[00:33] <jdub> hey hey
[00:33] <jdub> is there a loggerhead package in a ppa somewhere?
[00:34] <beuno> jdub, not yet. Hopefully I'll get to that next week
[00:34] <beuno> regular setup should be easy enough, if you don't want to wait
[00:34] <jdub> cool, ta :-)
[00:35] <beuno> jdub, all you need installed is python-paste and python-simpletal. Then just either run from source, or run: setup.py install
[00:36] <beuno> (optionally you can install python-sqlite and bzr-search, to get some caching for speed and super-cool searching)
[00:37] <jdub> tasty
[00:37] <jdub> is it pretty anal retentive about code, so i can display branches i don't control?
[00:38] <jdub> s/code/content/
[00:38] <beuno> jdub, define "don't control"
[00:40] <jdub> i'd like to suck down branches from other core developers (so, semi-trusted) and display them in the one place with loggerhead
[00:41] <beuno> without having the branches locally?
[00:41] <beuno> I mean, if you branch them to a dir, just serve that dir with LH, and you're set
[00:42] <jdub> yeah
[00:43] <lifeless> jdub: do checkout --lightweight URL
[00:43] <lifeless> jdub: then it should work
[00:44] <lifeless> jdub: or just branch and maintain a mirror, if you don't want LH sucking data over the network
[00:45] <jdub> and you're pretty confident LH won't screw me for 'trusting' that external data?
[00:45] <jdub> :-)
[00:45] <beuno> jdub, I can't really think of anyway you can get screwed. What do *you* have in mind?
[00:46] <lifeless> jdub: what do you mean by trust
[00:46] <jdub> evil co-developer putting evil in their branch that abuses my LH
[00:46] <lifeless> jdub: we use loggerhead do display 10^4+ branches on launchpad
[00:47] <mwhudson> loggerhead is reasonably careful to not allow xss by displaying user content unmunged, for example
[00:47] <lifeless> if you pull their branch to a local mirror and show that, you'll have precisely the same level of condom loggerhead on launchpad has
[00:48] <jdub> cool
[00:48] <lifeless> jdub: also, if you mirror locally you can setup a bzr-search index of the same branch
[00:48]  * jdub levels up his condom by grinding...
[00:48] <Peng_> If it's remote, will bzr-search still work?
[00:48] <lifeless> Peng_: yes
[00:48] <lifeless> Peng_: if there is a search index on the remote branch
[00:49] <lifeless> Peng_: or if you add a mapping to bzr-search telling it to use a different index for that branch, and then sync the index from time to time
[00:49] <lifeless> -> train to catch
[00:49] <Peng_> Oh, neat.
[00:49] <Peng_> lifeless: Bye. :)
[00:49] <lifeless> jdub: ring me if you want more info
[00:50] <jdub> beuno: congrats, btw :-)
[00:51] <beuno> jdub, not sure exactly for what, but thanks  :_
[00:51] <beuno> :)
[00:51] <jdub> canonical
[00:52] <beuno> ah, yes!  I'm thrilled  :)
[00:52] <cody-somerville> :]
[00:52] <beuno> how was your first week btw, cody-somerville?
[00:53] <cody-somerville> beuno, fairly productive
[00:53] <cody-somerville> beuno, Yours?
[00:54] <beuno> cody-somerville, productive too. Lots of setup stuff though, took away a chunk of time. I'm really enjoying it so far  :)
[00:56]  * cody-somerville nods.
[00:57] <cody-somerville> beuno, I dunno how I'm going to like getting paid monthly instead of biweekly though
[00:57] <Peng_> But do you get beanbag chairs?
[00:57] <beuno> Peng_, I have a beanbag 2m away from me!
[00:57] <cody-somerville> Peng_, I opted for the hoody
[00:58] <cody-somerville> Anyhow, I'm starving! :-) It has been a long day. One of the drawbacks of working from home is that you have to force yourself to actually stop working :P
[01:03] <mwhudson> cody-somerville: yeah, i find that a lot
[01:04]  * cody-somerville forms Canonical Workaholics Anonymous.
[01:32] <rocky> is there a standard way to "disable" a widget like a TitlePane so that all of it's inner contents become greyed out and can't be edited/focused ?
[01:33] <markh> it looks like I can get similar http weirdness with a 'http+urllib' url... - bzr: ERROR: Invalid http response for http://bazaar-vcs.org/....rix: Expected a boundary (squid/2.6.STABLE18:...9C890B4)line, got ''
[01:59] <Leefmc> Question: How do you create a working directory again?
