2008/01/21/#bzr.txt

=== doko__ is now known as doko
[02:03] <awmcclain> Is there a reason why bzr for os x doesn't give any feedback during a branch operation?
[02:25] <spiv> PyCon registration is open: http://us.pycon.org/2008/registration/
[02:25] <spiv> awmcclain: no good reason, it's a bug
[02:26] <spiv> awmcclain: (a bug that happens on all platforms)
[02:26] <awmcclain> spiv: When I branched on linux, i got a status bar. At least when branching from svn...
[02:27] <spiv> awmcclain: hmm, I don't think there should be any difference between mac and linux for status bars.
[02:27] <awmcclain> spiv: I'll look into it.
[02:43] <poolie> hello kiko
=== Verterok-laptop is now known as Verterok
[02:52] <kiko> hey poolie
[02:52] <poolie> hi, are you home now?
[04:16] <guillaumebokiau> trying to install with macports, but installing macports itself doesn't seem to work. (port: command not found)
[04:16] <guillaumebokiau> any advice?
[04:19] <poolie> guillaumebokiau, i think you need to install the macports dmg
[04:19] <guillaumebokiau> which i did  :|
[04:19] <poolie> guillaumebokiau, http://www.macports.org/install.php -- then you'll have a port command
[04:19] <poolie> hm
[04:19] <poolie> is it on your $PATH?
[04:19] <poolie> you might need to start a new terminal window?
[04:20] <guillaumebokiau> i even rebooted
[04:20] <guillaumebokiau> what is $path ?
[04:20] <guillaumebokiau> sorry, not very good at those things :/
[04:21] <poolie> please type 'echo $PATH'
[04:21] <poolie> into your terminal
[04:22] <poolie> also, 'echo $0'
[04:22] <poolie> (and press enter)
[04:22] <guillaumebokiau> \Library\Frameworks\Python.framework\Versions\Current\bin:\bin:\sbin:\usr\bin:\usr\sbin
[04:22] <poolie> see, http://guide.macports.org/#installing.shell
[04:23] <poolie> guillaumebokiau, ok, that's probably your problem
[04:23] <poolie> try the instructions there
[04:23] <guillaumebokiau> ok, thx
[04:30] <poolie> guillaumebokiau, did that work?
[04:30] <guillaumebokiau> trying to figure this out :)
[04:31] <poolie> guillaumebokiau, just for curiousity, why are you using macports rather than a dmg?
[04:31] <poolie> did we not have one that would suit?
[04:31] <guillaumebokiau> isn't the dmg for ppc ?
[04:31] <guillaumebokiau> i'm on intel
[04:31] <guillaumebokiau> 10,4
[04:32] <poolie> ok
[04:32] <poolie> yes, that's it
[04:32] <igc> the dmg should be universal I think
[04:32] <poolie> i mean, i wondered if that was the problem
[04:32] <poolie> we need someone with a intel 10.4 mac to rebuild it
[04:33] <guillaumebokiau> i tried with the ppc dmg anyway, but it didn't work
[04:36] <igc> guillaumebokiau: yes, checking again, the 10.4 installer is only ppc only right now sorry
[04:37] <guillaumebokiau> there's no other way to install it?
[04:37] <guillaumebokiau> than with macports?
[04:37] <igc> guillaumebokiau: running straight from source is really easy actually
[04:38] <igc> simply untar and put the directory on your path
[04:38] <igc> you'll need to install Python 2.4 or 2.5 separately
[04:38] <guillaumebokiau> installed that
[04:38] <igc> in not already installed
[04:38] <igc> s/in/if/
[04:38] <guillaumebokiau> er, what's my 'path' ?
[04:39] <guillaumebokiau> (i'm really new to all this)
[04:39] <igc> it's where your terminal looks for commands you type
[04:39] <guillaumebokiau> that : \Library\Frameworks\Python.framework\Versions\Current\bin:\bin:\sbin:\usr\bin:\usr\sbin ?
[04:39] <igc> you can create a file in your home directory to set it each time you start a terminal session
[04:40] <bob2> guillaumebokiau: when you type "bzr", your shell looks in each of the directorys in the PATH variable (:-seperated) for that command
[04:40] <igc> without setting a path just yet, let's test it works
[04:41] <igc> unpack the tar.gz
[04:41] <igc> change to the created directory and ..
[04:41] <igc> type "./bzr version"
[04:42] <guillaumebokiau> have a hard time following. where do I unpack?
[04:42] <igc> anywhere you like ...
[04:42] <igc> If you double-click ...
[04:42] <guillaumebokiau> ok, it's on my desktop :)
[04:43] <igc> cool
[04:43] <igc> in a terminal, change into that directory
[04:43] <igc> try "./bzr version"
[04:45] <igc> guillaumebokiau: did that work?
[04:46] <guillaumebokiau> hang on :)
[04:46] <igc> the directory to change into is something like "Desktop/bzr-1.1"
[04:47] <guillaumebokiau> ok, i'm there
[04:47] <guillaumebokiau> ~/Desktop/bzr-1.1
[04:48] <igc> great
[04:48] <igc> does "./bzr" work?
[04:48] <guillaumebokiau> yes!
