=== mark1 is now known as markh
[01:01] <speakman> hi folks!
[01:03] <speakman> I'm having this common webdevelop issue with bzr. A few users to commit to a central repo. There's still no authentication within Smart Server right?
[01:03] <Peng_> Right.
[01:05] <speakman> Another issue with this web developing (actually php), is there any "proper" way to get the remote working tree updated when committing?
[01:05] <Peng_> speakman: There's a bzr-push-and-update plugin. If you're using checkouts, I dunno if it would work.
[01:07] <speakman> But then it will be required on all clients?
[01:08] <speakman> Isn't there a post-commit hook to be used on the server side for such things?
[01:08] <Peng_> ...Possibly.
[01:09] <speakman> How about the bzr+http approach, will it work with simple htdigest for auth?
[01:13] <Peng_> I'm sure you could set up http auth in your web server.
[01:14] <speakman> And is there any reason not to use the mod_wsgi with bazaar? No howtos seems to cover that setup.
[01:14] <Peng_> It might be difficult to secure though..
[01:14] <Peng_> speakman: Eh. It's probably just that nobody's written the documentation.
[01:16] <speakman> Easy as that. :)
=== asac_ is now known as asac
=== joshuaAF1blount is now known as jblount
[03:14] <Pete_> Hello.  Can anyone help me get loggerhead to work on apache?  I have it running on the localhost:80, but cant get it to work when someone types in : http://bzr.website.com
[03:14] <james_w> are you using ReverseProxyPass?
[03:14] <james_w> oh
[03:14] <james_w> you're using mod_python and apache directly?
[03:15] <Pete_> i want to just use apache
[03:15] <james_w> ok
[03:15] <Pete_> or do w/e is needed to just get it working
[03:15] <Pete_> lol
[03:15] <james_w> can you access it at "<your ip>:80" ?
[03:15] <Pete_> no
[03:15] <Pete_> but.
[03:15] <Pete_> if i go terminal and use lynx, i can
[03:16] <james_w> using the IP address?
[03:16] <james_w> or using the domain name?
[03:16] <Pete_> one sec ill put it in pastebin
[03:17] <Pete_> http://pastebin.com/d676f76c7
[03:17] <james_w> oh, 8080?
[03:17] <Pete_> ya, sorry it was a typo earlier
[03:17] <james_w> you said 80 before
[03:18] <james_w> ok, so it sounds like you are not using mod_python, but running loggerhead by another method, is that correct?
[03:18] <Pete_> i suppose
[03:18] <Pete_> what is the mod_python way  of doing it?
[03:18] <james_w> I'm not positive you can
[03:18] <Pete_> i was just running serve_branches from the directory i wanted to start the branches in
[03:18] <james_w> I assume you can
[03:18] <james_w> ah, ok
[03:18] <Pete_> oh ic
[03:19] <james_w> if you want to keep using serve-branches then look for a ReverseProxyPass tutorial
[03:19] <james_w> I think that's what you need to use
[03:19] <Pete_> sorry, im also learning the whole server apache thing. :( sorry for the lack of knowledge there
[03:20] <Pete_> earlier
[03:20] <Pete_> i was able to get it to work by using a Proxy and mod_rewrite
[03:21] <Pete_> but it only worked halfway, half of the links would show up as localhost:8080/filename and not work....
[03:21] <james_w> hmm
[03:21] <Pete_> http://pastebin.com/m6f350e95
[03:21] <Pete_> that is the apache config file i used
[03:21] <james_w> there may be a serve-branches option to change what those links are generated as
[03:21] <Peng_> Pete_'s nick is really confusing me. D:
[03:25] <Pete_> --host=USER_HOST      Host Loggerhead should listen on.
[03:25] <Pete_> would that be my ip or the url?
[03:30] <Pete_> Ugh, and now when i try using the --host=bzr.petederemer.com option i just get an error: socket.error: (98, 'Address already in use')
[03:41] <nDuff> Pete_, I don't have the loggerhead source handy, but by convention the default should be 0.0.0.0, ie. to listen on all interfaces.
[03:41] <nDuff> Pete_, it sounds, though, like something else may have the port open.
[03:44] <Pete_> nDuff: i got it working. now i just have to figure out how to get http://173.45.224.235:8080
[03:44] <Pete_> to show up as bzr.petederemer.com
[03:45] <nDuff> Pete_, is doing a redirect from http://bzr.petederemer.com/ http://bzr.petederemer.com:8080/ unreasonable?