[01:59] <Leefmc> Or is it just a lightweight checkout.. i forget heh
[02:00] <Peng_> "bzr co --lightweight whatever/" to create a lightweight checkout.
[02:01] <Peng_> "bzr co" to create a working tree in the current branch.
[02:02] <Leefmc> I'm confused, a normal checkout creates a working tree?
[02:03] <Peng_> The  checkout command is overloaded.
[02:03] <Peng_> If you have a branch with no working tree, run "bzr co" to create one.
[02:04] <markh> and my hardy box on the same network can give the same error.
[02:08] <Leefmc> Peng_: Huh.. i gatta go, so i'll have to ask again later, because im still confused. I spose i must just be confused on BZR's version of a working tree. Anyway, thanks :)
[02:08] <Leefmc> later
[05:05]  * markh wonders where the release manager is when you need him :)
[05:06] <tacone> if he is european he could be in bed :-)
[05:14] <markh> I think he is in the US which means he would be getting ready for it (or better still, hopefully something even better on a friday night :)
[05:17] <meteoroid> so i have a relatively old bzr and bzr-svn from ubuntu hardy, should i build from source, or are there newer packages from alternative sources?
[05:18] <meteoroid> jelmer says the new bzr-svn has fixes for issues i have, and bzr-webdav says it "only works with 1.6"
[05:47] <lifeless> RAOF: http://piumarta.com/pepsi/?C=M;O=D
[05:47] <fullermd> markh: He's in my TZ, which means that if he's sensible, he's in bed   :p
[07:20] <lifeless> beuno: hi
[07:20] <lifeless> beuno: rant
[07:20] <lifeless> http://bazaar.launchpad.net/~jml/subunit/split-right/revision/51
[07:21] <lifeless> beuno: can't read the entire line, no scroll bar in my window to make things wider; kthxhelp
[07:24] <mwhudson> i think there's a bug report about that already
[07:24] <lifeless> mwhudson: great
[07:25] <lifeless> mwhudson: number plx?
[07:26]  * mwhudson looks
[07:27] <mwhudson> lifeless: maybe not, actually, i think it was just complained about on the mailing list
[07:57] <Jc2k> Zooming out FTW
[07:57] <Jc2k> (i have to zoom in to see the bug :P)
[08:18] <lifeless> Jc2k: :P
[08:18] <lifeless> beuno: ^^^^ more hai!
[08:35] <jml>         self.failUnless(self.suite._tests == [self.case1, self.case2,
[08:35] <jml>                                               self.case3, self.case4] or
[08:35] <jml>                         self.suite._tests == [self.case3, self.case2,
[08:35] <jml>                                               self.case1, self.case4])
[09:25] <Kazade> Hi everyone, I've just recently tried moving a project to Bzr (from GIT) so that I can use it on Windows easily. I've set up putty etc with my private key so I can ssh into my other PCs without entering a password, but no matter what I do, bazaar on windows crashes when I try to clone a project with the error "ERROR: socket.error: Socket is closed". Any ideas?
[11:00] <gour> jelmer: someone from #ghc asks me about bzr's rebase...
[11:01] <gour> jelmer: is it possible to drop shortly in #ghc
[11:14] <gour> lifeless: ping
[11:15] <gour> lifeless: can you visit #ghc shortly to explain about loom plugin, pls.
[11:16] <gour> ...or whoever use it and understand the purpose..
[12:15] <matkor> Hi ! Afer bzr update I see list of modified (M) files but bzr finishes run with bzr: ERROR: [Errno 13] Permission denied ... with not futher info which file permission are denied ... any hint how to find it out ?
[12:17] <matkor> ok, question cancelled - I found out ...
[12:20] <gour> considering http://andrew.puzzling.org/diary/2008/July/29/rebase-criticism i wonder whether rebase is actually needed in bzr?
[12:21] <thumper> gour: no, it isn't normally needed
[12:21]  * thumper going to bed now
[12:45] <LarstiQ> gour: there are usecases for it, but it usually isn't needed.
[13:01] <gour> LarstiQ: thanks
[13:02] <gour> thumper: ^^^
[14:51]  * rocky sniffs and wishes his "bzr commit" to a svn repo didn't make 100 requests and take about 1min
[15:18] <palango> is there an easy way to find out if a file is under bzr control?
[15:20] <Peng_> I guess you could use "bzr st" on it.
[15:20] <Peng_> You could also check for it in "bzr inventory".
