=== wgrant_ is now known as wgrant | ||
=== zyga-afk is now known as zyga | ||
zyga | mwhudson, ? | 06:46 |
---|---|---|
zyga | mwhudson, some paste | 06:46 |
gour | morning | 07:00 |
gour | co-located branches doc (http://doc.bazaar.canonical.com/developers/colocated-branches.html) lists few 'use cases' (http://doc.bazaar.canonical.com/developers/colocated-branches.html#use-cases) and i'm not sure whether co-located branches are real need in bzr (yesterday i played with --lightweight checkouts used for feature branches based on shared repo) or is it more of filling the gap to hg/git's features? | 07:03 |
spiv | gour: they are more convenient than lightweight checkouts + branchless repo, and just as good for many cases, so that's a pretty valuable improvement | 07:14 |
gour | spiv: i see...thanks...are they going to be ready for 2.6? | 07:15 |
spiv | They're already ready in 2.5: http://doc.bazaar.canonical.com/bzr.2.5/en/whats-new/whats-new-in-2.5.html#basic-colocated-branch-support | 07:17 |
spiv | No doubt they'll be even better in 2.6 :) | 07:17 |
gour | iirc, jelmer or someone else was speaking something about UI not polished (yet)...otoh, i was not aware they're already in...does it mean we should (try) to adopt 'em instead of lightweight co-s & branchless repos? | 07:19 |
spiv | I'd give it go, yes. | 07:19 |
gour | ok | 07:19 |
spiv | If you don't like them for some reason you can always change your mind later and use lightweight checkouts and boring old branches in treeless repo, etc :) | 07:20 |
gour | :-) | 07:21 |
gour | in some use cases, i like treeless repo and lw co-s...eg. i'm using weechat irc client and my ~/.weechat is lightweight co keeping track of weechat's config files, while the treeless repo is in ~/repos | 07:23 |
gour | hmm, colocated branches need new format...we just talked about bzr's formats yesterday :-) | 07:27 |
gour | 2.6 docs when speaking about colocated branches uses future tense, so wondering what's avaialable today... | 07:29 |
bob2 | suspect once you get used to them you'll want to use colocated all the time | 07:39 |
spiv | bob2: well, it's not especially well-suited for the .dotfiles use case mentioned above. | 07:42 |
spiv | But for general code hacking, yeha. | 07:43 |
jelmer | gour: huh? | 07:44 |
jelmer | gour: colocated branches need a new format? | 07:44 |
gour | jelmer: bzr init -h ==> --development-colo The 2a format with experimental support for colocated | 07:47 |
gour | branches. | 07:48 |
gour | or it means something else i'm not aware of? | 07:49 |
jelmer | gour: the development-colo format isn't necessary for colocated branches anymore | 07:50 |
jelmer | gour: you can just use them with 2a these days | 07:50 |
gour | jelmer: hmm, it seems that docs needs some love then ;) | 07:51 |
jelmer | gour: which docs? | 07:51 |
gour | jelmer: bzr-2.5.1 help text for 'init' command | 07:53 |
mgz | morning! | 08:00 |
jelmer | hey mgz | 08:03 |
jelmer | gour: ah, but I guess it's listed under the development formats still? | 08:03 |
gour | jelmer: well, it's listed under branch format section along with --2a etc., iow. valid option for 'init' in 2.5.1 | 08:07 |
jelmer | gour: sure, but there's lots of other uninteresting formats there too | 08:09 |
jelmer | *development | 08:09 |
jelmer | though we should probably hide development-colo | 08:09 |
* gour nods | 08:14 | |
anddam | do I need to register a launchpad account in order to just clone (branch?) a repository? | 12:44 |
jelmer | anddam: no, you can clone a branch (as long as it's not private) without an account | 12:44 |
anddam | jelmer: do I need to init a repo before running bzr branch? | 12:45 |
jelmer | anddam: no, see the mini tutorial | 12:46 |
jelmer | http://doc.bazaar.canonical.com/latest/en/mini-tutorial/ | 12:46 |
anddam | I tried both, "bzr branch lp:foo" and "mkdir foo; cd foo; bzr init; bzr branch lp:foo" | 12:46 |
anddam | k | 12:46 |
anddam | I just read that | 12:46 |
anddam | it shows how to init-repo and init but it doesn't says it that's *mandatory* or not | 12:46 |
anddam | with git a "git clone" would create the local repository as well | 12:47 |
jelmer | anddam: in bzr the repository is created if a shared one doesn't exist too | 12:48 |
anddam | I'm trying cloning lp:groundhog | 12:48 |
anddam | I get bzr: ERROR: Not a branch: "http://bazaar.launchpad.net/~juanjux/groundhog/groundhog-juanjo/" | 12:48 |
anddam | when running "bzr branch lp:groundhog", is that an issue on my side or not? | 12:49 |
jelmer | anddam: that branch doesn't really exist - it's never been pushed to according to the web page | 12:49 |
anddam | sounds like it isn't | 12:49 |
anddam | I see | 12:49 |
anddam | Recent revisions | 12:50 |
anddam | This branch has not been pushed to yet. | 12:50 |
anddam | so the launchpad repository for the project is empty, isn't it? | 12:50 |
jelmer | it doesn't actually have a control directory I think | 12:50 |
jelmer | so it's not even empty | 12:51 |
anddam | I apologize but I'm really new to bzr | 12:51 |
anddam | idk what a control directory is | 12:51 |
anddam | the point is that lp:groundhog has no code, is this correct? | 12:51 |
jelmer | it's the same thing as ".git" in git | 12:51 |
jelmer | I mean, in git the control directory is named ".git" | 12:52 |
jelmer | anddam: yes, lp:groundhog doesn't have any code | 12:52 |
anddam | I see lp:~juanjux/groundhog/trunk is listed as branch and that has code | 12:52 |
anddam | is there any difference between a subdir in user space or a root level project? | 12:52 |
anddam | I mean is that just a convention, a matter of taste or so? | 12:53 |
anddam | (fetching) | 12:53 |
jelmer | anddam: I'm not sure I follow | 13:00 |
jelmer | anddam: subdir in user space? | 13:00 |
anddam | jelmer: yes, ~juanjux/groundhog rather than groundhog | 13:04 |
jelmer | anddam: ah, that's just convention | 13:05 |
anddam | thx | 13:11 |
anddam | bye | 13:11 |
abentley | vila: Could you please review https://code.launchpad.net/~abentley/bzr/config-branchname/+merge/110825 ? | 13:38 |
=== zyga is now known as zyga-afk | ||
=== zyga-afk is now known as zyga | ||
abentley | jam: Changing branchname to fall back to basename is a 1-line change, and I'm happy to do that. Maybe rename to "shortname". But you're not proposing we remove "basename", right? So we'd still be introducing a new variable. | 14:51 |
jam | abentley: I'm fine with a new var, what I want is to have 1 var that can just be used, without needing to think "do I need X or Y in this case" | 14:51 |
abentley | jam: Okay. Does "shortname" work? ("branchshortname" seems...long) | 14:52 |
jam | branchname is fine for me | 14:53 |
jam | I think I prefer that | 14:53 |
abentley | jam: "branchname" was meant to refer to the colocated branch terminology. jelmer, is it cool if I introduce a config var called "branchname" that is ",branch=$NAME" but falls back to "parent/$NAME"? | 14:55 |
=== zyga is now known as zyga-afk | ||
mgz | Have we got a bug for `bzr config --remove launchpad_username` not working on 2.5.1? | 17:55 |
mgz | needs --scope=bazaar | 17:56 |
gour | /q | 20:33 |
=== pjdc_ is now known as pjdc |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!