[03:45] <nDuff> (if so, you certainly do have other options, but it's the easy/quick/cheap approach)
[03:45] <Pete_> nDuff: i would like to keep it cleaner
[03:46] <Pete_> i think my iptables are blocking port 8080, so im opening those up right now
[03:46] <nDuff> ...hmm, it's using paste... doesn't Paste support WSGI?
[03:46] <Pete_> what is wsgi?
[03:46] <nDuff> a protocol for communication between webservers (say, Apache) and webapps.
[03:46] <Pete_> well hot damn
[03:47] <Pete_> any tips on using it? or should i just hit the website straight up
[03:48] <nDuff> ...actually, looking at the README, looks like the preferred deployment approach is still mod_proxy...
[03:48] <nDuff> ...given that that's what the devs prefer, I'm surprised you were having trouble with redirects.
[03:48] <Pete_> what readme are you seeing this in? the one i looked at was like 10 lines long :(
[03:49] <nDuff> you *were* using a ProxyPassReverse directive, right?
[03:49] <Pete_> http://pastebin.com/m6f350e95
[03:49] <Pete_> i tried that
[03:49] <Pete_> but had very very limited succes
[03:49] <Pete_> but had very very limited success
[03:49] <nDuff> err, the README.txt in the loggerhead source directory I'm in is 117 lines per "wc -l".
[03:50] <nDuff> ...and has an explicit section "SERVING LOGGERHEAD FROM BEHIND APACHE"
[03:50] <Pete_> err, that file did NOT look lliek the one i read
[03:50] <Pete_> :(
[03:51] <nDuff> http://bazaar.launchpad.net/%7Eloggerhead-team/loggerhead/1.10/annotate/head%3A/README.txt
[03:51] <Pete_> Now im really mad... the different documentation wasted several hours of my time :(
[03:51] <Pete_> i think i read the readme in the debian package
[03:51] <nDuff> ahhh
[03:52] <Pete_> apparently, a very silly thing to do. haha
[03:52] <Pete_> ok thanks ill read and let you know if  i have any issues
[03:52] <Pete_> thanks!
[04:07] <Pete_> nDuff: http://pastebin.com/d4eea86df  i am now getting access forbidden errors :(
[04:08] <nDuff> Pete_, do you have any Allow clause applying to that Location?
[04:09] <Pete_> nDuff: Not that I know of.  but i dont have a deny either?
[04:10] <nDuff> Pete_, right -- so you're relying on default behavior. Smart (from a security perspective) default behavior for any server is to not allow access to anything unless it's been explicitly granted.
[04:10] <Pete_> nDuff: ah, didnt know that, will allow it right now
[04:11] <nDuff> Pete_, ...that said, the default Order deny,allow makes things which don't match either directive permitted (while Order allow,deny does the reverse)
[04:12] <Pete_> nDuff: the location here is the location relative to the site right?
[04:12] <Pete_> nDuff: or is it the location of the folder im trying to serve?
[04:14] <nDuff> The former.
[04:14] <nDuff> If you were talking about a location on-disk, you'd be using a Directory section, not a Location one.
[04:18] <Pete_> http://pastebin.com/d45f286ca
[04:18] <Pete_> still having issues :(
[04:24] <nDuff> Pete_, erm, I meant for you to put those directives in the Location clause, not in a separate Directory clause
[04:24] <nDuff> Pete_, my apologies if I was unclear.
[04:24] <Pete_> nDuff: oh
[04:26] <nDuff> ...my comment about "if you were talking about a location on-disk" was explaining *why* you're putting them in a Location clause rather than a Directory one, as when you're proxying something from a local service, it's not coming off a disk as far as Apache is concerned.
[04:27] <Pete_> ah, i see
[04:28] <Pete_> http://bzr.petederemer.com/ , still getting getting a service temperary unavailable
[04:28] <Pete_> rather now im getting
[04:28] <Pete_> nm just took a minute
[04:28] <Pete_> :-D
[04:29] <Pete_> nDuff: one issue. now if i click on a link it takes me to another subdomain for some reason :(
[04:29] <Pete_> http://pastebin.com/d55c13e3a
[04:30] <nDuff> Pete_, using it from here it looks fine.
[04:30] <nDuff> Pete_, can you describe very specifically how to reproduce?
