[00:42] <poolie> hi all
[01:00] <spiv> Good morning.
[01:01] <Jerub> morning.
[01:03] <Jerub> ah, found the button to upgrade the branch on lp
[01:23] <poolie> hi spiv
[08:15] <bialix> hi all
[08:15] <bialix> vbonjour vila
[08:15] <vila> bialix: hey !
[08:16] <bialix> no, I mean just bonjour
[08:16] <vila> oh, too bad, I thought it was a special bonjour just for me ;-)
[08:16] <bialix> vila: do you by any chance know does 2.2.1 will be released this week as https://launchpad.net/bzr/+milestone/2.2.1 page says?
[08:17] <bialix> I need to prepare new qbzr release then
[08:17] <vila> bialix: not that I know of :-/
[08:19] <bialix> ok, will try to catch poolie tomorrow
[08:22] <vila> bialix: or jam, AIUI he is the 2.2 RM (but IMBW, this occurred during my vacations and I may have missed something)
[08:34] <spiv> Good evening bialix, vila
[08:35] <spiv> bialix: poolie probably won't be around tomorrow, it's igc's funeral :(
[08:35] <vila> spiv: good evening
[08:35] <vila> spiv: can you confirm who is the RM for 2.2 ?
[08:35] <spiv> vila: not sure off the top of my head!
[08:35] <vila> :-/
[08:36] <spiv> I'd have to search my email...
[08:36] <spiv> We should put it in the /topic
[08:36] <lifeless> jam
[08:36] <vila> spiv: indeed
[08:37] <lifeless> IIRC, I'm moderately sure
[08:37] <bialix> you again have killed Kenny :-(
[08:37] <vila> bialix: Is that your poney ?
[08:37] <bialix> spiv: evening
[08:38] <bialix> spiv: I see
[08:38] <bialix> vila: win32 compatibility: https://bugs.launchpad.net/bzr/+bug/631350
[08:39] <bialix> I'm very very very unhappy
[08:39] <bialix> have to fix it now
[08:39] <spiv> bialix: :(
[08:40] <vila> bialix: and no report prior to yours ? Had nobody tested 2.2 on windows with any beta ?
[08:40] <bialix> with non-ascii characters?
[08:40] <vila> bialix: or all they all US... yeah
[08:40] <bialix> I'm reading almost all bug reports now, I don't remember any
[08:40] <spiv> Right :/
[08:41] <bialix> I suspect this bug caused by poolie's refactoring of ui
[08:41] <vila> bialix: one more reason to make windows tests run on babune (it uses a french localized xp and should catch such regressions if there are the right tests for them)
[08:41] <bialix> will try to look at the reasons tonight
[08:41] <vila> bialix: thanks
[08:41] <bialix> it's incredibly hard to write proper test for such things
[08:42] <vila> bialix: then we have to fix it, this shouldn't be hard (not implying it's easy today)
[08:42] <bialix> because it depends on difference between user_encoding() and terminal_encoding() and not all characters could be shown in french locale vs russian locale, and...
[08:43] <vila> bialix: but your report doesn't mention a crash right ? So it's a ui thing and should be (hopefully) shallow
[08:43] <bialix> it's not a crash, but this is very irritating
[08:43] <bialix> and this is definitely regression
[08:44] <vila> +1
[08:44]  * bialix just installed 2.2 final this morning
[08:44] <vila> the weird thing is (without knowing the actual cause) that we are aware of the trap, so it's weird we don't have at least one tests tripping up...
[08:45] <bialix> I understand what you mean
[08:45] <vila> hehe, 'one tests', cute freudian slip :)
[08:45] <bialix> yep :-)
[08:47] <bialix> vila: I'm sure we have non-ascii black box tests, but they don't check the exact output, because of encodings differences between machines
[08:48] <bialix> it should be possible to change, it's just not very trivial and very time consuming
[08:49] <vila> "not very trivial and very time consuming" that's the bug
[08:49] <bialix> but I might try to change this, cause I don't like when Kenny has been killed
[08:50] <bialix> I have no time to work on this right now, but I will catch you in next few days vila and ask you for your mind about specific tests
[08:50] <vila> bialix: ack
[08:52]  * bialix bbl
[09:37] <poolie> hi vila
[09:42] <vila> hey poolie, sry was otp
[09:42] <poolie> np
[09:42] <poolie> just saying hi, how are you?
