[02:36] Hey all. Just learned of bazaar and find it to be a very vague concept in my mind and Ive done some looking for information on wiki and the bar webpage, and its not helping. I was wondering if anyone had a link to some information that might help explain how bzr does what it does and such. Thanks! [02:46] stopie: have you used other SCMs before, like subversion or cvs? === UdontKnow is now known as ObiWanQueNoob [07:10] hi. can someone point me to a document which describes why version control is required? I need to convince a team that its needed.. would be nice to have all the arguments in one place [07:13] Theres a team that doesn't think its needed? :S [07:13] http://ianclatworthy.files.wordpress.com/2007/10/dvcs-why-and-how3.pdf <-- heres a good resource [07:32] i really dont know what to do other than bang my head against a wall [07:32] devilsadvocate: are you the team lead? [07:32] they want to do file renaming on ftp, etc [07:33] not the lead, but its my responsibility to make sure different parts of the system work together, etc [07:33] devilsadvocate: well, if folk want to do it hard, and the lead agrees, just use bzr yourself [07:33] and i also am the one with the most experience in writng actual code, so i have a feeling i'll be doing a chunk of that work as well [07:33] and show it off when ever a suitable time happens [07:34] it will spread ;) [07:34] the suitable time has already happened [07:34] we ended up submitting a report that had different sections at different levels of recent, since it turned out people were editing different docs [07:35] and the argument that is used against it is that "we wont be able to make people wont use it" [07:36] cody-somerville, thanks for the link. it might just be enough to tip it over the edge (added to my threat of leaving the project) [09:52] garyvdm: ping [09:52] Hi Alex! [09:53] bialix: pong [09:53] hi Gary [09:53] can you teach me about qlog internals a bit? [09:54] Sure - Thats why I did that review - and did not just fix [09:54] thank you [09:54] in log.py I see method update_search [09:54] there is special code paths for revid and revno [09:54] do you want me to place tag filter there? [09:55] We could. [09:55] It depends on how you want it to behave [09:56] for revid and revno - we don't filter - we just jump to the revision - cause there will only ever be 1 [09:56] my intent to provide the search for specif tags [09:57] For tag - your search might [09:57] brb [09:57] sorry - I'll be back in 10 min [09:57] btw, search for tags has one weak place: if tag in merged revision then it still collapsed. how can I force it to be visible? [09:57] ok [10:05] Dumb q: what's the best way of showing a change, ie. log + diff? [10:05] log -p [10:06] bialix: not in 1.6.1? [10:06] rusty: nope [10:07] upgrade? [10:08] garyvdm: perhaps I understand your suggestion: I need to use self.sr_index_matched_revids dict, and fill it with matching keys. Is that right? [10:08] bialix: thx, but sticking with ubuntu default pkgs. [10:17] bialix: Yes- if you want to filter. [10:17] btw, there is something strange [10:18] For revno, revid - we select, and not filter. For message, author, we filter [10:18] you're using self.sr_filter_re and self.sr_filter_regex in the same time [10:18] in constructor you define self.sr_filter_re [10:18] Oh [10:18] it seems like a typo or bug [10:19] self.sr_filter_re is the one that gets used. I'll fix that though. [10:21] I have working filter for tags: http://pastebin.com/m4a3a752c [10:21] garyvdm: how can I force merged revision to be visible? [10:21] Ah [10:22] um? [10:23] http://pastebin.com/d2cb3b507 [10:24] You need to set self.branch_lines[branch_id][1] = True [10:24] For the revisions, and the revisions that merge them [10:24] recursively [10:25] oh [10:25] it will play badly with multiple branches? [10:26] No [10:26] See invaladate_filter_cache_revs to see how I find the revisions that merge them. [10:27] back in 1min [10:28] * bialix -> lunch, back in 20 min [10:35] bialix: Please don't use sr_index_matched_revids. Create a new field, say sr_tag_matched_revids [10:37] why? [10:43] bialix: To make it easier to understand / read. [10:44] It will work just as well if you don't. [10:47] will do [10:47] I'll send the patch for review [10:49] how would i check this repository out ? https://launchpad.net/alarmclock [10:50] typically you can just do bzr branch lp:alarmclock [10:52] cool it worked ;) [10:52] but with a dash in it [10:52] alarm-clock [10:53] i'm starting to like bzr alot. the website looks/works like a few million dollars === mario_ is now known as pygi === thunderstruck is now known as gnomefreak [13:13] hi, I am a newbie of bzr, could some tell me how to make a patch [13:13] which command should I use [13:21] xiaohui: 'bzr diff' will show you what uncommited changes you have. [13:22] You can redirect it to a new file with: bzr diff > my_patch1.diff [13:25] oh , thank you jpds [13:26] or better, commit them and use: bzr send -o mypath.diff [13:27] luks: Oh, did not know that, thanks. [13:27] this is dvcs, offline commits are one of it's main point :) [13:28] and if the