[04:31] <Pete_> nDuff: click on the cs350hw link
[04:31] <Pete_> nDuff: it takes you to afrolegs.com
[04:31] <nDuff> Pete_, it brings me to http://bzr.petederemer.com/cs350hw/files... I don't imagine you'd be using an ancient browser that doesn't provide a Host: header?
[04:32] <Pete_> nDuff: Google Chrome...
[04:32] <nDuff> well, that's certainly not the problem, then...
[04:32] <nDuff> ...hang on, I'll try from my Windows VM.
[04:33] <nDuff> nope, I can't reproduce it here using Chrome either.
[04:33] <Pete_> nDuff: wtf?!
[04:33] <Pete_> ie does it fine
[04:34]  * nDuff remembers Chrome having a take-a-screenshot feature, but doesn't remember where it is.
[04:35] <Pete_> its chrome 2.0
[04:35] <Pete_> i dunno if that matters...
[04:35] <Pete_> but still thats just gay
[04:37] <nDuff> Pete_, that *is* bizarre. Can you reproduce it with other browsers?
[04:37] <Pete_> nDuff: Nope
[04:38] <Pete_> well this is REALLY strange
[04:38] <Pete_> but clearing thee cache worked
[04:38] <Pete_> im not sure why it was cached like that to begin with though...
[04:38] <Pete_> lol
[04:38] <Pete_> thanks sooo  much!
[04:38] <nDuff> Glad to help.
[04:39] <Pete_> nDuff: i still cant believe that i wasted a week because the debain readme was incomplete
=== BasicPRO is now known as BasicOSX
=== thekorn_ is now known as thekorn
[16:06] <wekt> What does LP error message mean?
[16:06] <wekt> "Series head on sane-backends has no branch associated with it"
[16:17] <avuton> How do I check the version of a remote repository without checking it out
[16:21] <phanatic> avuton: bzr revno remote_address
[16:21] <avuton> Aah, thanks :)
[16:22] <avuton> hrm, that isn't working on http://bazaar.launchpad.net/~lottanzb/lottanzb/
[16:22] <avuton> Says it's not a branch
[16:23] <luks> because it's not a branch
[16:24] <luks> the currect url is bazaar.lp.net/~USERNAME/PROJECTNAME/BRANCHNAME
[16:24] <luks> you are missing the last part
[16:24] <avuton> Oh, sorry, I'm so infamiliar with bzr
[16:25] <luks> no, this is about launchpad, not bzr
[16:45] <sender> hi, can anyone help me installing bzr on my debian box?
[16:46] <sender> aptitude show bzr gives me: Version: 0.11-1.1 - is this up to date?
[16:47] <phanatic> sender: that's a very old version
[16:48] <sender> phanatic, tx.. thought so :/
[16:48] <phanatic> sender: this page should help you a bit: http://bazaar-vcs.org/DistroDownloads
[16:49] <sender> tx, seen that.. trying to use backports.org now, it gives me a  GPG error ..
[16:49] <sender> "The following signatures couldn't be verified because the public key is not available: NO_PUBKEY EA8E8B2116BA136C"
[16:50] <sender> rather different problem.. any idea how do i fix this? :)
[16:51] <phanatic> you should be able to override that somehow. not sure how :/
[16:53] <sender> phanatic, this did it: apt-get install debian-backports-keyring.. thanx! off now exploring bzr :)
[16:54] <phanatic> sender: that's great, have fun! :)
[16:56] <sender> phanatic, shoot i'm now at Bazaar (bzr) 1.5 .. am i missing out with this version (1.10 most recent, right?) is this advanced enough or do you reckon that i compile from source?
[16:57] <phanatic> sender: i think 1.5 is still pretty old, so my advice would be to do a manual install (you can remove it later pretty easily)
[17:00] <sender> phanatic, ok I'll try this, will be a first
=== asac_ is now known as asac
=== beaumonta is now known as abeaumont
[18:45] <Sonderblade> is there something like git-svn for bzr that works really well?
[18:47] <LarstiQ> Sonderblade: bzr-svn
[19:03] <davidstrauss> lol
[19:43] <asabil> Sonderblade: and it works better :p
[19:44] <asabil> (it is only annoying sometimes when the bzr and bzr-svn versions don't match)
[19:45] <Sonderblade> good to know, git-svn is very nice and if bzr-svn is even better, then i'm happy
=== verterok is now known as verterok_
[20:13] <thumper> morning
[20:15] <lifeless> hi
[20:45] <Sonderblade> bzr diff is kind of slow, i thought it only operated on local data?