[15:20] <Peng_> I'm sure there's a better way, especially through bzrlib.
[15:23] <Necoro> bzr unknowns | grep -q $FILE  && echo $FILE is unknown
[15:23] <Necoro> ;)
[15:23] <uws> Necoro: that doesn't take ignored files into account
[15:24] <Necoro> true
[15:24] <Peng_> "bzr st" does, but it's not very good for a script.
[15:24] <uws> no
[15:24] <Peng_> I mean, usable by a script
[15:24] <uws> "bzr inventory $FILENAME"
[15:25] <Peng_> Oh, I didn't know you could do that. There you go then.
[15:25] <uws> bzr inventory $FILENAME 2>/dev/null && echo versioned || echo not versioned
[15:28] <palango> thanks
[15:40] <Glenjamin> does anyone have experience setting up tortoisebzr on vista?
[15:41] <Glenjamin> i'm getting an error about a missing entry point in pywintypes25.dll
[16:00] <Leefmc> Question: Is there a way to wildcard remove directories? I added a set of directories and my .bzrignore was formatted wrong, so it added a ton of directories and files i dont want.. and removing them by hand would suck
[16:02] <Leefmc> oo, seems wildcard works
[16:02] <Leefmc> :)
[18:01] <gour> how can one delete project on LP?
[19:07] <jbalint> Hi
[19:07] <jbalint> how can i debug this problem on windows? http://rafb.net/p/3vsOH366.html ssh.exe/plink.exe to the host works fine, bzr is working fine on the local and remote hosts
[19:08] <beuno> jbalint, I wonder why you're getting bzr: Command not found.
[19:08] <beuno> is bzr installed on the remote server?
[19:08] <jbalint> yes
[19:09] <LarstiQ> BZR_REMOTE_PATH?
[19:09] <jbalint>  /usr/bin/bzr
[19:09] <beuno> jbalint, also, My Documents/.bzr.log   has more debug info
[19:09] <jbalint> beuno: ah, right. thx, will check it
[19:10] <LarstiQ> jbalint: and I presume that is also where bzr actually resides?
[19:11] <jbalint> yes
[19:11] <jbalint> heres the info from the log http://rafb.net/p/1kzTVr41.html
[19:13] <LarstiQ> jbalint: hmm, no clue of command not found there.
[19:13] <LarstiQ> though, maybe, hmm.
[19:14] <jbalint> i branch its locally on the server, works. i can ssh to the exact from part of the url, and bzr version works fine
[19:15] <LarstiQ> jbalint: it might be a red herring, but how is ssh/plink setup? Key based login?
[19:15] <LarstiQ> jbalint: does the server side ssh log provide any insight?
[19:16] <jbalint> yes, keys, everything is working fine
[19:16]  * LarstiQ glances at the code
[19:21] <LarstiQ> jbalint: It comes back to BZR_REMOTE_PATH.
[19:22] <jbalint> set locally?
[19:22] <LarstiQ> jbalint: what happens if you set that to, say, /bin/true?
[19:22] <LarstiQ> jbalint: yes
[19:22] <jbalint> Dlklemme try
[19:23] <LarstiQ> jbalint: do I read that correctly to say that you previously hadn't set the env var in Windows?
[19:25] <jbalint> heh, /bin/true: Command not found.
[19:25] <LarstiQ> riiight.
[19:25] <LarstiQ> jbalint: is your ssh server jailing you?
[19:26] <jbalint> oi, i dont think so
[19:28] <LarstiQ> if it is a bzr bug, it is a strange one.
[19:28] <jbalint> i can ssh in and it works fine
[19:28] <LarstiQ> but right now I'd like some more confirmation on your environment :)
[19:28] <LarstiQ> rihgt.
[19:29] <LarstiQ> ugh, typo galore tonight.
[19:29] <LarstiQ> jbalint: So what is the difference in the two methods?
[19:29] <jbalint> i dont know
[19:30] <LarstiQ> jbalint: for openssh, you can add config options for particular hosts to ~/.ssh/config. Does plink have something similar?
[19:30] <jbalint> kind of, yeah
[19:30] <LarstiQ> jbalint: can you up the verbosity of the connection?
[19:31] <LarstiQ> jbalint: if not, does BZR_PDB=1 bzr .... drop you in a pdb prompt? (I'm not sure about connection errors)
[19:31] <LarstiQ> then again, if the connection is already closed..