[04:48] <igc> sweet
[04:48] <guillaumebokiau> cool
[04:48] <igc> now let's put it on your path
[04:48] <igc> type this:
[04:48] <guillaumebokiau> yeah
[04:49] <igc> PATH=~/Desktop/bzr-1.1:$PATH
[04:49] <igc> then change to another directory and try "bzr"
[04:49] <guillaumebokiau> it"s on the path when i echo it
[04:50] <guillaumebokiau> working
[04:50] <guillaumebokiau> great
[04:50] <igc> ok, let's make it permanent
[04:50] <igc> type "cd" to return to your home directory
[04:50] <guillaumebokiau> done
[04:50] <guillaumebokiau> should i copy the folder there?
[04:51] <igc> with your text editor, create a file called ".prfile" if it doesn't already exist
[04:51] <igc> oops ...
[04:51] <igc> make that .profile
[04:51] <igc> inside it put this text ...
[04:52] <igc> PATH=~/Desktop/bzr-1.1:$PATH
[04:52] <igc> export PATH
[04:52] <igc> (that's it)
[04:52] <igc> to test ...
[04:52] <igc> start another Terminal, in a tab or new Window
[04:52] <igc> type "bzr"
[04:52] <igc> sound ok?
[04:53] <guillaumebokiau> bzr command not found :(
[04:54] <igc> ok ...
[04:54] <igc> type "echo $PATH"
[04:54] <guillaumebokiau> it's not in it
[04:54] <bob2> maybe ~/.bashrc would be better, .profile is only sourced on "login"
[04:54] <igc> good point
[04:55] <bob2> depending on what os x considers login to be
[04:55] <guillaumebokiau> i hav a .bash_profile file
[04:56] <igc> guillaumebokiau: try putting those 2 lines in there then
[04:56] <guillaumebokiau> k
[04:57] <guillaumebokiau> ok, i got it working
[04:57] <guillaumebokiau> pfew
[04:57] <igc> good
[04:57] <guillaumebokiau> i think the correct line was PATH="~/.bzr-1.1:${PATH}"
[04:58] <guillaumebokiau> anyway, i thanks a lot
[04:58] <igc> that ok
[04:58] <igc> either syntax should work
[04:58] <guillaumebokiau> you saved me :)
[04:58] <igc> let us know if you need any more help
[04:58] <igc> no problem
[05:15] <ubotu> New bug: #184733 in bzr "Internal error running bzr selftest for bzr 1.1 on MacOS X 10.5.1" [Undecided,New] https://launchpad.net/bugs/184733
[05:24] <guillaumebokiau> where does bazaar save local copies ?
[05:24] <poolie> into .bzr in the top of the branch
[05:25] <guillaumebokiau> not sure you understood the question :) if I run bzr branch http://bazaar.launchpad.net/~manfre/xpattern/xpattern-1.0 where does it land ?
[05:26] <guillaumebokiau> ah, found
[05:26] <guillaumebokiau> ok, sorry, i'm officially stupid
[05:27] <poolie> right, into your working directory
[05:27] <poolie> no problem, questions are welcome
=== n2diy_ is now known as n2diy
[05:52] <eean> how come it says Unable to load plugin 'bzr_svn_0_4_6' from '/home/ian/.bazaar/plugins' ?
[05:52] <eean> when trying to do a bzr svn clone
[05:52] <eean> I do have the plugin there
[05:53] <spiv> eean: look in ~/.bzr.log
[05:54] <spiv> eean: it should have a more detailed error
[05:54] <eean> ImportError: The Subversion plugin must be installed as bzrlib.plugins.svn not bzrlib.plugins.bzr_svn_0_4_6
[05:54] <eean> so I'll rename the directory I guess
[05:54] <spiv> Right.
[05:55] <eean> there it goes
[05:55] <eean> I hear it doesn't have a memory leak anymore
[05:55] <eean> appears so, so far
[05:56] <eean> hmm
[05:56] <eean> it is increasing in memory steadily :/
[05:57] <spiv> eean: it depends on how new your subversion bindings for python are
[05:57] <eean> aaaah
[05:58] <eean> spiv: do you happen to know how new they have to be?
[05:58] <eean> it is not using crazy amounts of memory like it used to, but its certainly still leaking.
[05:59] <spiv> eean: SVN version of the as of about or week or so ago, IIRC
[05:59] <eean> that would make sense that the subversion bindings would be to blame, I didn't understand how a python script could leak. so its probably something in the c bindings
[05:59] <eean> haha crap
[05:59] <spiv> eean: I think the bzr-svn README links to the relevant bug
[05:59] <eean> riddell blogged before then that it had been fixed in bzr-svn
[05:59] <spiv> eean: I think newer bzr-svn tries to workaround the worst of it, but obviously it can only do so much...
[06:00] <eean> I see
[06:00] <spiv> eean: yeah, see the first part of the FAQ in the bzr-svn source :)
[06:00] <spiv> "bzr-svn can use a lot of memory cloning big branches when older
[06:00] <spiv> versions of the python-subversion bindings are used. This memory leak
[06:00] <spiv> has been fixed in the trunk of Subversion (r28544) and has been
[06:00] <spiv> proposed for inclusion in Subversion 1.4.7.
[06:00] <spiv> "
[06:01] <eean> oh wow, its subversion itself to blame?
