[07:26] <gour> morning
[07:26] <gour> considering there is 1.18rc1 does it mean next bzr release won't be 2.0?
[07:36] <poolie> gour, it means the one on monday will be 1.18 but we'll shortly follow with 2.0b1 or rc1
[07:36] <poolie> see the recent 'metronome' mail
[07:36] <gour> poolie: ohh, thanks
[07:36] <poolie> np
[07:56]  * gour is waiting for bzr-2 to see if it could work nicely with bzr-fastimport & darcs repos using LP
[08:07] <lifeless> gour: there's no need  to wait
[08:07] <lifeless> gour: just set the repo format to 2a
[08:07] <lifeless> 2.0's primary difference from the release before it will be to change the default format
[08:09] <gour> lifeless: hmm, so, 2a is available in 1.17?
[08:09] <lifeless> yes
[08:09] <lifeless> and 1.16
[08:09] <lifeless> but its best to use as recent a release ass possible - we've been bugfixing like made
[08:10] <gour> to be frank, constant change of repo-formats was the main reason why i went back to darcs :-/
[08:10] <lifeless> *mad*
[08:10] <gour> ok. i've 1.17 on arch
[08:10] <lifeless> gour: we hope 2a to be the last word for quite some time :)
[08:10]  * gour hopes as well...the old creation of new formats was, imho, disaster
[08:11] <gour> emacs is going to switch soon?
[08:22] <lifeless> gour: the thing that caused actual problems AFAICT was rich roots more than anything else; and thats addressed by 2a - everyone has to upgrade to rich roots when they move to 2a
[08:22] <lifeless> poolie1: did the talk go well?
[08:23] <poolie1> hasn't happened yet
[08:23] <poolie1> it's in about an hour
[08:23] <poolie1> hugh's was good
[08:23] <poolie1> most of it is in mandarin; aside from that it's a lot like an early lca
[08:24] <gour> lifeless: i could agree...still, introducing so many formats was, imho, bad PR
[08:25] <gour> lifeless: what about LP? works with 2a?
[08:25] <lifeless> yes
[09:16] <poolie> lifeless: still here? staging seems to be down btw
[09:16] <poolie> not necessarily an emergency
[16:05] <Colonel-Rosa> Can bazaar create patch files?
[16:22] <garyvdm> Colonel-Rosa: bzr bundle or bzr send
[17:44] <bialix> hi garyvdm
[17:44] <garyvdm> Hi
[17:44] <bialix> I've finished rework of commit data save/restore
[17:46] <garyvdm> Cool
[17:46] <bialix> I'd like someone (you or vila) will review it
[17:49] <garyvdm> ok
[17:52] <pygi> bialix, uh, if garyvdm has to review something that will take years :p
[17:52]  * pygi hides
[17:52]  * pygi waves
[17:52]  * garyvdm waves
[17:52] <bialix> pygi: this is no problem
[17:52] <garyvdm> bbl - dinner
[17:52] <pygi> garyvdm, bon appetit :)
[17:52] <pygi> bialix, it was a joke :)
[17:53] <bialix> you are joking about gary all the time
[17:53] <pygi> ok ok, I'll stop :)
[17:53]  * pygi stops
[17:53] <bialix> this looks suspicious
[17:53] <pygi> why?
[17:53] <bialix> never mind, it was joke
[17:54] <pygi> xD
[17:54] <bialix> I forget adding smilie
[17:54]  * pygi needs to stop joking with people
[17:54] <bialix> but this is just my observation
[17:54] <bialix> never mind
[18:07] <gsuveg> re
[18:44] <ronny> i guess im blind, but whats a good way to get all files added on a fresh commit (with no parents)?
[18:44] <ronny> for all the other cases i can mostly use revisiontree.changes_from(other_revisiontree)
[18:54] <ronny> hmm, i can just use list_files
[18:57] <ToyKeeper> D'oh.  I wonder YTF a branch-specific config value would be masked by a generic default from ~/.bazaar/locations.conf .
[18:57] <ToyKeeper> Seems backward to me.
[18:58] <ToyKeeper> It looks like locations.conf is even overriding the value I passed on the command line.
[19:17] <luks> ronny: I guess you can get changes_from against a NULL_REVISION revision tree
[19:33] <ronny> luks: works fine, thanks
[19:33] <ronny> \o/ the first diff support unittests for anyvc pass
[19:54] <garyvdm> pygi: ping?
[20:11] <garyvdm> bialix: I took a look at commit_data. I'm happy to merge now (so that we get user testing from trunk users.) Feed back from Vincent is important, but we get get that later.
[20:12] <bialix> garyvdm: thanks. I'll merge it soon (maybe tomorrow).