[09:43] <vila> fine, I've tweaked babune a bit this week-end so that some failed runs are restarted automatically, even less work there ;)
[09:45] <vila> and I should soon propose to some of us to run a subset of the test suite on all platforms for a given branch
[09:45] <vila> I can use it right now, I just have to find how to propose it to others
[09:53] <poolie> hm, i seem to have a clone
[09:53] <poolie> vila, one of us should do the announcement
[09:56] <vila> poolie: I haven't closely followed the various installers builds, checking lp
[09:57] <poolie> i think we're all up to date
[09:57] <poolie> we should check
[09:57] <poolie> i got sidetracked by working out why the web site wasn't updating
[09:57] <poolie> now i'm up in brisbane
[09:57] <poolie> at ianb's house
[10:00] <GaryvdM> Hi poolie, vila.
[10:01] <GaryvdM> poolie, vila: The following installers are done: Windows, Mac, and Ubuntu ppas.
[10:08] <vila> poolie: should the Release notes text be used as is for the announcement, it sounds fine as is
[10:08] <vila> poolie: at https://edge.launchpad.net/bzr/+milestone/2.2.0
[10:09] <poolie> good with me
[10:23] <vila> poolie: mail sent and announcement done on lp
[10:23] <vila> poolie: wikipedia already up-to-date...
[10:35] <poolie> thanks!
[10:57] <poolie> vila, see my pm?
[11:46] <HollyRain> hi! in .bzrignore file, can be used comment lines? #
[11:47] <HollyRain> comments in lines (which start with character #)
[12:33] <Odd_Bloke> I have a repository which I just removed a lot of branches from.
[12:33] <Odd_Bloke> Is there any way I can remove the now unreferenced revisions from it?
[12:33] <Odd_Bloke> s/repository/shared repository/
[12:37] <maxb> Only by branching each branch you want to keep into a new empty shared repository, AFAIK
[12:44] <MichealH> Hello?
[12:45] <MichealH> Can someone tell me how to login to Launchpad via #bzr
[12:45] <MichealH> *bzr
[12:47] <MichealH> Anyone?
[12:48] <maxb> What do you mean? Set your launchpad username to allow bzr+ssh access to branches?
[12:48] <maxb> If so, 'bzr launchpad-login your-lp-id'
[12:48] <MichealH> Okay
[12:48] <MichealH> I need a ssh key?
[12:49] <maxb> for that, yes
[12:49] <MichealH> How do I make one?
[13:04] <HollyRain> hi! in .bzrignore file, can be used comment lines?
[13:04] <HollyRain> comments in lines (which start with character #)
[13:07] <bialix> HollyRain: yes, o'course
[13:08] <HollyRain> bialix:  ok, I'd been looking for it in doc. but I din't find nothing
[13:09] <bialix> lemme check
[13:10] <HollyRain> at least here, no
[13:10] <bialix> HollyRain: yes, I think you're right.  I can't find this either. Can you file a bug report, please?
[13:10] <HollyRain> http://doc.bazaar.canonical.com/bzr-0.9/tutorial.html
[13:11] <HollyRain> I'll file the bug report
[13:12] <bialix> thank you
[13:13] <HollyRain> np
[13:18] <maxb> HollyRain: You know that's a very very old version of that document?
[13:23] <bialix> maxb: http://doc.bazaar.canonical.com/bzr.2.2/en/user-reference/ignore-help.html there is nothing about comments; http://doc.bazaar.canonical.com/bzr.2.2/en/user-reference/patterns-help.html there is nothing as well
[13:23] <maxb> Sure, I just didn't want HollyRain relying on potentially other out of date info
[13:24]  * bialix nods
[13:25] <HollyRain> This was looking at http://doc.bazaar.canonical.com/bzr.2.2/en/user-guide/index.html but nothing
[13:25] <HollyRain> anyway the bug has been reported
[15:04] <bialix> vila: thanks
 the weird thing is (without knowing the actual cause) that we are aware of the trap, so it's weird we don't have at least one tests tripping up...
[15:23] <mgz> it has been rather hard to write proper tests ticking this kind of thing as they tended to make the whole test suite fall over if they failed
[15:31] <vila> mgz: I don't get it. If you write a single test to reproduce the problem, there is no reason for it to... ohhhhh
[15:32] <vila> mgz: but with all your fixes, that should a thing of the past !