[20:46] <RAOF> It does, yes. It needs to hit the disc a bit, though.
[20:47] <RAOF> lifeless: What sort of debug information would be useful for a "bzr search displayed a bunch of binary data and messed up my terminal character map" bug?
[20:47] <Sonderblade> a lot it seems
[20:48] <Sonderblade> in git, diff, log and blame are instanteous but take a long time in bzr
[20:49] <RAOF> Some of that is the different type of output; bzr log shows the topological log, rather than the list of commits in chronological order.
[20:49] <RAOF> Some of that is python overhead.
[20:49] <RAOF> Perhaps some of it is an outdated repository format, or old bzr?  I don't find the diff time obnoxious.
[20:50] <Lo-lan-do> Speaking of which (I havent whined about it for quite some time)... is there work done on the reduction of the startup time?
[20:51] <Sonderblade> on the same repos, "bzr log | less" takes 7 seconds but "git log" less than one second
[20:51] <ronny> hmm
[20:51] <RAOF> Right. "git log" doesn't have to walk the repository in the same way "bzr log" does.
[20:52] <RAOF> In return, you get what is IMO more useful output.
[20:52] <ronny> LarstiQ: i had to change the abstraction, now i need repo<>workingset<>branch<>workdir
[20:53] <Sonderblade> RAOF: in this case it is not the "walk" that is taking the time, both repos are just one long svn HEAD branch
[20:55] <RAOF> We have, at this point, reached the limit of my bzr-internals knowledge.  Are they both native repositories, or is it an bzr-svn/git-svn thingy?  That might make a difference.  Maybe.
[20:57] <Sonderblade> one is bzr-svn, the other is git-svn
[20:57] <lifeless> so I have a ew queued questions I see :P
[20:57] <lifeless> *few* queued...
[20:58] <lifeless> Sonderblade: bzr diff slow - should be < 1 second across the board for working tree diffs; for historic diffs it will be slower (potentially radically slower) though that is being worked on
[20:58] <lifeless> RAOF: pointer to a repo and a search that does it would be best
[20:58] <lifeless> RAOF: patch that fixes it better still
[20:59] <Sonderblade> lifeless: historical diffs?
[21:00] <lifeless> Sonderblade: diff -r x..y
[21:01] <lifeless> Lo-lan-do: not much happening on startup reduction at the moment
[21:01] <lifeless> Lo-lan-do: various folk have done analysis of what makes it slow; seems to be most accurately summarised as 'python;
[21:03] <Lo-lan-do> :-/
[21:04] <lifeless> Sonderblade: gits raw extraction speed is extremely fast; I'm actually analysing exactly what at the moment; I suspect its a combination of less copies (e.g. use of mmap) and the IO ordering of packs which is pretty good (we tend to have more recent revisions at the end of the pack not the fount; this is a limitation in current storage not a design goal)
[21:04] <lifeless> Sonderblade: so when you say diff is kind of slow; are you talking milliseconds or seconds or ... ?
[21:05] <lifeless> Lo-lan-do: more specifically it appears to scale with codebase size
[21:05] <LarstiQ> lifeless: a bit higher up:21:51:07 < Sonderblade> on the same repos, "bzr log | less" takes 7 seconds but "git log" less than one second
[21:05] <Sonderblade> lifeless: i'm talking 7-10 seconds for bzr blame on a single file
[21:05] <lifeless> LarstiQ: thats log not diff
[21:05] <lifeless> Sonderblade: you said "
[21:05] <lifeless> 07:43 < Sonderblade> bzr diff is kind of slow, i thought it only operated on local data?
[21:05] <lifeless> "
[21:05] <lifeless> Sonderblade: did you mean blame?
[21:05] <Lo-lan-do> lifeless: I see.  There was talk of only loading the required modules, on demand, though.  Did that happen?  Would it help much?
[21:06] <LarstiQ> lifeless: though, I'm sorry, I'm not in the best of states atm
[21:06] <Sonderblade> lifeless: ALL operations are that slow, bzr blame, diff, log..
[21:06] <Sonderblade> they are "subversion slow"
[21:06] <lifeless> Sonderblade: thats unusual; blame *should* be pretty fast, status, diff, commit should be near instant
[21:07] <Jc2k> Sonderblade: did you bzr co or bzr branch from subversion?
