=== davidstrauss_ is now known as davidstrauss
[00:10] <jelmer> slangasek: hi
[00:11] <jelmer> slangasek: James was about to release another version of bzr-builddeb so I was waiting on that before uploading to sid
[00:11] <jelmer> james_w: ^
[00:13] <igc> hi lifeless, jelmer
[00:13] <jelmer> hey Ian
[00:13] <jelmer> igc: Congrats on bzr-explorer 1.0!
[00:13] <slangasek> jelmer: ok :)
[00:14] <igc> thanks jelmer
[00:15] <lifeless> hi igc
[00:15] <lifeless> jelmer: did you get the commitfromnews tests working ok?
[00:15] <jelmer> I haven't tried yet
[00:16] <lifeless> ok
[00:16] <lifeless> bzr.dev change is landed
[00:22] <lifeless> is it doc or docs we use for doc bugs?
[00:25] <poolie> Bazaar version control | try https://answers.launchpad.net/bzr for more help | http://irclogs.ubuntu.com/ | Patch pilot: vila | bzr 2.1.0final has gone gold, time to build installers!
=== poolie changed the topic of #bzr to: Bazaar version control | try https://answers.launchpad.net/bzr for more help | http://irclogs.ubuntu.com/ | Patch pilot: vila | bzr 2.1.0final has gone gold, time to build installers!
[00:26] <igc> hi poolie
[00:26] <poolie> hello igc
[00:26] <poolie> how are you?
[00:26] <igc> ok today
[00:26] <igc> time for a call sometime today maybe?
[00:27] <igc> poolie: ^
[00:27] <poolie> now's good if that suits
[00:28] <igc> poolie: it does
[00:29] <mwhudson> lifeless: yes
[00:29] <lifeless> mwhudson: its doc, I looked
[00:29] <mwhudson> although it redirects and that should work for a good while
[00:29] <lifeless> mwhudson: oh, the out of date question.
[00:30] <lifeless> ok, I won't stress about the link
[00:34] <mkanat> mwhudson: So, no crashes all this time?
[00:35] <mwhudson> mkanat: one, but i think a sysadmin was woken up for it, so no debugging :/
[00:35] <mkanat> mwhudson: :-(
[00:35] <mwhudson> yeah
[02:28] <dsuch> Hmm is there any notion of a session when working with WorkingTrees? My application accepts HTTP requests and upon receiving one I'd like to create some sort of a 'bzrlib session' which would go down through a couple of layers of the code,
[02:29] <dsuch> possibly ending in its committing if no exception have been raised. However, if an exception is raised, I'd like to rollback any possible changes.
[02:29] <dsuch> revert rather :)
[02:30] <dsuch> Something along starting a new SQL transaction I guess.
[02:32] <dsuch> Actually, not sure if it matters, but there would be only one commit made deeper in the code and I think I'd like to commit the commit right before finishing with the HTTP request.
[02:33] <dsuch> Or commit the session, if there is such a thing.
[02:35] <mwhudson> dsuch: there are write locks and write groups, which i think are at least a bit like that
[02:37] <dsuch> okay, I'll take a look at that, thanks
[02:43] <jmcantrell> how do i get the repo url that i branched?
[02:45] <dsuch> jmcantrell: bzr info ?
[02:46] <jmcantrell> dsuch: if i need to get it in a script, is the best way to just parse that output or is there a better way?
[02:47] <dsuch> in Python?
[02:48] <jmcantrell> dsuch: bash
[02:48] <dsuch> doesn't seem bad
[02:49] <jmcantrell> dsuch: yeah, i just didn't know if there were a way to get just the path
[02:50] <dsuch> I'm not sure, there are lots of commands but I tend to stick to a few mostly used, take a look at 'bzr help commands', maybe there is some shorter one
[02:54] <jmcantrell> dsuch: ok. thanks
[03:12] <poolie> thanks for all the reviews jelmer
[04:06] <parthm> hello. I have lp:bzr-grep working to grep a specified revision for a pattern. http://pastebin.com/AkccBvx8
[04:06] <parthm> I am working on extending it to take a range. I found builtins._get_revision_range(revision, wt.branch, 'grep') returns two RevisionInfo objects.