[06:01] <spiv> eean: it's possible this fix has already been included hardy, I haven't checked...
[06:01] <eean> well I use gutsy
[06:01] <spiv> eean: right, the python bindings are made by the subversion guys.
[06:01] <eean> ah ok
[06:02] <eean> I can probably make due if I quit and redo every 100000 commits
[06:02] <spiv> Yeah :)
[06:03] <eean> before it would eat all my memory at about 1000
[06:03] <eean> git svn is giving me shit, so this is a good time to switch ;)
[06:04] <spiv> :)
[06:05] <eean> I guess this bug has probably made bzr svn clone pretty robust at restarting
[06:05] <eean> git svn couldn't deal with the svn server timing out, after about 30 minutes of cloning :/
[06:07] <eean> pfft, or not. "bzr: ERROR: sqlite3.OperationalError: SQL logic error or missing database"
[06:08] <spiv> eean: ooh, I haven't seen that one before.
[06:08] <eean> http://pastie.caboo.se/141399
[06:08] <eean> on resuming
[06:08] <spiv> eean: could you file a bug report?  I expect jelmer would like to see that.
[06:08] <eean>  /dev/sda8              44G   42G  1.7M 100% /home
[06:08] <eean> might explain it :)
[06:08] <spiv> eean: ah :)
[06:24] <eean> will I be able to share this checkout with other devs?
[06:24] <eean> I'm kind of worried that it appears to be downloading into ~/.bazaar
[06:25] <spiv> eean: yes
[06:25] <spiv> eean: the stuff in .bazaar is just a cache of the mapping between data the SVN repo and the data converted to bzr
[06:26] <spiv> eean: if the cache is missing or incomplete, bzr-svn will just regenerate it as needed.
[06:30] <eeanm> bah my laptop power cord got loose
[06:30] <eeanm> spiv: did you say anything after 'yes'?
[06:30] <spiv> eeanm: I did, I'll paste:
[06:30] <spiv> 17:25 < spiv> eean: the stuff in .bazaar is just a cache of the mapping between data the SVN repo and the data converted to bzr
[06:30] <spiv> 17:26 < spiv> eean: if the cache is missing or incomplete, bzr-svn will just regenerate it as needed.
[06:31] <eeanm> um that mapping the hard part
[06:31] <eeanm> um *isn't that mapping the hard part
=== jamesh_ is now known as jamesh
[06:35] <eeanm> at least when you're cloning a 2.5k LOC, 200 commit module from a SVN with >700000 commits
[06:35] <bob2> yes, but people branching from your bzr repository won't need it, all they need is in the bzr repository
[06:35] <eeanm> the mapping is a big chunk :|
[06:36] <spiv> eeanm: people only need the mapping if they want to use the SVN repo
[06:36] <eeanm> well yea
[06:37] <spiv> eeanm: I can e.g. take your bzr branch, and make a branch off that, make some commits, and send those back to you, all without touching the upstream SVN repo.
[06:37] <eeanm> suppose I could do it that way, but it'd be nice if folks could push to svn themselves
[06:37] <spiv> They'll need the mapping then.
[06:38] <eeanm> won't be such a big deal once this memory leak fix is availabe
[06:38] <spiv> You can probably share the cache from your .bazaar directory.
[06:39] <eeanm> sounds like bzr has their act more together then git, we try publishing a git svn repo, someone commited something that they didn't end up pushing to SVN, it totally screwed it up
[06:39] <eeanm> would be nice to have feature branching in bzr that could be easily synced with svn
[06:41] <eeanm> (that's what we were trying to do with git)
[06:42] <poolie> eeanm, well, that should work through bzr-svn
[06:57] <eeanm> dang the cache is up to 771 megabytes
[06:58] <poolie> how big is the svn repo?
[06:59] <eeanm> >700000 commits
[06:59] <eeanm> 764k
[06:59] <eeanm> I'm up to 421k
[07:07] <eeanm> sqlite> select count( * ) from changed_path;
[07:07] <eeanm> 3578513
[07:08] <eeanm> probably has something to do with it :)
[07:25] <eean> 529215/764151 woot
[07:54]  * igc dinner
[08:33] <eean> the svn clone is soo near completion, and now its just hanging apparently :)
[08:33] <eean> erm
[08:33] <eean> :(
[08:38] <AfC> eean: Is the Subversion repository in question public? If it is, I'm sure people will be willing to attempt to replicate the problem.
[08:39] <eean> yea, its KDE
[08:39] <eean> ah ok cool
[08:39] <eean> nevermind
[08:39] <eean> now its "analyzing layout"
[08:40] <eean> about 4 minutes there when it didn't respond, I was  worried :)
[08:40] <eean> heh analyzing layout is going to take an hour or so it looks like. at least it doesn't leak memory.
[08:41] <eean> ah no, that's not true, more like 15 minutes I bet, its already 1/7 done
[08:49] <lifeless> hi
[08:55] <lifeless> jelmer: sorry, missed that in backlog, only saw your ^^, what was it ?