[21:07] <lifeless> Sonderblade: could you pastebin 'bzr info' - feel free to sanitise urls
[21:07] <Sonderblade> Jc2k: bzr checkout
[21:07] <Sonderblade> i was told bzr-svn worked very well:)
[21:08] <lifeless> Sonderblade: oh, in which case its svn slow cause its still doing svn on every operation
[21:08] <lifeless> Sonderblade: because you've told it you want to act like a svn client
[21:08] <lifeless> Sonderblade: run 'bzr unbind'
[21:08] <Lo-lan-do> I don't think bzr blame runs svn.
[21:08] <Lo-lan-do> Nor do bzr diff or log.
[21:09] <lifeless> Lo-lan-do: elimination of variables
[21:09] <Lo-lan-do> (Or maybe I've got a secret wireless card in my laptop that works even when in a train in a tunnel)
[21:09] <lifeless> Lo-lan-do: (and there is a release that does, it looks up the master ranch for the branch nickname)
[21:09] <Sonderblade> lifeless: i did checkout all revisions of the svn repos which took a really long time so i don't understand why it should act like svn?
[21:10] <Sonderblade> this is my bzr info: http://pastebin.com/m7ef9bc53
[21:11] <sportman1280> hello again.  How do i import a svn repository into bzr 11RC1?
=== TheMuso_ is now known as TheMuso
[21:13] <Lo-lan-do> sportman1280: bzr branch svn://foobar/trunk
[21:13] <sportman1280> Lo-lan-do:  what if it is a https page that requires login?
[21:14] <lifeless> Sonderblade: nothing alarming there
[21:14] <lifeless> Sonderblade: did you try unbinding?
[21:14] <lifeless> Sonderblade: 'time bzr st' would be useful for me
[21:14] <lifeless> Sonderblade: also are you running bzr packaged or from source?
[21:14] <sportman1280> Lo-lan-do: and i get a unsupported protocol for "svn://websitename.com"
[21:14] <Lo-lan-do> Then first do an "svn ls https://foobar/", then "bzr branch https://foobar/trunk" will reuse your SVN credentials.
[21:14] <Sonderblade> lifeless: bzr unbind did not help
[21:15] <Sonderblade> lifeless: time bzr st -> 0m0.337s
[21:15] <lifeless> what does bzr inventory | wc -l output
[21:15] <Sonderblade> lifeless: bzr --version -> 1.3.1 from ubuntu hardy
[21:16] <Sonderblade> lifeless: bzr inventory | wc -l -> 537
[21:17] <sportman1280> Lo-lan-do i still get an  unknown protocol for url
[21:17] <Lo-lan-do> Try svn+https then
[21:17] <Lo-lan-do> You did install bzr-svn, right?
[21:17] <sportman1280> Lo-lan-do i still get an  unsupported protocol for url
[21:18] <sportman1280> bzr-svn wont install with 1.11rc1
[21:18] <lifeless> Sonderblade: https://launchpad.net/~bzr/+archive
[21:18] <sportman1280> bzr-svn says it requires a build of bzr less than 1.8
[21:18] <lifeless> Sonderblade: has bzr 1.10 in it
[21:19] <lifeless> Sonderblade: this has had about 8 months of folk optimising it :P
[21:19] <Sonderblade> lifeless: yes but i'm not using latest ubuntu, 1.3.1 can't be so old can it?
[21:20] <lifeless> bzr 1.3.1 2008-04-09
[21:21] <Sonderblade> bzr can't be very mature yet if an 8 months old version is to old
[21:21] <lifeless> sportman1280: there are newer releases of bzr-svn too
[21:22] <Lo-lan-do> sportman1280: I guess you need to find a more recent version.  Like 0.5.
[21:22] <lifeless> Sonderblade: uhm its not too old; its had performance improvements and you seem to have a performance issue
[21:22] <sportman1280> Lifeless: where can i find a built one?
[21:22] <Lo-lan-do> Built for what?
[21:22] <sportman1280> Lo-lan-do: i386
[21:22] <lifeless> Sonderblade: we can sit down and analyse why you're having a performance issue, and hope its not one thats fixed; -or- we could as a start point get the current release
[21:23] <Lo-lan-do> What distribution?
[21:23] <sportman1280> Lo-lan-do: the ppa beta doesnt have it
[21:23] <sportman1280> Lo-lan-do: ubuntu intrepid
[21:23] <Lo-lan-do> Try the one in Debian experimental.