[19:33] <jbalint> it gets into the debugger after the error message
[19:33] <LarstiQ> ok, good
[19:34] <jbalint>  http://rafb.net/p/aCkOkR40.html
[19:38] <LarstiQ> jbalint: while I read more of the code to see what we could do, could you try executing ls or something to discover more about why we get these command not founds?
[19:39] <LarstiQ> jbalint: and, perhaps you could try with paramiko directly?
[19:39] <LarstiQ> jbalint: ie, BZR_SSH=paramiko
[19:45] <LarstiQ> jbalint: btw, which version of bzr is this?
[19:49] <jbalint> 1.4rc1 on the client and 1.5 on the server
[19:49] <jbalint> oh, paramiko wont work because of the plink aliases, bzr: ERROR: Unable to connect to SSH host linksys-route-FWD-nt77; (11001, 'getaddrinfo failed')
[19:50] <jbalint> let me try the ip
[19:52] <LarstiQ> jbalint: ooh, is that alias something over a forwarded ssh connection?
[19:52] <LarstiQ> that might be a clue.
[19:52] <jbalint> http://rafb.net/p/jI6CPD49.html
[19:53] <LarstiQ> jbalint: without BZR_PDB, does it still print 'command not found'?
[19:54] <LarstiQ> jbalint: and in pdb, could you 'print dir(self)'?
[19:54] <jbalint> no, just the bzr: ERROR part
[19:54] <jbalint>  http://rafb.net/p/8a351019.html
[19:54] <LarstiQ> jbalint: that to me sounds like an ssh login problem.
[19:55]  * LarstiQ looks at _request
[19:58] <LarstiQ> jbalint: the .pyc in the traceback sounds like you use the windows binary installer, and not the bare python based one, is that right? (Thinking of editing .py files instead of doing post-mortem debugging)
[19:58] <jbalint> yeah, i can install something else
[20:00] <whitemice> Hi, new (potentially) to bzr.  I am the patch gatekeeper for a large codebase kept in subversion(OpenGroupware).  I'd like to front-end the subverzion repo with bzr so I can allow people to use BZR and push patches, then look at the patches and potentially push them to svn.
[20:00] <whitemice> I've looked at a couple of pages like <http://blog.tplus1.com/index.php/2008/03/22/use-bazaar-and-subversion-together/> but the examples don't quite work. :(
[20:00] <uws> whitemice: install bzr-svn and   "bzr branch svn+ssh://....."
[20:00] <whitemice> Is this a possible setup?
[20:01] <whitemice> I tried "bzr branch svn+http://developer.opengroupware.org/OpenGroupware.org/trunk" and get "Unsupported protocol for url "svn+http://dev..."
[20:02] <jbalint> LarstiQ: i just branched from another box on the lan, and it worked fine, so it must be a client issue
[20:02] <uws> whitemice: do you have bzr-svn?
[20:02] <uws> whitemice: bzr: ERROR: Invalid http response for http://developer.opengroupware.org/OpenGroupware.org/trunk/.bzr/branch-format: Unable to handle http code 401: Authorization Required
[20:02] <uws> whitemice: That's what I get
[20:02] <LarstiQ> jbalint: ok, still would like to find out what is going on
[20:03] <LarstiQ> jbalint: could you go 'up' in pdb until you hit bzrlib/smart/medium.py:protocol_version
[20:03] <jbalint> ok
[20:03] <uws> jelmer: ping
[20:04] <uws> jelmer: recent bzr and bzr-svn complain about svn+ syntax being depreacted
[20:04] <LarstiQ> jbalint: and then a dir(self) again, looking for the backing transport
[20:04] <uws> jelmer: but on the url just mentioned by whitemice it DOES make a difference
[20:04] <jbalint> LarstiQ: print some specific member of self? or dir is enough?
[20:04] <LarstiQ> jbalint: dir is enough for now
[20:05] <LarstiQ> jbalint: what I want to do is call transport.list_dir('.')
[20:05] <LarstiQ> or '/'
[20:05] <whitemice> uws: yep, the repo requires auth.  I must (openSUSE 11) be missing something...  I have all the available bzr packages installed.
[20:05] <LarstiQ> jbalint: but first we have to find the transport
[20:05] <uws> jelmer: in the svn+http:// case it asks for a password, while the http:// one (recommended by the warning message) gives an error about .bzr/branch-format needing auth
[20:05] <uws> whitemice: does "bzr plugins" list svN?