[08:56] <AfC> eean: You might want to reply to poolie's RFC about finer grained progress reporting on the mailing list today and note your experience of long waits without feedback
[09:30] <asabil> hi all
[09:32] <lifeless> hi
[10:43] <lifeless> abentley: hi
[10:43] <lifeless> abentley: how is sydney :)
[10:43] <abentley> Quite nice.
[10:43] <abentley> Weather's definitely superior to Toronto's.
[10:44] <abentley> Although pretty similar to Toronto in summertime.
[10:44] <thumper> lifeless: it is hot and muggy
[10:44] <thumper> too hot and too muggy
[10:44] <thumper> but that's from my Dunedin point of view
[10:44] <abentley> This is the first sprint I've ever done without you, though, which is a bit wierd.
[10:45] <abentley> thumper: Toronto in the summertime can be quite muggy, too.  We're on the edge of Lake Ontario, one of the Great Lakes.
[10:45] <lifeless> thumper: unedin is easily as humid; its the heat that does it
[10:49] <muffinresearch> Greetings chaps.
[10:51] <muffinresearch> Are the London sprint dates set in stone? Just it clashes with SXSWi in part.
[10:51] <muffinresearch> Though in any case I would try a pop along on the Monday if I can
[10:52] <thumper> muffinresearch: what is SXSWi?
[10:52] <muffinresearch> south by south west
[10:52] <muffinresearch> http://sxsw.com
[10:56] <muffinresearch> Is coming to the sprint, a good idea for people who want to start getting more involved?
[10:58] <asabil> http://ifaedi.insa-lyon.fr/~asabil/bzr-viz-tags.png
[11:00] <muffinresearch> Or is it more of a closed event for current core developers?
=== cprov-afk is now known as cprov
=== mrevell is now known as mrevell-lunch
=== mrevell-lunch is now known as mrevell
[13:44] <eean> does bzr svn not follow svn moves?
[13:44]  * eean looks at the website again
[13:45] <jrydberg_> svn does not have move
[13:46] <eean> um, yes it does
[13:46] <jrydberg_> it has copy, iirc.
[13:46] <eean> and move
[13:46] <eean> but anyways
[13:46] <eean> its possible either way
[13:46] <Akheron> it has sime kind of add with history?
[13:46] <eean> does bzr not follow them?
[13:46] <eean> Akheron: yes
[13:47] <jelmer> eean: svn only has copy+delete
[13:47] <eean> that sounds awfully like moving to me
[13:47] <eean> and there is a svn move command
[13:47] <jelmer> eean: in order to be able to convert a copy+delete to a rename you have to prove there is only one copy of a file and that the original was removed
[13:47] <eean> I don't care if its seen as a rename or not
=== brilliantnu1 is now known as brilliantnut
[13:48] <jelmer> "svn move" is an alias for "svn copy"+"svn delete"
[13:48] <eean> but it thinks this 8 month old project is only a day old
[13:48] <eean> since it was copied+deleted yesterday
[13:48] <jelmer> it should follow copies of branches
[13:49] <jelmer> (if you copy /trunk to /tags/2.0 it should recognize that)
[13:49] <eean> ah ok, well it didn't. I'll double check the svn log
[13:50] <jelmer> is this a public repository?
[13:50] <eean> yes
[13:50] <eean> svn://anonsvn.kde.org/home/kde/trunk/KDE/kdemultimedia/dragonplayer
[13:51] <eean> well svn log tracks its history through all the reneames
[13:51] <jelmer> ah, the KDE repository
[13:52] <eean> this project is entirely post-CVS if that's what you're thinking :)
[13:52] <eean> I heard that conversion messed some things up
[13:52] <jelmer> At the moment the copy tracking only works for the current "branching scheme"
[13:52] <eean> ah
[13:52] <jelmer> by default that's the standard /trunk ; branches/* ; tags/* structure
[13:53] <asabil> jelmer: is that what you wanted : http://ifaedi.insa-lyon.fr/~asabil/bzr-viz-tags.png ?
[13:53] <eean> jelmer: can you add this as a known problem?
[13:53] <jelmer> eean: We're trying to make that deal with other situations as well, but that won't happen until 0.5.0
[13:53] <eean> cause I just wasted 3 hours last night, restarting bzr svn constantly :|
[13:53] <jelmer> eean: There's already a bug for it
[13:54] <asabil> (iirc I had a discussion yesterday with you and Schierbeck)
[13:54] <eean> I mean, on the website
[13:54] <jelmer> asabil: Ah, that looks quite nice
[13:54] <jelmer> eean: will do, thanks
[13:55] <asabil> jelmer: want a branch to give it a try?
[13:55] <eean> cool thank you
[13:55] <jelmer> asabil: any chance you can send an email to the bzr-gtk list?
[13:55] <asabil> jelmer: I am not subscribed :/
[13:55] <jelmer> asabil: Should be easy to do :-) Otherwise I can manually approve your email
[13:56] <asabil> jelmer: I was subscribed a while ago
[13:56] <asabil> but the list is very high traffic for me
[13:56] <jelmer> asabil: it's a lot quieter these days :-/
[13:56] <eean> jelmer: do you plan on optimizing how you use the sqlite database? I suppose personally I don't optimize until after I'm finished, so I can't blame you for not. :)
[13:57] <eean> cause for KDE its 1.9gigs
[13:57] <jelmer> eean: the focus so far has mostly been on correctness
[13:57] <jelmer> eean: hopefully we can cache less in the near future
[13:58] <rjek> corrections is paramount: performance is a bonus.  Especially in this sort of product.