[04:07] <parthm> Whats a good way to get a list of inclusive revision for this range.
[04:14] <poolie> hello parthm
[04:14] <poolie> that'd be a nice featuer
[04:15] <parthm> hi poolie
[04:15] <poolie> *feature
[04:15] <poolie> there should be a repository method for that
[04:15] <poolie> i'd look in log
[04:16]  * parthm goes digging into log code
[04:17] <poolie> i'd probably call something like _graph_view_revisions from log.py actually
[04:17] <poolie> igc, hi?
[04:17] <igc> hi poolie
[04:18] <poolie> igc i think bug 524177 is now fixed, if you want to try again
[04:18] <ubottu> Launchpad bug 524177 in bzr "Doc rebuilds are deleting chm files" [Critical,Fix released] https://launchpad.net/bugs/524177
[04:18] <igc> poolie: thanks
[04:19] <parthm> poolie: thanks. i will look at that.
[04:20] <poolie> igc, do you want me to try this thing of removing the ga urchin from the html templates? or are you ok with that?
[05:25] <parthm> hello. For lp:bzr-grep I have a simple loop to iterate through the rev range given by the user http://pastebin.com/qaQyVsjX This works fine and I am able to grep the rev range.
[05:26] <parthm> However, I am seeing a weird failure in a test http://pastebin.com/Nqtr2kCu which tries to grep on an unversioned file. it seems to be failing in at the 'for' clause http://pastebin.com/BfM0Hvhu
[05:27] <parthm> interestingly, running the exact same step from command line works fine (and issues a skip warning as expected). any ideas?
[05:29] <parthm> the run from the command line: http://pastebin.com/NLQhsihG
[05:34] <parthm> never mind. this seems to be the edge case of the repo having only 1 revision. http://pastebin.com/dYJHRxSE Any suggestion on what may be a good fix? Do I just catch the error and ignore?
[05:48] <poolie1> hi parthm
[05:48] <poolie1> sorry did anyone answer before?
[05:49] <lifeless> poolie1: check out lp:bzr-plugin-info
[05:49]  * lifeless goes offline
[05:49] <parthm> poolie1: np. i located and fixed the error. it was a special case for revno:0
[05:50] <parthm> i am done with the rev range support also (just adding tests). you can try it at lp:bzr-grep. i can try to make it a builtin and propose a merge if you think its upto it.
[07:14] <vila> hi all !
[07:21] <poolie1> hi vila
[07:21]  * poolie1 is writing stuff to plot the review queue length
[07:22] <vila> hey poolie
[07:22] <vila> interesting
[07:24] <lifeless> :!bzr git-import
[07:24] <lifeless> Command git-import is supplied by the following plugins:
[07:24] <lifeless>   git
[07:24] <lifeless> poolie1: ^
[07:25] <poolie1> good
[07:26] <poolie1> lifeless: makes it sound like it's actually supplied by them
[07:26] <poolie1> but for you, not so much :)
[07:26] <lifeless> hmm ?
[07:26] <poolie1> "Command '%s' is not available at the moment but might be provided by the following plugins: %r.  See <http://doc......>"
[07:26] <poolie1> oh also make sure it exits nonzero
[07:27] <lifeless> I think might is a little waffly
[07:28] <poolie1> delete it then
[07:28] <poolie1> use something like what ubuntu's thing does
[07:28] <lifeless> thats what I have done;)
[07:28] <poolie1> The program 'darcs' is currently not installed.  You can install it by typing:
[07:28] <poolie1> sudo apt-get install darcs
[07:29] <lifeless> yes
[07:29] <lifeless> we don't have that slick a system yet
[07:29] <poolie1> you don't understand the difference?