[20:05] <jbalint>  http://rafb.net/p/O1seq080.html
[20:05] <whitemice> uws: no
[20:06] <LarstiQ> whitemice: you cn use http://user@../ to force authentication prompt
[20:06] <uws> whitemice: then you don't have the bzr-svn machinery so bzr won't be able to speak to SVN servers
[20:07] <uws> LarstiQ: wasn't that supposed to be directed to me instead?
[20:07] <LarstiQ> jbalint: you can check self._bzr_remote_path but I don't think it will help us much
[20:07] <jbalint> '/bin/true'
[20:07] <LarstiQ> uws: ehm, maybe, not paying close attention, trying to debug jbalint's issue
[20:07] <uws> LarstiQ: no worries
[20:08] <LarstiQ> jbalint: ah. hmm. Have you tried the combo paramiko//usr/bin/bzr ?
[20:08] <LarstiQ> jbalint: since paramiko /bin/true doesn't print a command not found
[20:09] <jbalint> trying
[20:09] <jbalint> ah, success!!
[20:10] <LarstiQ> jbalint: ok, so: paramiko works with the ip address, but plink does not with the alias?
[20:10] <jbalint> right
[20:10] <LarstiQ> jbalint: how about plink + ip?
[20:11] <LarstiQ> at least we have isolated the problem somewhat
[20:11] <jbalint> hrm, yup works with IP
[20:11] <LarstiQ> next question would be, does it work with other aliases? Or, what makes this one different.
[20:12] <jbalint> let me try a new/clean alias
[20:12] <LarstiQ> If it doesn't work with _any_ alias, then we know which bug to file.
[20:12] <LarstiQ> jbalint: thanks for your patience going through this with me.
[20:13]  * LarstiQ is also simultaneously trying to translate a text.
[20:13] <uws> LarstiQ: from klingon to dutch?
[20:14] <LarstiQ> uws: Finnish to English
[20:14] <uws> LarstiQ: that's roughly the same for me :)
[20:14] <LarstiQ> haha :P
[20:14] <LarstiQ> I admit one of the Finnish 2-seat row athletes had some suspicious crinkles in her forehead.
[20:15] <uws> heh.
[20:15] <LarstiQ> but Klingon I don't speak at all.
[20:16]  * uws afk for a book (in English, not the original ancient greek *phew*)
[20:16] <whitemice> uws: ran through one-click install (nice!) and now have svn plugin
[20:16] <whitemice> uws: "bzr branch svn+http://adam:******@developer.opengroupware.org/OpenGroupware.org/" fails with "subversion/libsvn_ra/ra_loader.c:940: svn_ra_get_log2: Assertion `*path != '/'' failed."
[20:17] <jbalint> LarstiQ: yup same problemw tih a fresh alias :(
[20:17] <uws> whitemice: ask jelmer about that one. I think it's a Subversion bug
[20:17] <LarstiQ> jbalint: alias to the same system?
[20:17] <uws> whitemice: (recompile svn, comment out that assertion might work ;))
[20:18] <uws> whitemice: (or try upgrading svn to 1.5)
[20:18] <LarstiQ> jbalint: I have no clue how plink aliases work. But if you can list a set of steps to reproduce this, a bug report would be nice.
[20:18] <whitemice> uws: I'm using subversion 1.5.0 as is the server (I believe)
[20:18] <LarstiQ> whitemice: what version of bzr-svn are you using?
[20:18] <whitemice> uws: bzr-svn-0.4.10-6.2
[20:18] <LarstiQ> whitemice: svn changed their api in 1.5, but there is a patch for bzr-svn to deal with that.
[20:19] <uws> whitemice, LarstiQ: https://bugs.launchpad.net/bzr-svn/+bug/234010
[20:20] <LarstiQ> whitemice: for a big gun approach, you could try to apply http://ftp.de.debian.org/debian/pool/main/b/bzr-svn/bzr-svn_0.4.10-2.diff.gz (but -6.2 seems to imply some local patching)
[20:27] <jbalint> LarstiQ: add a bug on launchpad?
[20:28] <LarstiQ> jbalint: yes please.
[20:28] <jbalint> ok
[20:29] <LarstiQ> ala "bzr complains about command not found when using plink aliases", listing steps to reproduce and something like your original paste, that would be sweet.
[20:30] <jbalint> ok
[20:30] <jbalint> thanks again
[20:31] <LarstiQ> jbalint: glad to be of help, and thank you for the report.
[20:41] <LarstiQ> whitemice: how easy is it for you to patch the bzr package you are using?