[13:58] <eean> right
[13:58] <rjek> s/corrections/correctness/
[14:04] <asabil> jelmer: https://code.launchpad.net/~asabil/bzr-gtk/viz-tags-fancy
[14:04] <asabil> jelmer: please feel free to mail the list about it, if you want
[14:04] <asabil> I am busy with other things right now sorry
[14:05] <asabil> I edited the branch summary to point to the screenshot
[14:32] <schierbeck> asabil: hello!
[14:33] <asabil> hey schierbeck
[14:33] <asabil> schierbeck: https://code.launchpad.net/~asabil/bzr-gtk/viz-tags-fancy
[14:33] <schierbeck> i've just seen it, great job!
[14:33] <schierbeck> are you on the mailing list?
[14:35] <asabil> nop schierbeck
[14:35] <asabil> please feel free to mail the list about it if you want
[14:35] <asabil> and thanks :)
[14:36] <schierbeck> i'm currently looking into some small changes, like decreasing the font size, having the tags stack horizontally so they're all visible, and using a single color for all tags
[14:37] <asabil> decreasing the font size is as simple as
[14:38] <asabil> tag_layout.set_markup("<small>" + tag + "</small>")
[14:38] <asabil> instead of tag_layout.set_text(...)
[14:38] <asabil> (but am not sure if you want to use tags)
[14:39] <asabil> making the tags stack horizontally is a matter of changing x0 and leaving y0 fixed
[15:05] <schierbeck> asabil: ok, i've made a few changes, they should be at http://bazaar.launchpad.net/~bzr-gtk/bzr-gtk/viz-tags
[15:10] <asabil> schierbeck: did you try the code with many tags ?
[15:10] <asabil> there is too much space between the tags I think
[15:10] <schierbeck> asabil: only with two -- i'll check it out now
[15:11] <asabil> but it looks nicer indeed
[15:11] <asabil> there is just one issue
[15:11] <asabil> I am not sure about using markups
[15:11] <schierbeck> yeah, it screws up
[15:11] <asabil> as it might blow up if ^evil^ developers decide to use pango markup in their tags :D
[15:12] <schierbeck> asabil: we'll just escape the tag text
[15:12] <schierbeck> we do that with the summary already
[15:12] <asabil> if it is just about making the text smaller
[15:12] <schierbeck> no, it's also about tooltips and such
[15:12] <asabil> you can set the font directly
[15:12] <asabil> oh ok
[15:13] <cr3> how can I apply a bzr diff between releases 239 and 240 for example?
[15:13] <luks> "apply"?
[15:14] <cr3> bzr revert -r239..240?
[15:14] <dato> cr3: bzr merge -r 239..240
[15:14] <dato> or -r 240..239, depending on what you want
[15:15] <cr3> dato: cheers, exactly what I was looking for
[15:16] <ubotu> New bug: #183821 in bzr-svn "bzr branch fails with AssertionError in parse_revid_property" [Undecided,New] https://launchpad.net/bugs/183821
[15:17] <schierbeck> asabil: try pulling again
[15:18] <schierbeck> it seems to be working now
[15:18] <asabil> imho it looks better
[15:19] <asabil> :)
[15:19] <schierbeck> asabil: would you be able to make the label have a single, common color?
[15:19] <schierbeck> like a yellow, post-it-ish one?
[15:19] <asabil> I am not sure if it is good idea
[15:20] <schierbeck> i think it would be more aesthetically pleasing
[15:20] <asabil> so we need to decide on a color then
[15:20] <schierbeck> yup
[15:20] <asabil> a hardcode color :p
[15:20] <schierbeck> :D
[15:21] <fullermd> You could make it a red/blue offset, and sell 30-cent cardboard/gel glasses, so people can see them in 3-D!
[15:21] <schierbeck> :x
[15:22] <schierbeck> there may lie a few usability issues right there... :)
[15:22] <schierbeck> although it would be kick-ass!
[15:23] <fullermd> I can just see the reviews, though...   "I liked bzr overall, but it felt like the tags were attacking me."
[15:23] <schierbeck> fuck, i always seem to get excited about these projects when i'm studying... my exam is on wednesday!
[15:24] <fullermd> Heck, that's _days_ away...  plenty of time   ;>
[15:24] <schierbeck> fullermd: we'd still be less scary than git...
[15:24] <fullermd> Yeah.  git would be 3-d, but would sneak up from behind you.
[15:26] <schierbeck> yup, and it would hold you down until you iterated all available git commands
[15:26] <jelmer> schierbeck: That sounds familiar...
[15:27] <fullermd> "Say 'git-fmt-merge-msg'!  *SLAP*  SAY 'git-fmt-merge-msg'!!!"
[15:27] <jelmer> for some reason the time I'm most productive is the days before I have a deadline or an exam or something
[15:27] <schierbeck> jelmer: me too, it's just that i'm not studying
[15:27] <schierbeck> everything else seems so damn interesting in comparison!
[15:28] <fullermd> Well, I've got this project you could do some work on...  would cure that little problem RIGHT up   :p
[15:28] <schierbeck> zzzzzZZZzzz
[15:28] <schierbeck> huh.. what?