[07:29] <lifeless> I was going to add the URL
[07:29] <poolie1> please make it more clear that the command is not actually available yet
[07:30] <lifeless> ok
[07:30] <lifeless> I'm glad you like it enough to offer feedback.
[07:30] <poolie1> i do :)
[07:30] <poolie1> this is just polish :)
[07:31] <lifeless> yes; you're coming across a bit aggro. Sorry if its me being slow or whathaveyou.
[07:32] <poolie1> sorry
[07:33] <lifeless> I just spent an hour twisted sideways on the train; only pasted it here to show that I had actually got <something> done
[07:35] <lifeless> take #2
[07:35] <lifeless> :!bzr git-import
[07:35] <lifeless> Command 'git-import' is not available at the moment.
[07:35] <lifeless> The following plugins include this command:
[07:35] <lifeless>  * git (branch available ../../git/trunk)
[07:36] <poolie1> sweet
[07:37] <lifeless> its pushed
[07:42] <lifeless> vila: beuno: ^ you might enjoy that. (install lp:bzr-plugin-info)
[07:42] <lifeless> poolie1: gnight
[07:43] <vila> lifeless: nice
[07:43] <vila> lifeless: Do^H^H When you plan to put it into core ?
[07:45] <lifeless> it needs discussion
[07:48] <poolie1> vila i think we should ship it somehow
[07:48] <poolie1> whether it's in core or a shipped plugin is not a big deal
=== poolie1 is now known as poolie
[07:49] <poolie> lifeless: you might be interested to know in the last week 38 bugs were filed, 11 were fixed, and 54 left state New
[07:49] <vila> poolie1, lifeless : sure, keeping it as a plugin is fine with me, one more example
[07:49] <poolie> 23 were closed
[07:49] <poolie> 17 became inprogress
[07:49] <vila> poolie: wow, excellent, hydrazine ?
[07:49] <poolie> psql :)
[07:49] <poolie> could be done more slowly through hydrazine
[07:51] <poolie> however i want to plug it into tuolumne aka lpstats
[07:52] <poolie> and that is easier done here
[07:57] <lifeless> vila: grats on hudsonification
[07:58]  * vila returns the compliment ;-)
[08:00] <parthm> poolie: lp:bzr-grep plugin is tested and has the most common features http://pastebin.com/ijCs02kr . Feel free to try it out. I am thinking this can be proposed as a builtin. If you agree I could do the porting.
[08:00] <parthm> I plan to add --external-grep, --include/exclude but if a merge is to be proposed I would like to add these after to keep the proposal size small.
[08:01] <vila> parthm: thanks for resurrecting this plugin !
[08:01] <parthm> vila: welcome :-) this one uses a builtin grep and i have tested it out on windows also.
[08:01] <poolie> wow that's great
[08:02] <poolie> i'll install it now
[08:03] <bialix> hi all
[08:03] <parthm> bialix: hi
[08:04] <vila> hey bialix
[08:04] <bialix> bonjour vila
[08:05] <lifeless> vila: you did all the wrk:)
[08:10] <bialix> jelmer: ping
[08:22] <lifeless> vila: https://code.launchpad.net/~lifeless/bzr-gtk/setup/+merge/20341
[08:35] <poolie> vila, lifeless: https://code.edge.launchpad.net/~mbp/tuolumne-lp-configs/bzr-graphs/+merge/20357 - plot mps in progress etc
[08:35] <vila> poolie: Not allowed here
[08:35] <vila> Sorry, you don't have permission to access this page.