[15:32] <mgz> yup, there's now no reason not to.
[15:32] <vila> mgz: right, at least that's a reasonable explanation (bialix ? Saw that ?)
[15:33] <bialix> wwwhat?
[15:34] <vila> bialix: mgz more or less said that unicode-related failing tests had the tendency to break the whole test suite. That may be why we don't have enough of them to catch the problem you reported earlier
[15:34] <bialix> you guys read the mind of each other, I forgot my mind reader at home
[15:34] <bialix> vila: hmm, I don't think so
[15:34] <bialix> or maybe mgz means testtools quirks?
[15:35] <vila> bialix: mgz fixed a couple/bunch of issues regarding tests where the *log* contained unicode, yeah, testtools, subunit, pyjunitxml
[15:36] <bialix> yep, it bit me couple of times. glad it's  fixed now
[15:36] <bialix> but the problem with unicode output is not new, it was there long before testtools and other zoo
[15:37] <mgz> yeah, there have been a bunch of issues, and I may still have to hunt a few more down.
[15:37] <mgz> but I have my trusty bug spear!
[15:37] <vila> bialix: you said it's a regression, that sounds like something not covered by a test. And there may be a good reason for that
[15:39] <bialix> the reason is in the fact that different windows machines have different oem codepage, so we have to create expected result non-ascii string. it could be a bit boring
[15:40] <mgz> we can fake that a little too though, alexander, run_bzr takes an encoding parameter for the streams, so we could make a test that fails on nix as well.
[15:41] <bialix> actually it's a stab in the dark right now. I think at least status should check exact string. unless we force utf-8 in the tests :-/
[15:41] <mgz> (ie, the output should be in the stream encoding, not utf-8 as the user encoding will still be)
[15:41] <bialix> mgz: the crucial point is to have get_user_encoding() and get_terminal_encoding() return different values
[15:42] <mgz> but that might not be needed to make a test that will fail if the wrong one is used, was my point. anyway, I'll have a look at the bug in more detail.
[15:48] <vila> bialix: we know how to force get_user_encoding() and get_terminal_encoding() for a given test
[15:48] <bialix> mgz: I'm not quite understand your point, honestly. that bug is all about using wrong encoding (ansi instead of oem)
[15:48] <bialix> vila: yes
[15:49] <bialix> we have private hook points in osutils to do so
[15:49] <vila> yup, so we should be able to reproduce your bug
[15:50] <bialix> I can reproduce it even without this (joking joking)
[15:51] <mgz> ...I'm not having any luck on that front at the moment...
[15:51] <vila> lol
[15:51] <mgz> I've got a cp437 terminal on a cp932 system, and bzr 2.2 r5082
[15:52] <vila> mgz: use astral chars ?
[15:52]  * vila wonders what astral chars are....
[15:52] <vila> klingon ?
[15:52] <mgz> ehe, they're ones that get encoded as \uXXXX\uXXXX in utf-16 :)
[15:53] <maxb> surrogate?
[15:53] <mgz> yup, with surrogates.
[15:55] <bialix> cp932??? http://msdn.microsoft.com/ru-ru/goglobal/cc305152(en-us).aspx
[15:55] <bialix> japanese?
[15:55] <GaryvdM> mgz: Any chance I can put uncode in the log, and have it printed as uncode, and not as a \uXXXX string?
[15:56]  * bialix waves hello on GaryvdM
[15:56] <jelmer> GaryvdM: please tell me more about this new "uncode" character set :-P
[15:56] <mgz> bialix, yup, that's what my system uses... I'll try a russian terminal codepage and c/p your example
[15:56]  * jelmer also waves
[15:57] <GaryvdM> mgz: The reason is, for my qlog tests, I can print a much more informative picture if I can use unicode.
[15:57] <GaryvdM> Hi bialix
[15:57] <bialix> cp866 is russian oem,
[15:57] <GaryvdM> :-P -> jelmer
[15:57] <bialix> mgz: try any accented character
[15:57] <mgz> garyvdm: should Just Work now
[15:57] <mgz> but you'll want new versions of several things possibly.
[15:57]  * GaryvdM installs the latest testtools
[15:59] <mgz> ...bialix, why is bzr init 5 in your example creating a 1.9 format tree? I get a 2a one with 2.2
[16:00] <bialix> mgz: because I have special plugin to enforce this
[16:00] <mgz> okay.