upstream is using bzr, they can just use bzr merge/pull mypatch.diff [13:37] jpds: I have done as your said , but I havn't seen the patch show the info about the added file [13:37] welcome to patches [13:37] what's the patch for? [13:38] xiaohui: did you 'bzr add' the file? [13:38] and seriously, use bzr send [13:38] oh luks , I haven't do that [13:39] I just used the bzr from this moring [13:40] the usul way of contributing patches to a project using bzr is to branch the remote branch, work on the branch, make a commit, bzr send the resulting patch [13:40] the patch produced by bzr send will then contain your name, the commit message, etc. [13:41] and the maintainer will just merge it like if it was a normal branch [13:41] *ussual [13:41] bah, I can't type [13:43] it works, thank you guys [13:43] I will read the document later [13:56] How do I change the default push location for a branch? [13:57] edit .bzr/branch/branch.conf or bzr push --remember somethingelse [13:57] Is the --remember switch documented? [13:59] bzr help push === thunderstruck is now known as gnomefreak === ObiWanQueNoob is now known as UdontKnow [15:48] hi, is there a gui frontend for bzr on os-x? I want to use bzr on a project, but I need to provide a nice gui for my friend on os-x or he will refuse :) [16:08] Is there some way to configure bzr so that if I do « bzr branch foo », then it remembers the parent branch as the push location so that I can do « bzr push » without having to specify where I want to push? In short, I want to be able to do: « bzr checkout foo; *hack*; bzr push », instead of: «bzr checkout foo; *hack*; bzr push foo » [16:08] erm. s/checkout/branch [16:09] no [16:10] Is there a reason other than "no one has contributed the code" why that is? Or at least why a configuration setting does not exist? [16:32] jrib, probably to prevent accidental pushing [16:32] jrib, however, I'm pretty sure checkout does remember. Infact, with a checkout, when you commit, you're committing to the remote branch [16:32] Whereas branch is actually creating a new branch that you commit to [16:35] jrib: bzr push :parent [16:46] thats a neat trick [16:53] jpds: that will do, thanks [16:53] 'tis a nifty shorthand, true. :) [17:03] is there a gui frontend for bzr in os-x? [17:10] I think there is good bzr integration with netbeans [17:10] jrib: Looks like QBzr runs on it: http://bazaar-vcs.org/MacOSXBundle [17:11] thanks again === goden is now known as eugeneoden [18:02] hi, I added a .bzrignore file in my project directory and typed on the terminal "bzr ignore ./config.cfg" [18:02] but I got this: "Warning: the following files are version controlled and match your ignore pattern: config.cfg" [18:02] if I write bzr ignored, config.cfg is not showed. what should I do to ignore it? [18:04] BlackLukes: euh [18:05] BlackLukes: the warning tells you that the file is under version control [18:06] BlackLukes: in which case, even if it matches an ignore rule, it will not be ignored. [18:06] I have to keep it under version control but it should ignore changes [18:06] that doesn't make a lot of sense [18:06] BlackLukes: if you're not interested in versioning it, why keep it under version control? [18:07] no it's not that case [18:08] I have a configuration file that shouldn't change if a developer commit a changes to other files that automatically changes the configuration file (it's system dependent) [18:10] BlackLukes: so, why does it need to be versioned in the first place then? [18:11] I have to keep the default value every time the application run.. if a developer commit a change, the conf file is changed so I have to manually change the conf file to default values [18:12] BlackLukes: ignores do not work the way you seem to want to use them, and I don't completely understand your situation so have trouble recommending a solution [18:13] yeah you're right, it's very difficult to explain and maybe it's even an incorrect thought. I'll try to do a code workaround [18:15] just rename the file in version control to "config.cfg.default" [18:17] and then I have to put under version control.. but what if I want to keep it unchanged? [18:22] BlackLukes: as stbuehler said, a common pattern with config files is to version a template, copy that to something unversioned (and ignored), changes to that won't propagate [18:24] ah yeah that make sense [18:24] thanks, I'll do in this way [18:24] BlackLukes: ok, I hope that solves your use case :) [18:25] another idea (if you have "include" available for your configs), is to have a versioned "config.default" and a "config"; the "config" includes "config.default" and tries to include "config.custom" if it exists [18:27] yeah, another good way. I think I can do this also through code === abentley1 is now known as abentley [19:38] How can I fix these errors: http://flare183.pastebin.ca/1375983 [19:41] [#| ] None ? 