[08:36] <poolie> blah
[08:36] <poolie> oh well you'll have to trust
[08:36] <poolie> me
[08:37] <vila> lifeless: a bit of explanation for that bzr-gtk mp ? (to save me the time to install your plugin and encounter the failure)
[08:37] <vila> poolie: hehe
[08:40] <lifeless> vila: http://doc.bazaar.canonical.com/bzr.dev/developers/plugin-api.html#plugin-metadata-before-installation
[08:42]  * igc dinner
[08:42] <vila> lifeless: excellent
[08:42] <slangasek> james_w: when you're about, I'd appreciate having your eyeballs on bug #529900; I have an opportunity to nudge the Debian Samba packaging in the direction of bzr from svn, and that bug's a blocker for it
[08:42] <ubottu> Launchpad bug 529900 in bzr-builddeb "merge-upstream --v3 tries to unpack .tar.bz2 with tar xzf" [Undecided,New] https://launchpad.net/bugs/529900
[08:43] <poolie> night all
[08:43] <vila> poolie: g'night
[08:56] <jml> lifeless, no
[09:13] <james_w> slangasek: haha
[09:14] <james_w> thanks for the fix, I'll get this uploaded soon
[09:14] <slangasek> ok, cheers :)
[09:38] <soren> Is there a simple way to get bzr-svn to allow me to check out some stuff over https if the certificate on the server is expired and/or signed by an unknown CA?
[09:39] <lifeless> if you tell svn itself that its ok
[09:39] <lifeless> I think it will work
[09:39] <lifeless> YMMV
[09:39] <soren> Oh.
[09:39]  * soren tries again
[09:41] <soren> Ah, I was just doing "bzr branch https://blahblah/"... switching to "bzr branch svn+https://blahblah/" helped. It didn't even ask me to confirm the cert, which I suppose is a bit worrying.
=== Adys is now known as Adys`
=== Adys` is now known as Adys
=== gerard_1 is now known as gerard_
[11:29] <lifeless> vila: 120K tests is kinda scart
[11:29] <lifeless> *scary*
[11:29] <lifeless> gnight
[11:30] <vila> lifeless: yeah, matrix job fallout, I know
[11:31] <vila> lifeless: just divide by the number of slaves :-D
[11:31] <lifeless> :P
[13:55] <igc> night all
[14:44] <mthaddon> can anyone enlighten me on whether there's a way to specify which identity file to use for a given bzr operation (the use case is when the home dir for the user is on a tmp filesystem)?
[14:51] <bialix> what?
[14:53] <vila> mthaddon: identify file as in ssh ?
[14:53] <vila> grr identity
[14:54] <mthaddon> yeah
[14:55] <vila> we obey ~/.ssh/config ... you want a way to force it without having a ~/.ssh directory ?
[14:56] <mthaddon> vila: yeah, similar to -e for rsync
[14:57] <vila> mthaddon: nothing comes to mind... I'm not sure ssh itself obey an appropriate env var...
[14:58] <vila> hmm, there has been talk about specifying
[14:59] <vila> grr delete not enter
[14:59] <mthaddon> vila: in rsync you can just do rsync -e="ssh -i file" and it does what we're hoping for in this case, but that sounds like bzr doesn't support that, right?
[14:59] <vila> mthaddon: I don't think so, I'm checking
[14:59] <mthaddon> thx
[15:00] <vila> There is BZR_SSH but it's supposed to be the executable name only, no options
[15:01] <vila> mthaddon: the command is built in bzrlib/transport/ssh.py
[15:02] <vila> mthaddon: I see to env var used there
[15:03] <vila> mthaddon: two possible tricks: 1) patch b.t.ssh.py to add your -i param, 2) define your own ssh wrapper
[15:03] <vila> both are a bit ugly, the later is just a bit less ugly :-/
[15:04] <vila> mthaddon: feel free to file a bug with your use case
[15:05] <mthaddon> thx, will do :)
[15:06] <vila> errq, s/I see to env var used there/I see *no* relevant env var used here/
[15:07] <vila> mthaddon: out of curiosity what *is* your use case (i.e. why can't you build a usable ~/.ssh dir ?) ?
[15:07] <mthaddon> vila: running a nagios check - user has tmp home dir as it's a role account
[15:07] <vila> tmp as in empty ?
[15:19] <GaryvdM> Hi all
[15:19] <GaryvdM> Hi bialix
=== khmarbaise_ is now known as khmarbaise
[15:21] <bialix> Hi GaryvdM !!!