[15:29] <schierbeck> :)
[15:30] <jelmer> :-)
[15:30] <schierbeck> jelmer: have you checked out my additions to asabil's whoop-cake awesome work?
[15:30] <schierbeck> http://bazaar.launchpad.net/~bzr-gtk/bzr-gtk/viz-tags
[15:30] <jelmer> not yet
[15:31] <schierbeck> it's on the team repo, so you guys can write to it
[15:31] <jelmer> thanks, will have a look
[15:31] <schierbeck> :)
[15:34] <asabil> schierbeck: you can pull from my branch
[15:34] <asabil> the colour is yellow now
[15:36] <schierbeck> cool!
[15:38] <schierbeck> asabil: that looks great! i've merged it into the ~bzr-gtk branch
[15:38] <schierbeck> you should be able to pull
[15:38] <jelmer> schierbeck: ? Your branch you mean?
[15:39] <schierbeck> http://bazaar.launchpad.net/~bzr-gtk/bzr-gtk/viz-tags
[15:48] <Bwoaas> Hi all. I have a question regarding 'commit --local' in 0.90.0. When I use it, I keep getting conflicts, and now I am wondering if I am using it in the correct way.
[15:48] <Bwoaas> Currently I do:
[15:48] <Bwoaas> bzr co sftp://[snip]
[15:48] <Bwoaas> bzr ci --local
[15:48] <Bwoaas> bzr ci
[15:48] <Bwoaas> and this is where I get conflicts
[15:48] <asabil> schierbeck: can you pull again please ?
[15:49] <schierbeck> asabil: sure
[15:49] <asabil> not very sure if it looks better
[15:49] <asabil> fell free to test it
[15:49] <asabil> also maybe you want to delete the tag that I added and forgot to remove :
[15:49] <asabil> :D
[15:49] <asabil> the Woot tag
[15:50] <schierbeck> cool!
[15:50] <schierbeck> the holes could be a bit bigger, though
[15:50] <schierbeck> jelmer: new patch sent to the ml, it's pretty trivial
[15:50] <asabil> schierbeck: they don't look nice when bigger
[15:51] <asabil> maybe you can try
[15:51] <schierbeck> ok
[15:57] <schierbeck> asabil: okay, i've made them a bit larger
[15:57] <schierbeck> pull them in
[15:57] <asabil> yep that's better :D
[15:58] <schierbeck> do you think we can adjust the position of the hole?
[15:58] <schierbeck> it's just a tad too high
[15:58] <asabil> I think it looks perfect to me
[15:58] <asabil> it is right in the middle
[15:58] <schierbeck> yeah, could be i'm wrong -- such things depend on the screen and whatnot
[15:58] <asabil> oO how can it be high ?
[15:59] <schierbeck> asabil: many such things are psychological -- sometimes you need to bend things to make them look even
[15:59] <asabil> :/
[15:59] <schierbeck> :)
[16:00] <asabil> oh and btw, I am not sure you needed a bzr merge
[16:00] <asabil> you could have used bzr pull
[16:00] <schierbeck> nah, you forgot to pull my merger
[16:00] <schierbeck> i tried
[16:00] <asabil> because there was no divergence
[16:01] <schierbeck> look at the graph -- i had already merged with your branch once, hence there was a revision in my branch that wasn't in yours
[16:02] <asabil> oh ok sorry
[16:02] <asabil> didn't notice
[16:27] <schierbeck> np :)
[16:46] <ubotu> New bug: #178108 in bzr-svn "Implement SvnRepository.find_branches()" [Wishlist,Fix committed] https://launchpad.net/bugs/178108
[16:59] <schierbeck> asabil: do you think the branch is ready to be merged into trunk?
[16:59] <asabil> schierbeck: yep, I don't see any big issue
[16:59] <schierbeck> then i'll make a merge request
[17:00] <asabil> cool :)
[17:00] <asabil> next time I might make the tags corner rounded :D
[17:00]  * asabil is a bling addict
[17:04] <schierbeck> asabil: i thought about it, too -- it would be cool
[17:04] <schierbeck> btw, sign up for the mailing list
[17:05] <fullermd> And it should be a cube!  That you can spin!
[17:05] <asabil> schierbeck: the ML is too high traffic for me
[17:05] <asabil> schierbeck: I was subscribed
[17:06] <asabil> fullermd: lol
[17:06] <schierbeck> it's not that high traffic atm...
[17:06] <asabil> ok ok I will subscribe again
[17:06] <jelmer> it's about 40 messages per month these days
[17:06] <lifeless> jelmer: hi, sorry I haven't gotten you on pqm yet
[17:07] <jelmer> 'morning lifeless
[17:07] <jelmer> lifeless: no hurry; would be nice to be able to use pqm at some point though
[17:07] <lifeless> jelmer: I'm in London this week; so we can chat in the same timezone (FSVO same)
[17:08] <jelmer> ah, ok
[17:09] <asabil> schierbeck: I am subscribed
[17:09] <asabil> to bazaar@lists.....
[17:10] <schierbeck> nah, it's bzr-gtk@lists.canonical.com
[17:12] <asabil> ah oki
[17:14] <asabil> done
[17:28] <welterde> is there a command to view the gpg-signature of a commit?
