll probably notice it doesn't Just Work quite yet :-) [11:39] (I reported the bugs last night) [11:40] I'm off to lunch, but I'll be back to discuss them afterwards :-) === jszakmeister is now known as jszakmeister|awa [12:12] james_w, ah, cool - how? [12:12] jelmer: do you use export-upstream in merge mode, or not? [12:13] james_w, no, I don't use merge mode anymore these days [12:13] jelmer: to add a new mode, so one can just re-use the tarball if already present [12:15] we could also make it a but easier as well, as when you are not in merge mode the upstream branch is not necessary to retrieve the revision in question, as you will have merged it [12:15] though noting the upstream branch will still be useful [12:15] I haven't thought it all through yet though [12:17] anyone know why on ubuntu hardy, bzr depends on python-central (>= 0.6.7) but only 0.6.5ubuntu1 is available [12:19] james_w, Is there a good use case for "export-upstream" without "export-upstream-revision" ? [12:20] jelmer: it was intended for nightly build [13:26] Three lines: $ bzr upgrade \nFormat for file:///blah/.bzr/ is deprecated - please use 'bzr upgrade' to get better performance \nbzr: ERROR: The branch format Bazaar-NG meta directory, format 1 is already at the most recent format.\n [13:26] "please change! sorry, nothing to do!" Why? [13:27] CardinalFang, maybe the repo is old format? [13:27] are you using shared repos? [13:27] Yes, I am. [13:28] CardinalFang, then you have to upgrade bith the branch and the repo [13:28] Hrm. Ugly. Okay. [14:32] hi [14:32] is there a command to remvoe unknow files? [14:32] from the repo? [14:32] *remove [14:33] goshawk: `bzr clean-tree`, though that may be in bzrtools. [14:33] tahnks [14:33] thanks [14:33] works great! [14:33] :) [14:34] wait... it deleted the whole repo... [14:38] goshawk: That's surprising. Are you sure that you had added the files you were expecting to keep? [14:38] <_mathrick> goshawk: repo or tree? [14:38] wait [14:39] <_mathrick> also, bzr serve just tried to murder my RAM, why does it hate my RAM? [14:40] <_mathrick> it used up above 500M or so, unsure how much exactly, because I was too busy killing it before it took all the mem [14:41] <_mathrick> the repo has only 2 revisions, although it has quite a few files [14:41] <_mathrick> 10K files, ~150MB working tree [14:42] <_mathrick> how do I find out what's up? [14:51] Odd_Bloke: it was my fault [14:58] how can i revert before deleting? [15:00] done :) === rocky1 is now known as rocky === beaumonta is now known as abeaumont [17:08] hi guys. About to try about bzr for the first time === _mathrick is now known as mathrick [17:10] does bzr support keywords at all? I namely want to use the $id $ feature of svn if possible [17:13] CaMason_: not currently, it is in development though [17:13] james_w: ok. It's not that important for me at the moment :) [17:14] I love the fact this is portable. I can now continue my dev and commits whilst on the train :D [17:14] I did like tortoiseSVN though. I'm on vista x64, so I can't use the gui [17:14] although the eclipse plugin seems good [17:14] I thought there was a TortoiseBZR too? [17:15] yes but it doesn't support x64, I believe [17:16] Ah. Pardon me for being ignorant of these things, I haven't touched Windows for years. [17:17] there's only a couple of things holding me back in windows, and thats the Adobe suite, and better multi-monitor support [17:18] I need access to the Adobe stuff, and there's no elegant working solutions for the suite. Also, I use 3 or 4 monitors, and the support with distro's like ubuntu isn't as good as the windows implementation, yet [17:19] I wasn't tryng to elicit a discussion on that subject, just apologising, really ;-) [17:19] Lo-lan-do: I was just voicing my dissapointment with Adobe's lack of linux builds, as usual :) [17:21] hrm.. I've got a couple of symbollic links in my project. I just did a bzr add, and commit. Now 'bzr status' shows those 2 links as 'unknown' with @ appended [17:21] Any ideas on the best approach now? [17:26] CaMason_: what happens if you "bzr add "? [17:27] it seems it added the folder, but not contents (desired) [17:27] and commit succeeded :) [17:37] What's the preferred mechanism for seperating lines/sentences with a commit message? [17:41] CaMason_: I like to do a one line summary, a blank line, then paragraphs explaining it [17:41] sorry for my lack of understanding. How would I insert a newline when I'm typing at command line? (linux) [17:41] that works well with things like bzr log --line, loggerhead etc. [17:42] oh [17:42] don't pass -m [17:42] and don't type a message [17:42] it is possible at the command line, but if you do that you will be given an editor in which to type [17:43] oh I see! and thanks to uncommit, I can go back and try it on a real commit :D [17:43] liking bzr already [18:20] james_w: congrats on the Ubuntu package branches [18:57] hi, anyone know the status of bzrp4? Is the bzr -> perforce functionality available to test/hack? [19:22] thanks LaserJock [19:36] james_w: I was looking at a failure and see: UnicodeDecodeError: 'utf8' codec can't decode bytes in position 286-291: unsupported Unicode code range [19:36] james_w: isn't that a common problem? [19:37] yeah [19:37] it assumes