[15:31] <bialix> GaryvdM: I'm a bit busy with current work, so I was unable to release qbzr 0.18.3 this weekend
[15:31] <bialix> maybe next weekend
[15:31] <GaryvdM> Maybe I can do it this week.
[15:32] <bialix> will be nice
[15:33] <vila> hey GaryvdM !
[15:34] <LeoNerd> I think my dist-upgrade just broke bzr-svn... I now get: bzr: ERROR: exceptions.AttributeError: 'BranchConfig' object has no attribute '_get_change_editor'
[15:34] <GaryvdM> Hi vila
=== salgado is now known as salgado-lunch
[15:47] <jelmer> LeoNerd, please file a bug against bzr-svn
[15:48] <LeoNerd> OK willdo
[16:05] <LeoNerd> jelmer: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=572109
[16:05] <ubottu> Debian bug 572109 in bzr-svn "AttributeError during 'bzr shelve'" [Normal,Open]
[16:08] <RainCT> Hi
[16:09] <RainCT> About that websites thing in Explorer, I'm not sure how it works but wouldn't it make sense to let projects include a dot-file directly in the branch with that info (instead of having to get it merged)?
=== deryck is now known as deryck[lunch]
[16:13] <jelmer> LeoNerd, thanks
=== beuno is now known as beuno-lunch
=== kfogel is now known as kfogel-loonch
[16:31] <shortlord> Is it possible to create a daily build with bzr-builder for software that is not hosted on Launchpad and does not use bzr? (SVN for example?)
[16:32] <james_w> yes
[16:32] <james_w> if you have bzr-svn installed
[16:35] <shortlord> james_w, oh wow, that sounds great. thx :)
=== deryck[lunch] is now known as deryck
=== beuno-lunch is now known as beuno
=== salgado-lunch is now known as salgado
=== kfogel-loonch is now known as kfogel
[17:27] <ubuntujenkins> can you push branches to launchpad with out an ssh key?
[17:28] <maxb> no
[17:28] <rubbs> ubuntujenkins: not that I know of
[17:28] <ubuntujenkins> fair enough that just means a little more work thats all , thanks rubbs
[17:29] <rubbs> np
=== bialix_ is now known as bialix
[18:14] <anddam> hello
[18:14] <rubbs> hello
[18:14] <anddam> woa 2.1.0
[18:15] <anddam> I guess 1.4 is considered old
[18:17] <anddam> was binary called baz back then?
[18:18] <nDuff> I don't believe so
[18:27] <GaryvdM> anddam: baz and bzr have unrelated version numbers. bzr started at 0.1
[18:27] <anddam> what's the difference? I installed bazaar with macports and I see just a single binary "baz"
[18:27] <anddam> 1.4.2
[18:28] <anddam> well there's annotate-baz too, but no bzr
[18:28] <anddam> oh lol
=== radoe_ is now known as radoe
[18:30] <anddam> it's a different app,  "NOTE: this is baz, aka Bazaar 1. If you want Bazaar 2, use the bzr port" I should have read the long description
[18:30] <GaryvdM> anddam: I just checked, there was never a bzr 1.42
[18:30] <GaryvdM> *1.4,2
[18:30] <GaryvdM> *1.4.2
[18:32] <GaryvdM> anddam: I would recommend this: http://wiki.bazaar.canonical.com/MacOSXDownloads
[18:32] <anddam> no, I'm using mp
[18:33] <anddam> I mean I don't need further instructions, I just wrongly assumed bazaar was the right port
[18:33] <GaryvdM> anddam: ok
[18:33] <anddam> thanks anyway
[18:34] <anddam> is baz distributed?
[18:34] <GaryvdM> Yes, but really really old
[18:34] <GaryvdM> and unsupported
[18:35] <GaryvdM> anddam: The reason I'm recomending our download in that our download has everything for the gui. Installing the gui via mp will be difficult. (not sure whether this is important to you)
[18:36] <anddam> no really
[18:36] <GaryvdM> ok
[18:36] <anddam> I'd not be using mp in that case
[18:36] <GaryvdM> anddam: it is also very easy to run bzr from source.