[21:23] <Sonderblade> lifeless: ok, i'll recompile it from source and will try again
[21:24] <Lo-lan-do> (Or just grab it from upstream into your ~/.bazaar/plugins, there's no compiling required)
[21:24] <lifeless> Sonderblade: huh! you don't need to do that
[21:24] <sportman1280> Lo-lan-do: oh i didnt realize there was no compiling
[21:24] <lifeless> Sonderblade: just add that PPA to your apt sources and ubuntu will track it
[21:25] <lifeless> sportman1280: intrepid has bzr-svn 0.4.13-2;
[21:26] <sportman1280> lifeless: which doesnt work with 1.10
[21:26] <lifeless> sportman1280: cause intrepid shipped 1.6.1
[21:26] <sportman1280> lifeless: it apparently only likes bzr 1.8 or less
[21:26] <sportman1280> lifeless: i was getting an error when importing
[21:27] <sportman1280> lifeless: it was an invalid character '\'
[21:27] <sportman1280> lifeless: and would terminate
[21:27] <lifeless> sportman1280: how did you install 1.10 ?
[21:30] <sportman1280> lifeless: i installed the 1.11rc1 from the bzr beta ppa
[21:31] <lifeless> sportman1280: ah, we don't build bzr-svn in the beta ppa
[21:31] <lifeless> perhaps we should
[21:31] <sportman1280> Lo-lan-do: the plugin isnt being detected when i put it in the plugins file
[21:31] <lifeless> sportman1280: anyhow, I would just use the released ppa, which has bzr-svn included in it
[21:31] <sportman1280> lifeless: yes please do. they conflict and it actually removes any bzr-svn u have installed
[21:32] <sportman1280> lifeless: but that bzr-svn doesnt work
[21:32] <sportman1280> lifeless: because it conflicts with bzr 1.10.
[21:32] <lifeless> sportman1280: in the release ppa?
[21:32] <sportman1280> or anything higher than 1.8
[21:32] <sportman1280> https://launchpad.net/~bzr/+archive
[21:32] <lifeless> sportman1280: yes
[21:33] <sportman1280> yep it conflicts
[21:34] <lifeless> +COMPATIBLE_BZR_VERSIONS = [(1, 10, 0)]
[21:34] <lifeless>  
[21:34] <lifeless> sportman1280: looks fine to me; I would wager breakfast that you were using the beta ppa which *will conflict*
[21:35] <sportman1280> lifeless: gah. apparently i cant read
[21:39] <sportman1280> lifeless:  Thanks lifeless and Lo-lan-do. Hopefully i dont get an error this time importing it
[21:40] <sportman1280> lifeless: cuz then i really have no idea what to do haha
[21:40] <lifeless> sportman1280: well if an import fails file a bug please
[21:40] <sportman1280> there was a bug filed when i looked
[21:40] <sportman1280> but its been there since 2007 i think so theres little hope of it being fixed
[21:40] <sportman1280> :(
[21:41] <kumi> Hi guys, I pulled an upstream revision into a standalone tree, and the working tree is now 1 revision behind the tip of the branch.
[21:41] <kumi> What command will tell me what files will change when I use "bzr up"? Or which command will give me a diff of the changes that "bzr up" will perform?
[21:42] <Lo-lan-do> kumi: bzr diff ../upstream --new .
[21:43] <kumi> Lo-lan-do: thanks. But what if I don't have read access to upstream?
[21:44] <kumi> (I mistyped earlier, I didn't pull from upstream, I pushed from upstream)
[21:44] <Lo-lan-do> Then you won't be able to update anyway :-)
[21:44] <kumi> hm, maybe I'm not wording things right
[21:45] <Lo-lan-do> Oh, no, it's just that I didn't read you correctly.
[21:45] <lifeless> kumi: bzr diff -r -2..-1 probably
[21:45] <Lo-lan-do> I'm actually wondering about the same question from time to time.
[21:45] <lifeless> we really sould add that ree: revspec
[21:46] <lifeless> *tree:*
[21:46] <kumi> Lo-lan-do: me too, I stumble upon this on occasion
[21:47] <Lo-lan-do> lifeless: Also probably a "upstream:" revspec or so, for use in "bzr missing"
[21:47] <Lo-lan-do> (For bound branches)
[21:48] <Lo-lan-do> Currently, one has to do bzr info, paste the "bound_location" out of that, and do a bzr missing $that_url, which is a bit cumbersome for the equivalent of "svn status -u"
[21:56] <lifeless> Lo-lan-do: parent:
[21:57] <Lo-lan-do> Won't that ignore the new revisions on the parent side since the last update?
