[00:21] Jc2k: thanks a lot for your help, you put me on the right track. turns out "use of a system shared libexpat.so/expat.dll is not advised" in python. No wonder I couldn't get the tests to pass. :p === Spaz is now known as _ === _ is now known as Spaz [00:37] :p [00:37] np fdv === fta_ is now known as fta === jszakmeister is now known as jszakmeister|awa === mark1 is now known as markh === snaggen1 is now known as Snaggen === Mario__ is now known as pygi === asac_ is now known as asac [11:35] hey, does anyone else have a problem with directory tab completion on windows with 1.10? [11:36] (as in, it doesnt happen) [11:37] Glenjamin: what do you mean exactly? tab completion is generally handled by the "shell" (ie, the command-prompt) [11:37] oh, in bzr shell [11:37] should have mentioned that [11:38] I've never used it (and don't even know *what* it is :) - did it ever work, or this is just something you first tried now? [11:39] its always worked fine, but now the tab completion only seems to do commands [11:39] either way though, I'm afraid I'm not going to have a clue [11:39] fair enough [11:39] if it is a regression, it is certainly worth reporting in as much detail as you can [11:39] (ie, to launchpad) [11:40] mm, i thought i'd better try and see if its affecting anyone else, since i'm not using the proper win32 release [11:40] as i built my own while the build machine was broken [16:17] hello. is it possible to change commiter of a revision? i forgot to set my email and a few commits are attributed to a wrong user [16:34] I added a local copy of a huge library to my branch and committed it, and now I've changed my mind. 'bzr revert' seems only to change my working tree; how can I revert the entire branch so it's as if those files were never added and the branch committed? [16:34] You have to back the branch up to before all that happened. [16:35] Generally, you do that with either pull or uncommit. [16:35] Is this really recent (in branch history terms)? [16:35] ah! uncommit - excellent. I hadn't yet found that in my meanderings through the help. Thx. [16:36] fullermd: yes. This will be no problem. [16:36] thx again. [16:36] Note that 'uncommit' takes you back to the state right before you committed. [16:37] Which means that the library is still "add"'d. So if you commit again right away, you'll be right back where you started ;) [16:37] right, so the files will still be there. [16:37] * moquist nods [16:37] You could revert the tree, though that would lose any other changes you had since the last commit. Or use 'rm --keep' to un-add the library. [16:38] I'll uncommit to one rev before what I want, then I can recommit that and manually merge in the tiny number of changes I've made to files I actually want to keep. [16:38] Another option is to make a new branch, from the rev before the one you want. [16:39] Then manually cherrypick the changes from the 'original' branch over into the new one, and finally just mv them around so you move onward in the 'new' branch. [16:39] That way you can keep the old one around untouched, just in case. [16:39] That seems like the best idea. [16:40] beautiful. I'm almost done. [16:49] what's the simplest workflow case of having non-committed changes in one branch and then wanting to move those changes over into another new branch without committing them to the current one and then removing those changes from the currnet branch? [16:49] i would have thought bzr shelve in the first branch and the bzr unshelve in the second branch but that doesn't seem to work [16:50] bzr 1.10 [16:50] rocky: "bzr merge --uncommitted"? [16:51] Peng_: perfect, thanks :) [16:52] :) [18:05] hi, is bzr-svn compatible with bzr 1.10 ? [18:10] ok, I'm officially stupid, nevermind [18:12] my second question - do you know of any tool which can help me find the latest common ancestor of two branches when they are not linked from bzr's point of view? [18:13] epsy: you're using a different definition of common ancestor than bzr in that case? [18:16] well, the last revision that's common in two branches [18:19] epsy: if needed once, I'd figure that out from `bzr missing` output [18:21] neat..can I check if a branch has diverged from an official branch using bzr missing? or is another command better-suited for this