[20:13] <bialix> when I worked on bugs.py I thought that we have to use too much private bzrlib config API (with leading underscore)
[20:13] <bialix> maybe we need to provide some patches for bzrlib/config.py, but later
[20:14] <bialix> garyvdm: about find_branches/iter_branches: I'd prefer to have new method iter_branches
[20:14] <garyvdm> Yes - I think lets merge it, and then you can continue to work on improving it in those regards.
[20:14] <bialix> in this case backward compatibility will be easier for us
[20:15]  * bialix nods
[20:17] <garyvdm> I'm working on Bug 412029 and then Bug 412035. Still not sure how I'm going to do the later.
[20:17] <garyvdm> But it is so important. That bug so often forces me to the command line
[20:19] <bialix> bug 412029?
[20:19] <bialix> ah yes
[20:20] <bialix> it's nasty bug
[20:20] <bialix> garyvdm: about 412035
[20:21] <bialix> we have context menu for invoking diff, right?
[20:21] <bialix> why not force user to select parent here
[20:21] <garyvdm> Yes, but also a button, and the file list.
[20:21] <bialix> we can force users to select parent all the time
[20:21] <garyvdm> I hope I can find a way for all 3.
[20:22] <bialix> it could be nagging
[20:22] <bialix> and people start crying
[20:23] <bialix> so we either should provide more options in context menu (as you did to select between different diff tools)
[20:23] <bialix> or use special dialog
[20:24] <bialix> garyvdm: what about creating short-cut diff option: default diff tool, left-hand parent always
[20:24] <bialix> and rich diff option: with all possible options to select
[20:24] <bialix> something like Advanced search in LP bugs UI
[20:24] <bialix> garyvdm: ?
[20:24] <garyvdm> Yes - a combo box would provide that to.
[20:25] <garyvdm> (default to left hand parent)
[20:25] <bialix> no, I mean special dialog
[20:25] <garyvdm> Oh - ok
[20:25] <bialix> how often you need special options?
[20:26] <bialix> so with short-cut people will have old behavior
[20:26] <bialix> and with new rich options dialog they will have to select and run diff
[20:26] <garyvdm> What I'm really looking for is a way to select the parent, when selecting the revison(s) in the list.
[20:26] <bialix> in 2 steps: dialog with options -> OK -> launch diff
[20:27] <bialix> You can show short summary info about parent revisions?
[20:27] <bialix> e.g. revno + author + summary
[20:27] <bialix> something like bzr log --line
[20:27] <bialix> put this into combobox
[20:28] <bialix> radiobuttons could be unoptimal for octopus merge case
[20:31] <garyvdm> I'm going to do bug 412029 first - Because I think that that might affect my ideas...
[20:32] <bialix> ok
[20:32] <bialix> I'm planning to work on Bug #392920 because it beats me on win32 quite regularly
[20:33] <bialix> garyvdm: after this I think working on auto screenshots and docs + site
[20:35]  * bialix mutters: maybe auto screenshots could be used as semi automated tests later
[20:35] <garyvdm> bialix: I was thinking the same thing...
[20:36] <bialix> garyvdm: I hope to use qbzr trunk history for generating screenshots
[20:36] <garyvdm> wow!
[20:36] <bialix> just need to figure out better places/revisions
[20:36] <bialix> and maybe your help for programmatically manipulate with twisties in qlog
[20:40] <bialix> garyvdm: I'm done for tonight. Good night and easy hacking!
[20:45] <kenichi> "An attempt to access a url outside the server jail was made"  ...  just upgraded the bzr and bzrtools on a host that was serving through apache httpd, and now get this from clients.  can anyone help? (1.10 -> 1.17)
[20:45] <beuno> kenichi, are both client and server 1.17?
[20:45] <kenichi> yes
[20:46] <pygi> garyvdm, pong?
[20:51] <kenichi> beuno: anything else?
[20:53] <pygi> garyvdm, once you're around, tell me what you needed :)
[21:17] <garyvdm> Hi pygi: I think that you were work together with vila on the bzr-gtk, save and restore commit messages.
[21:18] <pygi> garyvdm, I'm not sure, but I don't think so?
[21:18] <garyvdm> bialix has been working on and improved version for qbzr. I know that vila is interested in making some of the code common, so I though that you would like to take a look at what bialix has done.
[21:19] <garyvdm> see: https://code.launchpad.net/~qbzr-dev/qbzr/commit_data/+merge/10040
[21:22] <pygi> vila, were we working on that? I don't think so. Perhaps Szi?
[21:25] <lifeless> moin
[23:43] <igc1> morning