[21:58] <lifeless> Lo-lan-do: no, why would it?
[21:58] <igc> morning
[21:58] <Lo-lan-do> I'm probably mixing with ancestor:
[21:59] <Lo-lan-do> And I'm still on 1.5, where there's no parent:, I guess it was added later.
[21:59] <Lo-lan-do> Ah, yeah, the changelog for 1.6 has :parent.
[22:01] <mwhudson> what's the story with my non system bzr trying to load plugins from the system location?
[22:01] <Wurblzap> Hi all -- I'm getting "No such file" on an initial init-repo using sftp. I can mkdir using sftp manually, though... Can you give my some pointers?
[22:02] <spiv> mwhudson: that's the intended behaviour, I think.
[22:03] <mwhudson> even if BZR_PLUGIN_PATH is set?
[22:03] <mwhudson> also, just no
[22:03] <mwhudson> it seems nuts
[22:05] <LarstiQ> it can have it's use, but I agree it's often more annoying
[22:06] <mwhudson> i guess i can avoid it by calling load_plugins with just the paths i want
[22:08] <mwhudson> i wonder if this is why bzr trunk fails tests when run against launchpad trunk...
[22:12] <lifeless> Wurblzap: -Dtransport may add more detail to .bzr.log
[22:13]  * Wurblzap tries
[22:13] <lifeless> Wurblzap: other than that I suggest filing a bug
[22:14] <Wurblzap> -Dtransport literally doesn't add anything (identical logging). Server logs don't shed light either. Unfortunate...
[22:15] <lifeless> Wurblzap: oh, what url are you using
[22:15] <Wurblzap> A private one, lifeless... What details do you need?
[22:16] <lifeless> basically do you have a dedicated directory, or is it in your home dir, if its in your home dir, do you have ~ in the url
[22:16] <lifeless> e.g. sftp://HOST/~/INMYHOMEDIR
[22:16] <lifeless> or sftp://HOST/ABSPATH
[22:16] <lifeless> are the syntax for sftp urls in bzr
[22:16] <Wurblzap> It's in a home dir, lifeless. There's no ~. It's sftp://user@hostname/path/from/homedir/subdir/
[22:17] <lifeless> add /~
[22:17] <Wurblzap> All right, that sounds promising, lifeless
[22:17]  * Wurblzap tries
[22:17] <Wurblzap> All right, no error msg... Let's see what happened on the server
[22:18] <Wurblzap> Yeah, lifeless, this seems to have done the trick. Thanks a lot!
[22:28] <mwhudson> spiv: i think this behaviour means that if you have an old bzrtools installed in the system location, bzr.dev's tests will fail
[22:28] <mwhudson> (because of the blackbox tests of shelve)
[23:37] <markh> do I still have time to make doc changes and brand tbzr and still have them make 1.11?
[23:38] <lifeless> I'd say so
[23:40] <sportman1280> hello.  i am getting the following error. is there anyway around it?bzr: ERROR: Unable to convert Subversion path loud-lintian/trunk/testset/filenames/files/'\  because it contains characters invalid in Bazaar.
[23:52] <sportman1280> Can anyone help?
[23:53] <lifeless> sportman1280: bzr doesn't permit every single possible character in file names
[23:53] <lifeless> in particular '\' isn't permitted
[23:54] <sportman1280> lifeless: how do i get around this if i am trying to import an svn?
[23:54] <lifeless> (as well as \r, \n, \x00)
[23:54] <lifeless> I don't know
[23:54] <lifeless> you could fastexport it and write a awk filter for the fastexport stream
[23:54] <lifeless> bzr can import fastimport streams
[23:56] <sportman1280> what is fastexport a google search isn't really showing me much.  to much random in the results
[23:57] <spiv> You could probably filter that file out of an svnadmin dump of the repo too.
[23:57] <lifeless> http://bazaar-vcs.org/BzrFastImport
[23:57] <spiv> I've no idea if that'd be better/worse or easier/harder than the fastexport option.
[23:58] <lifeless> note that lintian has changed its repo to remove these files now
[23:59] <mwhudson> as is almost always the case
[23:59] <mwhudson> it makes the restrictions a little frustrating for my vcs-imports hat :)