[16:00] <bialix> try look for format1 plugin in ml year or so ago
[16:03] <mgz> so, no luck: http://pastebin.ubuntu.com/489284/
[16:03] <mgz> there must be some other factor involved than just the osutils functions
[16:04] <bialix> mgz: https://bugs.launchpad.net/bzr/+bug/631350/comments/3
[16:04] <GaryvdM> mgz: Sweet - it works: http://pastebin.ubuntu.com/489285/
[16:05]  * bialix pulls latest changes from lp:bzr/2.2
[16:05] <bialix> GaryvdM: I suspect we need to rebuild 2.2 installere
[16:06] <mgz> one note of caution gary: you need to be actually using unicode objects not utf-8 strs if you want it to work elsewhere, and still may print funny on terminals of limited ability
[16:06] <GaryvdM> bialix: Do we need to do that before 2.2.1?
[16:06] <mgz> ah, good idea bialix, I'll grab the 2.2 installer and try that
[16:07] <bialix> GaryvdM: it depends on the date of 2.2.1
[16:07] <GaryvdM> mgz: Yes - my next question is how can I detect If I can safely use unicode, or if I should rather output ascii?
[16:08] <mgz> check the terminal encoding! :)
[16:08] <GaryvdM> bialix: What reason for 2.2.0-2 ?
[16:08] <bialix> mgz: ha! 2.2 branch is not broken!
[16:08] <mgz> we're trying to track down an encoding bug gary
[16:08] <GaryvdM> mgz: How can I check for testtools support?
[16:09] <bialix> mgz: http://pastebin.ubuntu.com/489290/
[16:09] <mgz> __version__ greater than 0.9.5, but I think we'll want to be enforcing that soon anyway
[16:09] <bialix> GaryvdM: waut, why you said 2.2.0-2?
[16:09] <bialix> GaryvdM: wait, why you said 2.2.0-2?
[16:10] <mgz> +or equal to
[16:10] <bialix> GaryvdM: https://launchpad.net/bzr/2.2/2.2.0 has only 2.2.0 without -1 or -0
[16:10] <bialix> what I'm missing? *blinks*
[16:11] <GaryvdM> bialix: I just meant a new ver of the 2.2.0 installer.
[16:11] <vila> bialix: x.y.z-n 'n' if for installers if something goes wrong but doesn't require a change in x.y.z
[16:11] <bialix> GaryvdM: what's new docs bug
[16:11] <GaryvdM> ok
[16:12] <bialix> GaryvdM: https://bugs.launchpad.net/bzr-windows-installers/+bug/631470
[16:12] <GaryvdM> ok
[16:12] <bialix> and I'm not sure yet what's wrong with encodings
[16:12] <mgz> oaky, yup, the installer is borked
[16:12] <mgz> C:\Program Files\Bazaar2.2>bzr.exe mkdir b:\alex2\Тест
[16:12] <mgz> added b:\alex2\'?aa
[16:13] <bialix> finally
[16:13] <mgz> I'll pdb in an poke around, see if I can work out what's up.
[16:14] <mgz> user and terminal encoding are right...
[16:14] <bialix> using pdb on bzr.exe is...
[16:15] <bialix> can't find the word
[16:15] <bialix> a bit maso
[16:15] <mgz> :)
[16:19] <vila> bialix: you're talking to mgz you know...
[16:19]  * bialix have to go now, hopefully bbl tonight
[16:19] <bialix> vila: err?
[16:19] <vila> bialix: joke :)
[16:19] <bialix> I suspect that
[16:19] <mgz> se you later bialix, I'll see if I can make some progress on this.
[16:19] <bialix> I know mgz is fearless
[16:37] <mgz> okay, this is kinda funky.
[16:40] <mgz> the bazaar ui object contains a codecs wrapper, contains stdout
[16:40] <mgz> the codecs wrapper and stdout both have the encoding cp866 set
[16:41] <mgz> and the wrapper correctly encodes the unicode russian to a cp866 bytestring, and writes it to the python file
[16:41] <mgz> ...which somehow is coming out mangled as-if it's being decoded as... something random latin-ish then encoded as mbcs or similar
[16:42] <mgz> sys.version is 2.6.4 ... can we build an installer with something else?