3257KB 24KB/s | copying revision 176/2287 [19:41] * SamB wonders if he needs to pull a new version of bzr-svn [19:42] Flare-laptop: did you actually install bzr? [19:42] SamB: Yeap [19:42] where is bzrlib? [19:42] And it gives me this crap too: ImportError: Bad magic number in /usr/lib/python2.5/site-packages/bzrlib/__init__.pyc [19:42] OH [19:43] maybe try reinstalling? [19:43] ... [19:43] Ok, I'll try that [19:43] fail [19:43] It still gives me the same error [19:43] still bad magic number? [19:43] Yeap [19:43] that's not good [19:44] Yeah I know [19:44] It works fine on my Ubuntu Laptop, but not on my Arch Linux Server [19:44] :( [19:44] % python --version [19:44] what's the output for you? [19:44] Python 2.6.1 [19:44] oh. [19:44] I have BOTH python 2.5 and 2.6 installed [19:44] and what is the shabang line on bzr? [19:45] Sorry for the cpas [19:45] caps* [19:45] shabang? [19:45] the first line [19:45] it has a "#!" at the beginning [19:45] I'm confused [19:46] i.e., run "head $(which bzr)" [19:46] read off the first line [19:46] oh ok [19:46] hold on === serg_ is now known as serg [19:46] Output: http://flare183.pastebin.ca/1375998 [19:47] SamB: ^^ [19:48] okay ... so it looks like you have python 2.6 trying to load bytecode compiled by python 2.5 [19:48] Yeah, and I'm also having problems with my supybot not wanting to load either [19:48] out of the python 2.5 library directory, no less ... [19:48] Same problem too, I think [19:49] what does "echo $PYTHONPATH" tell you? [19:49] oh, btw, this is the shabang line: [19:49] Ok, here is stupid part; when I go to install supybot, it installs python2.6 [19:49] #!/usr/bin/python [19:49] hold on [19:49] ahh ok [19:49] called that because it starts with a sharp and a bang [19:49] (# and !) [19:49] lol like crunchbang [19:50] whoa wt* [19:50] [jesse@flare183 ~]$ echo #PYTHONPATH [19:50] [jesse@flare183 ~]$ [19:50] There's nothing there! [19:50] er, $, not # [19:50] oops [19:50] * Flare-laptop facepalms [19:50] hehe [19:50] :S [19:50] Same thing [19:51] oh, okay [19:51] that is very wierd ... [19:51] I agree [19:51] * Flare-laptop wishes he had enough RAM to install Ubuntu Server [19:51] jam: Please can I chat to you about Bug 350796? [19:51] Launchpad bug 350796 in bzr "merge_depth from merge_sorted_revisions incorrect." [Undecided,New] https://launchpad.net/bugs/350796 [19:51] maybe you should ask in a forum (channel, mailing list) for Python on Arch [19:52] alright [19:59] * SamB wishes the gcj runtime library on Debian was in smaller pieces ... [20:01] SamB: But what exactly should I say/put on the forums? [20:46] * cody-somerville pouts at how difficult git is. [20:47] * cody-somerville huggles his bzr. [21:41] Why does bzr try to push through http on Windows? [21:42] Instead of through bzr+ssh? [21:42] Is pushing through http even possible (Launchpad here) [21:42] magcius: are you specifying a http:// url? [21:42] jelmer, lp:* [21:42] magcius: it will use http:// if you haven't logged in [21:43] try "bzr lp-login" [21:47] jelmer, he tried it. [21:57] bzr noobie here [21:57] I did a bzr pull, now how do I merge the changes with the local working copy? [21:57] magcius: bzr pull will automatically update the working tree [21:58] jelmer, hmm... [21:58] jelmer, it didn't appear to change anything in emacs. [22:10] jelmer, bzr says to use "merge" then "push". Even after I do that it does the same thing. [22:10] jelmer, oh, I had to merge, commit, then push. [22:11] Just a usability issue, that's all. [22:11] magcius: "bzr pull" only works if you are behind on the branch you're pulling from, not if you have any local commits [22:12] if you have any additional commits locally, you have to merge [22:12] jelmer, yeah, I fixed the pull issue. I am trying to push, it said to "merge" then "push" [22:12] I had to merge, then commit, then push. [22:12] Just an unhelpful error message, that's all. [22:45] Is there a quick how-to on bzr merging? [22:53] jelmer: just wanted to give some feedback: bzr dpush does *not* give you the svn-revisions back, even if you had new commits pushed to upstream svn [22:54] perhaps that works if you never used the normal push, didn't try that [22:54] magcius: maybe in the user's guide - it's mostly just a matter of 'merge, fix any conflicts, commit', though [22:56] jelmer: ping [22:59] jml: ping [22:59] lifeless: pong. [22:59] lifeless: I'm on a call with thumper at the moment. [22:59] I have a tiny review I'd like done when you can - lp:subunit, the polish branch. Its a no-brainer [23:00] I just got a mail titled 'using SubUnit for Python on Parrot' :) === Spaz_ is now known as Spaz [23:39] lifeless: moin [23:39] lifeless: one sec [23:44] jelmer: hi [23:44] jelmer: subunit, you put up an RFP [23:44] lifeless: ah, yeah [23:46] jelmer: its come a ways since then, and with my polish branch installs correctly AFAICT [23:46] lifeless: did you request a merge or anything for the polish branch? [23:46] yes [23:46] lifeless: since I didn't see any email yet [23:47] are you suscribed to trunk for reviews [23:47] ah, probably not [23:48] https://code.edge.launchpad.net/subunit/+activereviews [23:48] thanks [23:49] hmm, subunit developers is the standard reviewer for lp:subunit [23:49] subscribe to trunk [23:49] the default subscribtion status changes when reviews were added [23:49] s/changes/changed/ [23:50] existing subscriptions were not updatd [23:53] ahh