task, considering I would use it in a shell script? [18:34] epsy: bzr missing is a good choice for detecting divergence [18:35] only if tree is not a checkout, right? [18:38] epsy: what do you mean? missing operates on branches, not trees [18:39] well, actually, I will just need a way to dertermine the other branch properly [19:31] Silly question no doubt, but I'm coming from SVN... Does Bazaar support access control? I want to push a branch up to a server, but I want only people with access to get to it. [19:35] derekmounce_, access control is based on who can read/write [19:35] so you'd setup ssh accounts or ftp [19:35] beuno: Ok, cool, so it's not "built in" to bzr. That makes sense... [19:36] Thanks for the answer. :) [19:36] right, although there is a script in controb/ that helps you do some of that with ssh keys [20:05] Another question... I've pushed via ftp, but I'd also like to have a working tree up in FTP. Is it possible to force the push to also create a working tree? === _thumper_ is now known as thumper === Mario__ is now known as pygi [21:44] is there any way to have --using meld open the entire diff instead of files by file? [22:04] Hi! I've got a number of random error messages from running the selftest suite. Where should I send them, for best results? [22:34] beuno, around? [23:06] Does the bzr support in meld work for anyone? [23:24] rockstar, hi [23:25] beuno, have you seen the loggerhead branch for better linking of files? [23:26] rockstar, I don't think so. Which one? [23:26] lp:~ree/loggerhead/abstract-paths-fix [23:29] looks interesting [23:29] Basically, I wondered if there was a policy on XXX statements in Loggerhead, but basically, I don't think it needs to be there at all. [23:30] yeah, it's kinda odd [23:30] I do add FIXME's now and then, which I guess is the same thing [23:32] the comment is useful though [23:32] branch looks good otherwise [23:32] are you reviewing it? :) [23:35] beuno, yup. [23:35] rockstar, awesome [23:36] I have a branch 80% ready that uses YUI to get diffs through ajax [23:36] instead of generating them every time === Mario__ is now known as pygi [23:37] What does this change for users with JS disabled? [23:38] Peng_, the URL will take you to a diff [23:38] so you loose seeing all diffs in the same page I suppose [23:38] is that terrible? [23:39] beuno, aren't we using javascript for that as it is? [23:39] rockstar, yeah, for expanding and contracting [23:39] beuno: Ehh, it's not great. What about taking you to an ?diffs=all#somefile page or something? [23:40] Peng_, I guess I could... [23:40] Peng_, it's not great, no, but for those who don't want javascript, they are already experiencing a less than great internets. [23:41] rockstar: Well yeah, but.. [23:41] Peng_, especially since you're our testsuite ;) [23:41] beuno, We _can_ remedy that. [23:42] Whenever I use a diff page in Trac or something, I like to see all the files at once, but I don't do it frequently. [23:42] I more use them to see "what happened there?" than to actually debug stuff or anything like that.. [23:42] Or maybe that is the normal usage. After all, if you wanted to do something more in-depth, that's what the VCS is for. [23:43] rockstar, yeah, we do need to start writing tests... [23:43] Peng_, I'll have the option to see all the diffs for a revision [23:43] it's not too hard [23:43] Oh, that sounds good. [23:44] For both Ajax and non-Ajax? Or would that not be very useful? [23:44] beuno, I say non-Ajax tests. [23:45] Peng_, I think that the link will be there when avascript is disabled [23:45] rockstar, yeah, so did I [23:46] Wait, what page is this for? /revision? [23:48] Peng_, yup [23:49] So it'll default to having every file collapsed? [23:49] that's my idea, yes [23:50] Will you keep the "expand all/collapse all" button? [23:50] yeap [23:50] it will bring in all the diffs [23:51] Cool. :) [23:51] Possible suggestion: If only one file was changed, expand it by default? [23:51] that would be interesting, yes [23:52] although maybe confusing [23:52] sometimes ahving diffs, sometimes not [23:53] True. [23:55] Peng_, I'll poke you before merging into trunk [23:55] 'kay [23:58] I also experimented with adding syntax highlighting, but it's a PITA to do with the way we currently do things [23:59] Oh, too bad.