[16:50] <mgz> ...not that I can find any likely upstream bug
[16:54] <GaryvdM> mgz: The 2.1 installers have python 2.5 - I don't know if that help?
[16:54] <mgz> I'll try one, pretty sure this isn't related to bzrlib code changes
[16:56] <mgz> py2exe seems most likely, to be honest
[16:59] <mgz> (Pdb) ctypes.cdll.msvcrt.printf("\x92\xa5\xe1\xe2")
[16:59] <mgz> Тест4
[16:59] <mgz> (Pdb) os.write(sys.stdout.fileno(), "\x92\xa5\xe1\xe2")
[16:59] <mgz> '?aa4
[17:05] <mgz> so, certainly a build problem, printing through msvcr90.dll is borked, and worked through msvcrt.dll both of which are linked
[17:07] <mgz> unfortunately that means I'm not entirely sure what the right fix is.
[17:10] <maxb> Sounds a bit wrong to have 2 CRTs linked
[17:11] <mgz> it's not as wrong as it sounds from nix perspective, but it's possibly indicative of something
[17:12] <mgz> as I understand it, starting with 2.6 you have to ship some vc 9 dlls with python
[17:12] <mgz> the older threaded runtime is from my system.
[18:41] <docoptix> hi. does anyone know of a way to limit the bandwidth bzr uses on commit?
[19:24] <Muscovy> If I've made edits to an older revision of a branch, is there a way to automatically merge my changes with the new version?
[19:28] <beuno> Muscovy, sure, bzr merge should do that for you
[19:28] <beuno> or do you *just* want to cherrypick that change?
[19:29] <Muscovy> merge is probably what I'm looking for.
[19:30] <Muscovy> I don't have an issue right now, but I've manually merged things a few times now.
[19:30] <Muscovy> I figured it was time to learn the proper way. :P
[19:31] <mgz> are your changes committed or uncommitted?
[19:31] <mgz> if uncommitted, just `bzr update` and resolve the conflicts
[19:32] <Muscovy> bzr update wouldn't overwrite my edits?
[19:32] <mgz> no, it's nice like that.
[19:33] <Muscovy> Ok, thanks.
[20:57] <maxb> jelmer: Hi. Are you around? I wanted to bring bug 397526 to your attention.
[21:08] <bialix> jam: ping
[21:08] <jelmer> maxb: Yeah, I've seen it
[21:09] <jelmer> maxb: Haven
[21:09] <jelmer> 't had time to roll a new tarball yet
[21:09] <bialix> vila: are you still here?
[21:09] <maxb> Ok - as it was an old bug which got reopened, I just wanted to check it had been noticed
[21:12] <jelmer> maxb: Unfortunately bzr-gtk is severely lacking developers at the moment
[21:13] <maxb> I pretty much only use it for bzr-notify :-/
[21:13] <jelmer> maxb, what about bzr viz?
[21:14]  * jelmer would still like to make the nautilus integration really rock at some point
[21:14] <maxb> qlog is too good not to use :-)
[21:15] <maxb> In fact, that's pretty much the sole reason I even have Qt installed at all :-)
[21:16]  * jelmer has qt installed for mumble, skype and twinkle
[21:16] <jelmer> for some reason proper VOIP apps only come in Qt form
[21:19] <mac9416> How can I remove a file from version controlling without deleting it.
[21:20] <mgz> bzr remove --keep
[21:21] <mac9416> mgz, thanks.
[21:27] <bialix> mgz: thanks for your analysis
[21:27] <mgz> I'm a bit stuck on how to fix it though unfortunately...
[21:28] <mgz> well, apart from big things like "go back to Python 2.5"
[21:29] <bialix> mgz: I'm going to check py2exe faq
[21:30] <bialix> and I'd like to look on the build machine, at least on build product before it get packed into installer. but for this I need either Garyvdm or jam
[23:48] <dorins> Hi. Any qbzr devs around?
[23:51] <dorins> I made some improvements to qdiff UI in this branch: lp:~dorins/qbzr/qdiff-changes . How do I go about requesting that it be merged into main branch?
[23:52] <lifeless> click on 'propose for merge' in the web ui
[23:55] <dorins> Do I leave the default merge target- lp:qbzr? Or should I set a specific version branch as merge target?
[23:55] <lifeless> the default is the main branch
[23:58] <dorins> That was easy. Thanks lifeless!
[23:58] <lifeless> no problem