[18:37] <anddam> run==build?
[18:37] <luks> no, run :)
[18:37] <anddam> scripted?
[18:37] <GaryvdM> python
[18:37] <luks> bzr is written in python
[18:37] <anddam> I mean it's interpreted rather than compiled?
[18:37] <nDuff> anddam, Python with C modules for speed
[18:37] <anddam> I see
[18:37] <anddam> I'll let mp do the magic tho', that's the whole poing in a package manager
[18:38] <anddam> s/poing/point/
[18:38] <GaryvdM> I see
[18:38]  * anddam builds dependencies, that's the whole point in using a package manager… 
[18:38]  * anddam sighs
[18:39] <anddam> I can't recall if I already asked, is bzr suited to work with mercurial repository?
[18:39] <luks> hg is probably better at that :)
[18:39] <anddam> I read the migration doc and I saw it has hg module but can I "pull" (is that the right action?) into an hg repo?
[18:39] <nDuff> anddam, there's a plugin: http://wiki.bazaar.canonical.com/BzrForeignBranches/Mercurial
[18:40] <luks> I'm not sure if bzr-hg supports at least dpush now
[18:40] <anddam> luks: yes but the "Why is Bazaar better" page got me with the disk usage comparison
[18:41] <GaryvdM> anddam: Yes, with bzr-hg you would just do branch/pull
[18:41] <anddam> svn's commit being "push"?
[18:42] <GaryvdM> anddam: svn commit ≈ bzr commit + push
[18:42] <GaryvdM> anddam: or bzr push in a bond branch (aka checkout)
[18:42] <GaryvdM> * bzr commit in a bond branch (aka checkout)
[18:43] <nDuff> s/bond/bound/
[18:43] <anddam> so my question was meant with "push" rather than "pull", namely I'm thinking about a Google hosted project
[18:43] <Raim> anddam: use bzr instead of bazaar port with macports
[18:43] <anddam> Raim: lol, hi man
[18:43] <Raim> hi anddam :)
[18:43] <anddam> I just wrote that myself a few lines up
[18:43] <luks> anddam: bzr-hg doesn't support that
[18:44] <luks> if it's your project, then you can use Google's svn with bzr-svn
[18:44] <anddam> Raim: I'm looking at the portfile now, namely the tcl split you made
[18:44] <luks> that will work jut fine
[18:44] <anddam> luks: this way loosing the distribution in DVCS, isn't it?
[18:45] <luks> anddam: depends on how you look at it
[18:45] <luks> of course if you want the best DVCS experience, then I'd suggest you to use bzr natively
[18:46] <anddam> the point is I'm DVCS agnostic atm and I'd like with one tool rather than several, from what I can read I like bazaar more but I'd like to work easily with G Project Hosting
[18:46] <luks> is hosting the code launchpad a problem?
[18:47] <GaryvdM> anddam: bzr-svn support is excellent
[18:47] <anddam> luks: I don't think it'd be
[18:47] <GaryvdM> anddam: bzr-hg is in development...
[18:47] <anddam> Raim: it's installing the whole py26 deps…
[18:47] <anddam> Raim: from what I heard on chan the main issue in adopting a DVCS rather than svn for trunk is GPL, right?
[18:48] <anddam> Raim: did you follow (one of) the last discussion?
[18:48] <Raim> anddam: kind of, MacPorts is using svn as that is what is being shipped with OS X
[18:49] <anddam> to build a mp system out of the box?