[17:30] <jelmer> welterde: not yet afaik
[17:32] <welterde> jelmer: hmm... how strong is the authentication of the commits? eg. do they depend on a hash of the commit-data, on the previous commits?
[17:32] <welterde> (like git for example)
[17:34] <jelmer> welterde: afaik both
[17:35] <welterde> hmm.. so manipulating a older revision would be detected?
[17:35] <jelmer> yes
[17:35] <jelmer> at the moment, there isn't any functionality to check signatures though
[17:37] <welterde> but gpg-signatures are checked afair, or not?
[17:37] <jelmer> I don't think they are either
[17:38] <welterde> ah.. is an extra plugin
[17:50] <lifeless> welterde: commit signatures do not transitively verify older history
[17:50] <lifeless> welterde: this is deliberate
[17:51] <lifeless> welterde: as it allows partial history environments to be verified
[17:53] <yacc> Just wondering, is it possible to store the whoami value on a per project basis?
[17:56] <lifeless> sure; you can probably set it in branch.conf; or url based in locations.conf
[17:58] <welterde> lifeless: so no protection against a manipulated bzr branch?
[17:58] <lifeless> welterde: uh, sure there is protection; just sign the commits for the history as well
[17:59] <welterde> hmm... yeah... but besides that there is none?
[18:00] <lifeless> I'm not sure what you're getting at
[18:01] <welterde> if someones stops by and manipulates the unsigned history, that wouldnt be detected, right?
[18:02] <welterde> (manipulates in a way the checkout changes)
[18:14] <welterde> does bzr leak ip-addresses on any operation?
[18:19] <lifeless> welterde: 'leak'?
[18:19] <welterde> lifeless: like, being noted somewhere
[18:19] <welterde> inside of commits or something like that
[18:20] <lifeless> it may record them in your !/.bzr.log; and the commiter name is derived from your username and hostname by default
[18:20] <welterde> which can be changed with bzr whoami
[18:22] <lifeless> right
[18:23] <welterde> lifeless: but signing one commit doesnt automaticly "sign" all of the commits before :/
[18:23] <lifeless> right
[18:23] <lifeless> you can bzr sign-my-commits though if you haven't been signing them
[18:24] <welterde> and when the person, of whom the commits are, is not reachable?
[18:24] <welterde> thus cant sign his commits?
[18:27] <welterde> lifeless: and can i sign commits which just have a name?
[18:27] <welterde> eg. from a cvs-import
[18:28] <lifeless> sure; I'm off to dinner now, have fun
[18:31] <Peng> welterde: You might also want to change your bazaar.conf/locations.conf/branch.conf to require signatures. See the docs.
[18:32] <welterde> Peng: yeah... know that
[18:52] <Enquest> is there a way to get some stats out of bazaar... Like how many lines of code has been changed the last day or period?
[18:54] <james_w> Enquest: apart from 'bzr info' there is nothing that I know of.
[18:54] <Enquest> maybe that would be a cool thing to get into it.
[18:55] <james_w> However the example you mention could be approximated by piping something like 'bzr diff -rdate:yesterday' to diffstat
[18:56] <james_w> I agree it would be interesting. It would work very nicely as a plugin I expect.
[18:56] <beuno> Enquest, and there is a stats plugins which shows mostly per-user stuff:  http://bzr.arbash-meinel.com/plugins/stats
[19:31] <elmo> Copying repository content as tarball...
[19:31] <elmo> bzr: ERROR: Tags not supported by BzrBranch5('file:///home/james/scratch/trunk/'); you may be able to use bzr upgrade --dirstate-tags.
[19:31] <elmo> why does bzr hate me?
[19:32] <fullermd> It's your shoes.
[19:42] <elmo> wtf
[19:42] <elmo> if I rsync the entire dir, and bzr branch locally it works
[19:42] <elmo> bzr makes me CRY
[19:43] <Peng> elmo: Bzr version? Smart server? Shared repo?
[19:43] <fullermd> Smart server, I suspect.
[19:44] <elmo> bzr 1.0, bzr+ssh (so I guess smart?), and yes, shared repo I think
[19:44] <fullermd> An incarnation of bug 173002.
[19:44] <ubotu> Launchpad bug 173002 in bzr "Branching from hpss doesn't preserve non-repository formats" [High,New] https://launchpad.net/bugs/173002
[19:44] <Peng> Ah.
[19:44] <Peng> Does that affect bzr+http too?
[19:45] <fullermd> I presume it affects bzr+anything.
[19:45] <elmo> shall I followup to the bug?
[19:45] <fullermd> I dunno why your bzr 1.0 would try to make a branch5 branch, though.
[19:46] <fullermd> It should go with the default, which is branch6.
[19:46] <fullermd> Please.
[19:48] <Peng> You may have been the person who said "no" last time I asked, but is there an easy way to "upgrade" from branc6 to branch5?
[19:48] <Peng> branch6*
[19:48] <elmo> I didn't create the repo on the other end
[19:48] <fullermd> You can pull.  It may even not bother giving you an error if there are no tags.
[19:51] <dysinger> Heh! morning.
