[00:30] Hi, I have a question about bzr-git. My project is using bzr, but I'd like to embed another one in it that uses git. Would it be best to store the git one in another folder outside and just copy over the contents periodically, or can something be fancy done with bzr-git to help with this? [00:36] vadi2: you would be able to do that with bzr-git [00:36] it requires nested tree support in bzr, which isn't quite there yet [00:36] so trying to do it right now would be a bit tricky [00:37] alright. I can wait [00:37] is there any report I can subscribe to? [00:38] I don't think so [00:38] I think there's a wiki page though [00:39] and there's been plenty of discussion on the mailing list in the last couple of months [00:41] https://blueprints.launchpad.net/bzr/+spec/nested-tree-support ? [00:42] yep, looks like what I need [00:43] so, there't the "join" command that can join two trees together [00:43] but I'm guessing that you're not really joining them, the git project is still external, you just want to embed a copy of it? [00:43] yeah. [00:43] yeah, then you need that nested tree support unfortunately [00:44] alright. I've subscribed to the blueprint, so we'll switch as soon as that's in [00:44] cool [00:45] thanks for your time [01:04] hi guys [01:04] I'm thinking of implementing a new protocol for bazaar [01:04] where should I start ? [01:10] Just out of curiousity, what kind of protocol? Wh? [01:10] Err, why*? [01:10] Apparently I'm curious enough to ask, but not enough to stay around for the answer. See you later! :D [01:10] * Peng_ /away! [01:17] visik7: I assume you're current with all the discussion, analysis, and work that is currently being done with the existing smart server protocol. What do you want to do that is different? [01:18] I want to implement bazaar over an IM protocol === Kissaki is now known as Kissaki^0ff [01:18] ... well that's a novel concept [01:19] so I could push and pull from my collaborators without an intermediate server and without forcing others to install servers or anything harder than a plugin :) [01:19] wouldn't be cool if you could just say hey pull from me my latest changes [01:19] something like bzr avahi but over IM [01:21] avahi isn't actually a transport, just a discovery mechanism. [01:21] Can IM networks do multiple Mbps of traffic? [01:21] Seems like a doable, though terribly inefficient, idea. [01:21] * Peng_ shrugs. [01:21] * Peng_ goes back to being /away. [01:22] I think is great [01:24] you'll not discourage me [01:24] :) [01:28] visik7: well, if you can "send files" between you and your buddies, then just use `bzr bundle` to create patch bundles and send those across. It's not quite push and pull, but once you receive a bundle you just merge (or pull) as you would normally. [01:28] (we do that all the time) [01:29] Otherwise, yes, I imagine the bzr dbus/avahi work that Robert did a few years ago might be a place to start. === KX is now known as zombor === zombor is now known as KX === KX is now known as Clippit === Clippit is now known as KX [08:08] fullermd: around? [08:08] fullermd: https://edge.launchpad.net/libcpuinfo/ needs a BSD module [10:21] lifeless: Hm. I'll take a look... [10:21] fullermd: only if you care ;) [10:22] Well, see, I'm thinking of the work I should be doing tonite... it's tempting to care about classifying toenail clippings instead of that, so... ;p [10:25] Hm. "the University" is listed in the 3rd clause of the license? [10:27] oh foo [10:27] I'll correct that [10:27] Not to get into a license discussion, but why 3-clause instead of 2? [10:27] Isn't 3-0clause the modified? 4-clause was evil [10:27] 3-clause is what debian carries as the 'stock BSD' [10:28] I tend to use a 2-clause. [10:29] so just 1 and 2? [10:29] Yah. It ends up being isomorphic to the MIT/X license. [10:29] I'll look into that later [10:29] * fullermd nods. [10:29] for now the obvious change is sufficient [10:30] How'd you have in mind handling the OS choice? [10:30] Just #ifdef's? [10:31] if trying to read /proc/cpuinfo will segfault or something [10:31] but just chaining the ifs should be simple and easy [10:31] if it won't compile a particular check on other platforms then yes ifdefs are solid [10:32] Probably won't link, I'd use sysctl(3) (not (1)) on BSD, so... [10:32] specifically change [10:32] if (result < 1) goto fail; [10:32] insert before that [10:32] if (result < 1) result = cpuinfo_MODULE_processor_count(); [10:32] But chaining the if would work OK, there's nothing in /proc but pid's and a curproc symlink (if it's even mounted) [10:33] Oh, I was thinking in cpuinfo_proc_cpuinfo_proccessor_count() [10:33] I prefer to build as much as possible because it avoids things bitrotting as much [10:33] adding a bsd module should be as simple as adding a new .c and .h [10:33] and changing cpuinfo_processor_count to call your new function [10:33] which itself would return -1 on 'cant tell' or otherwise a count [10:34] * fullermd nods. [10:34] proc_cpuinfo is a specific module [10:34] Gotcha. [10:34] I'll take a poke after I get some coffee... [10:34] sweet === Kissaki^0ff is now known as Kissaki [11:26] lifeless: Push up a branch and figure out how to send a review request? [11:28] sure [11:34] 'k, sent. [11:37] lol copy n paste :) [11:37] I don't have any convenient non-BSD systems to check it on, but it should build cleanly I think. [11:37] +/* Probe proc/cpuinfo */ [11:37] testing it [11:37] Darn network corruption! [11:38] Why did that show up in my bzr mailbox... [11:38] Oh, frell. Freakin' LP... [11:41] ta! [12:37] I have a *local* svn repo. how can i import it to bzr? it's huge (> 4GB) so i'd prefer to not to do it over the network, but it's not a svn:// url.. [12:38] bazaar supports just one .bzrignore file per tree? [12:57] * AfC wonders if there is a fast-export that takes a svn-dump as an import? [12:57] fta: (I know that Jc2k used svn dumps to import all of GNOME into bzr) [12:58] oh, wait... can't you use file:/// URLs with Subversion? If so, that should Just Work with $whichever bzr conversion tool you're using (maybe)? [12:59] well, my 1st try was with bzr branch svn:// [13:03] My first use of Subversion was with a file:/// repo, but that was *long* ago. [14:33] ok, i figured out how to start my svn import from my local copy, but for some reason, it's doing a lot of http :( [14:34] [\ ] http 268885KB 1270KB/s | copying revision 17/14146 [14:34] will take a while :( [15:58] it died after 2000 revisions :( [20:00] is there a way to get a mail on commit ? [20:12] hmm, where is jelmer? === sabdfl1 is now known as sabdfl [22:01] There's a story behind the code name for 1.16, isn't there? [22:20] davidstrauss: yes [22:20] jml was wanting to get it out on Thursday [22:21] thumper: Makes sense now :-) [22:21] but it ended out being Friday (Sydney time) which was still Thursday on the west coast of the us [22:21] US [22:21] I figured it was a time zone thing [22:21] :-) [22:22] anyone using buildout ? [22:30] bzr.dev revision:4439 seems really slow, just a local issue or others notice this too? [22:30] I haven't noticed any issues, not that I've used it much. [22:31] looking at check-ins [23:45] good morning #bzr [23:46] morning jml [23:47] moin [23:47] Morning jml. [23:47] RAOF, lifeless, james_w: g'day [23:48] spot the odd one out [23:50] hello all === lifeless is now known as all [23:50] hi === all is now known as lifeless [23:56] ronny: hi [23:56] hello ronny [23:56] and jelmer [23:56] fta: hi [23:56] moin poolie [23:57] Hmmm, all of you here? I smell productivity! :D [23:57] nooooooooo [23:58] poolie: hi [23:58] james_w: :)