debian/changelog is valid utf-8 [19:37] I'm going to try and remove that assumption or relax it somehow [19:38] Is there anything like SNV's $Id$ (I think that's how it is in SVN at least) in bazaar? [19:39] Dvyjones: not currently, but it's in development [19:50] And there's nothing like CIA(.vc) in Bazaar? [19:50] Dvyjones, sure, check out the CIA plugin: http://bazaar-vcs.org [19:50] er [19:50] http://bazaar-vcs.org/BzrPlugins [19:59] james_w: this one seems a bit weird: bzrlib.errors.NoSuchTag: No such tag: upstream-ubuntu-0.12 [20:00] Do oyu know any good PHP software management softares / CMSes that support showing last bazaar pushes? [20:23] LaserJock: yeah, that's the most common one, it's a logic error I haven't had chance to look in to yet [20:24] james_w: something about the way it figures out what the upstream release is I suppose === tetha_ is now known as tetha [21:46] hello [21:46] hello poolie [21:55] poolie: hi [21:55] poolie: I was trying to remind myself of the --fixes syntax, and went `bzr help bugs` [21:56] poolie: I was somewhat surprised not to see a launchpad example in there [21:56] poolie: it had three or four other examples, but no lp :( [21:56] that sounds like a bug [21:56] i wonder why [21:57] ah so it does talk about lp, but doesn't give a good example [22:15] Hi guys. I've got a shared-repo with a project in it that I no don't need. I want to delete it. Should I just rm -rf it, or use bzr commands to do so? It has never been branched, just created is all. [22:15] if it has nothing else in it that you want, you can just rm it [22:16] Ok. [22:17] If I were to delete and recreate that project, as far as the shared repo is concened, it is not the same as the old one? [22:19] right === fta_ is now known as fta [22:20] thanks poolie. bye. [22:25] poolie: huh! [22:25] mm? [22:25] poolie: that change used to pass tests, but now it trips on your new consistency chekc [22:26] ah [22:26] (the change to remove "len(self.target._fallback_repositories) > 0") [22:27] yes, that's what i understood you to mean [22:27] good thing that check was there i guess [22:27] I guess that makes sense, it can't be a straight pack-to-pack copy, some delta records have to be fulltexts. [22:40] thumper, spiv: just or anything more? [22:40] poolie: fixes rather than addresses perhaps? [22:40] but yes, that is all I was looking for [22:41] i'd kind of like to say what effect this will [22:41] have [22:42] but, that's mostly defined server side and may go out of date [22:43] and i'm not sure off hand what will happen [22:45] poolie: that looks reasonable to me [22:46] thumper: i just got subscribed to a very active project using code reviews etc [22:46] so, congratulations on your feature's popularity [22:46] Off hand I believe --fixes lp:NNNN causes that branch to be associated with the bug, and that association's status is set to "Fix Available". I think "addresses" is an ok way to summarise that ;) [22:46] but it's also an eyewatering amount of email [22:46] and i don't really really care about it [22:46] possibly i should fix this in my mail client [22:47] but i wonder, if i unsubscribed (if i can) whether I'd get a good replacement through the web [22:47] basically a timeline view of all merge proposals, comments, and landings [22:55] poolie: so I've updated my original InterPackRepo.fetch hack for bug 294479 to add a comment and a NEWS entry. [22:55] Launchpad bug 294479 in bzr "Vast number of round-trips pushing stacked branch" [High,In progress] https://launchpad.net/bugs/294479 [22:55] poolie: I'm just running a full test suite now, in case there's something unexpected, but I don't expect there to be. [23:03] so this is setting it through a callback? [23:03] i think a comment about why we're doing it, and maybe also something on the constructor parameter to say we don't really want it would make sense [23:03] i sent that doc patch to pqm [23:04] poolie: right. [23:04] poolie: I'll add a comment to the constructor too. [23:29] you know what would be great? [23:29] being able to re-order threads in a loom [23:30] (I guess doing it properly depends on proper cherrypicking, but I'd be happy with a half-arsed solution that was a joy to use) [23:46] hey guys, i've just installed bzr1.9 on windows, and i'm getting the error "Unable to load bzr-svn extensions - did you build it?" "cannot import name client" [23:46] anyone know what this actually means? [23:47] jml: its in todo I think, file a bug, yada yada yada [23:51] Glenjamin: It's a packaging bug [23:51] Glenjamin: You installed the python2.5 version? [23:51] yes [23:51] Glenjamin: Delete c:\python25\lib\site-packages\bzrlib\plugins\svn [23:51] Unless you actually want to use bzr-svn, of course [23:52] and if i want svn i should just install it? [23:52] "just install it" is relative on windows [23:52] It's rather tricky to get it built [23:52] ah [23:52] i probably dont need it for now [23:52] For windows I think you are going to be looking at prepackaged binaries [23:52] yeah, i'm fine with that [23:53] while i'm here, i dont suppose anyone knows how to get tortoisebzr working on 64bit windows? [23:58] poolie: http://bundlebuggy.aaronbentley.com/project/bzr/request/%3C20081127231800.GH23611%40steerpike.home.puzzling.org%3E is the updated patch, btw