[19:52] <dysinger> Anyone in here on OS X Leopard? (I presume most of you are on Ubuntu/Debian.)
[19:53] <dysinger> Jelmer I updated the wiki page several times last week.  I never did get it 100% right.  Hopefully I can keep bugging people on Leopard to try it out....
[19:53] <dysinger> (it being bzr-svn)
[19:53] <elmo> bah.  sorry, I somehow ended up on gutsy bzr instead of 1.0
[19:53] <elmo> 1.0 works
[19:53] <fullermd> Well, 1.0 works by coincidence.
[19:54] <fullermd> It happens to use branch6 in its default format, so you don't hit the same problem.
[19:54] <Peng> God, Canonical owns our souls. Bazaar, Ubuntu, Launchpad...
[19:54] <fullermd> It's till a bug that it doesn't preserve the remote format.
[19:54] <elmo> fullermd: ah, ok - so still follow up?
[19:54] <dato> Peng: and that is ontopic how?
[19:54] <Peng> dato: It's not. :)
[19:54] <fullermd> He's fishing for bribes to keep quiet about it.
[19:55] <fullermd> Maybe the common criticism is wrong; bzr needs MORE format changes, to keep those sort of bugs in the forefront and high on the priority stack   ;)
=== kiko is now known as kiko-afk
[20:15] <schierbeck> jelmer: ping
[20:25] <ubotu> New bug: #184898 in bzr "AttributeError: 'NoneType' object has no attribute 'abort'" [Undecided,New] https://launchpad.net/bugs/184898
=== jw2328_ is now known as james_w
[21:19] <welterde> are gpg-signatures pushed yet?
[21:21] <welterde> *over existing non-signed commits
[22:24] <BasicOSX> Is there a way to test bzr's email plugin? I think my upstream smtp server is rejecting relaying, but I want to confirm. Some sort of debug output from the plugin?
[22:30] <james_w> BasicOSX: you could look in ~/.bzr.log
[22:50] <igc> morning
[23:00] <luislavena> hello everybody.
[23:00] <luislavena> is there a way (simpler) to change the commit messages? (4 or 5 revisions)
[23:01] <aklaver> I asked yesterday about the location of the latest debs. Set my repositories to PPA. The problem is there does not seem to be a *.deb for Dapper.
[23:03] <jelmer> dysinger: thanks for the wiki updates
[23:03] <dysinger> jelmer np - I just wish it worked perfectly -
[23:04] <jelmer> dysinger: what's not working?
[23:04] <dysinger> for some reason ssl svn access through bzr dosen't work
[23:04] <dysinger> neon installs, svn 1.5 installs, svn 1.5 + ssl works
[23:04] <dysinger> but not bzr-svn ssl
[23:04] <jelmer> even using the svn+https:// prefix?
[23:04] <dysinger> n
[23:05] <dysinger> I am just hoping for more people trying it soon.
[23:05] <dysinger> It dumps out of python.
[23:05] <jelmer> oh, what error?
[23:06] <dysinger> some stack trace - I didn't study it closely.  I can go back to it soon and give specifics - it's nothing that was real clear to me what the problem was.
[23:07] <dysinger> You can see from the wiki, I am using bzr-svn "stable" repo - is this ok?
[23:07] <jelmer> dysinger: yeah, "stable" is just a symlink to the 0.4 branch
[23:07] <jelmer> dysinger: seeing the backtrace may help debug this problem
=== thumper_laptop is now known as thumper
[23:14] <dysinger> ok hold on 2 minutes jelmer, i'll get it.
[23:21] <jelmer> ok
[23:25] <dysinger> I am installing again fresh
[23:25] <dysinger> I spent time reading through the svn/neon INSTALL docs
[23:25] <dysinger> so I think I have that part right
[23:26] <dysinger> I think this is some sort of problem with Apple's python or something.
[23:26] <jelmer> if it'
[23:26] <jelmer> s giving you a python backtrace, chances are it may be a bzr-svn bug
[23:26] <dysinger> ah ok
[23:38] <dysinger> OK I have svn 1.5 + neon installed with openssl support - all check-swig-py tests pass.  svn --version reveals 1.5 with https support.  I can checkout https://svn.collab.net/repos/svn/trunk with svn 1.5 no problem.  No for the stack dump on bzr-svn
[23:46] <dysinger> So just like it says in the wiki bzr checkout --lightweight http://svn.collab.net/repos/svn/trunk test works
[23:46] <dysinger> jelmer notice no ssl
[23:46] <dysinger> one sec and running ssl
[23:50] <dysinger> jelmer I get
[23:50] <dysinger> bzr checkout --lightweight \
[23:50] <dysinger> >   https://svn.collab.net/repos/svn/trunk test
[23:50] <dysinger> Assertion failed: (g->gc.gc_refs != _PyGC_REFS_UNTRACKED), function instancemethod_dealloc, file Objects/classobject.c, line 2285.
[23:50] <dysinger> Abort trap
[23:52] <dysinger> The apple core dump pop up gives me this
[23:52] <dysinger> http://pastie.caboo.se/141720
[23:57] <dysinger> that's all - I don't know what to do know.
[23:57] <dysinger> now
[23:57] <dysinger> s/know/now/g
[23:57] <dysinger> lolz