[18:49] <Raim> yep
[18:49] <anddam> Raim: I mean one could provide a bootstrap pkg
[18:49] <anddam> Raim:  I kinda ends messing up repo with svn, like the infamous email address commit
[18:50] <anddam> fetching bzr now :-)
[18:50] <Raim> anddam: unfortunately bzr-svn lacks keyword expansion for svn:keywords, so if you plan to use that it will not work correctly
[18:51] <anddam> Raim: well, no, I'm trying to switch to distributed, otherwise I can stay with plain svn client
[19:18] <jelmer> Raim: keywords work but you have to do some additional local configuration
[19:20] <KhaZ> Hello: I'm wondering if this is a crazy idea, and how one would go about implementing it.  We use Perforce at work (sadly; particularly at times like today when it's down), and I'd like to work around it's outages.  I was thinking it would be neat to 'double-version' our source code - once in Perforce, and once in bzr.  Ideally I'd work in bzr, have my commits to bzr 'queued' to be put into perfroce at some
[19:20] <KhaZ> time, and then be able to commit them all to Perforce some time in the future.
[19:21] <KhaZ> Does this sound like something automatable, or does this sound purely crazy? :)
[19:22] <jelmer> KhaZ, it sounds reasonable
[19:22] <jelmer> I think somebody (Matt?) once started on a plugin for this sort of thing named bzr-p4
[19:22] <jelmer> I'm not sure what its status is though
[19:22] <rubbs> I"m not entirely sure, and I'm definately not an expert (I'm better at helping complete newbies), but you could automatically pump out your changes as patches and then commit them with preforce
[19:23] <rubbs> perforce*
[19:25] <KhaZ> Hrmm, interesting.  I'll give that some thought.
[19:26] <rubbs> I'd checkout jelmer's suggestion of bzr-p4 first.
[19:26] <nDuff> bzr-p4 looked subject to severe bitrot last I checked
[19:26] <nDuff> which was within the last week
[19:26] <rubbs> ah
[19:36] <KhaZ> Still, might be a good launching point - I've never written a plugin for bzr before.
[19:36] <KhaZ> Would be nice to add a true 'disconnected mode' for Perforce.
[21:15] <poolie> hi jam?
=== salgado is now known as salgado-afk
[21:46] <poolie> hello lifeless
[21:53] <lifeless> hi poolie
[21:54] <poolie> hey
[21:57] <lifeless> rsync is having trouble with my BackupPC llink farm
[22:01] <poolie> too many hardlinks?
[22:10] <lifeless> yeah
[22:10] <lifeless> that and an assertion in the rsync code base
[22:11] <lifeless> I hope the too many links issue is ext3 vs btrfs
[22:22] <lifeless> poolie: http://permalink.gmane.org/gmane.comp.file-systems.btrfs/4614
[22:23] <poolie> istr rsync has some kind of hardcoded limit on the number of links
[22:23] <poolie> or did ten years ago
[22:28] <lifeless> thats likely too
[22:28] <lifeless> this is rsync3
[22:28] <lifeless> which is pretty nice
[22:30] <bob2> are you using btrfs on your desktop or backup server?
[22:31] <lifeless> backup USB2 connected drive on my home server
[22:31] <lifeless> main drive is ext3
[22:31] <lifeless> actually, main drive is raid 5 on 4x320GB disks w/ext3.
[22:42] <lifeless> poolie: I've discussed this with dbarth; I'm going to be timeshifted to evenings this week to help dx with a sprint
[22:55] <poolie> k
[23:04] <jelmer> hey lifeles, poolie
[23:04] <jelmer> does either of you know if the config editing code changed recently?
[23:04] <poolie> hi
[23:04] <jelmer> I'm suddenly finding my mock Config object needs to implement _get_config_editor()
[23:04] <poolie> not as far as i know
[23:04] <lifeless> I don't remember anything
[23:04] <poolie> check the log i guess
[23:05] <poolie> log -p|less may help
[23:05] <poolie> or bzr search
[23:05] <jelmer> lifeless, that's worrying :-)
[23:18] <jelmer> poolie: I'll have a look
[23:20] <poolie> jelmer, iirc you're going to uds with the soyuz team?
[23:26] <jelmer> poolie, yes
[23:33] <spiv> Good morning.