[00:02] I previously planned to work on getting the %2B displayed unescaped, until I got scared off by the problems %7E vs. ~ was causing [00:02] But yes, there's no real justification for escaping that + [00:16] abentley: hi? [00:16] poolie: hi [00:16] hi i think i have better tc instructions, i was wondering if you could try them on your machine [00:17] poolie: sure thing. [00:18] http://doc.bazaar.canonical.com/bzr.dev/developers/testing.html#simulating-slow-networks [00:18] basically, just omit the third line [00:19] i think that's all that's needed [00:19] like http://typewith.me/7hBAzi2MYN [00:21] poolie: tc qdisc add dev lo parent 30:1 handle 40: prio [00:21] RTNETLINK answers: Operation not supported [00:22] yeah, leave that one out [00:22] just add the root discipline, the netem, and then the filters [00:23] poolie: If I continue, it errors: http://pastebin.ubuntu.com/658315/ [00:23] oh, omit the 'handle 800::800' [00:24] i realize randomly cutting things out does not generally seem to work but in this case it does :) [00:24] i think more stuff is now done automatically and is no longer allowed to be done manually [00:24] poolie: The typewith.me version doesn't error for me. [00:24] hooray [00:24] and does it work? [00:24] presumably you'll want different ports to 4155 [00:25] poolie: dunno, I don't have anything on 4155 :-) Let's try again with 80... [00:25] you might need more than 500ms for it to be clearly visible [00:28] poolie: when you have a moment, do you have the new PQM output for ? [00:29] If you can't notice 500ms latency then you aren't making enough roundtrips ;) [00:29] poolie: I think it's working. [00:29] I'd like to build of that branch to resolve some remaining issues and get it landed at last. [00:29] btw i'm so happy to hear of lp people testing with that [00:29] were you going to use it for html/js ui stuff, or something else? [00:29] mgz, sure, me too [00:31] mgz, i put it on the mp; it looks very similar to what happened before [00:31] abentley, thanks, i'll update the bzr docs [00:31] poolie: yes, it's working. Thanks! [00:33] https://code.launchpad.net/~mbp/bzr/doc/+merge/70386 [00:34] poolie: thanks! I'll have a crack at that tomorrow. [00:37] thanks for fixing the knownfailure mismatch [00:37] i think there's a bug for that? [01:12] hi poolie, did you have a moment to talk about bug #54624? [01:12] Launchpad bug 54624 in Bazaar "warn when committing large (binary) files" [Medium,Confirmed] https://launchpad.net/bugs/54624 [01:13] oh hi [01:13] sure [01:13] so, you asked about doing it from 'add' [01:13] in bzr, add just means 'mark this as versioned' of course [01:13] so, we could usefully give a warning there i suppose [01:13] but, perhaps people also want to be warned if an already-added file grows very large? [01:13] maybe not [01:14] ah, hadn't thought of that [01:14] well, the original idea seemed to be to prevent people from adding large files they didn't really mean to [01:15] yeah, i think you're right [01:15] also it's a bit easier to undo at that point [01:15] and just a warning will do [01:16] ok, so it seems it should check the file during the list it's running through during an add, and warning at that point, skipping it if it's over the threshold.. ? [01:16] which should probably be configurable [01:16] yes [01:16] and this needs to cover files implicitly added by recursion [01:16] i hadn't dug into it, but it seemed there was a per branch config file. is that the right place? [01:17] that sounds good [01:17] you should get the branch config stack, which will look there and in the globla config etc [01:17] ok. and how are those adjusted usually - by manually editing a file? should there be a command line to override and force this? [01:18] either by editing the file or by 'bzr config' or the gui [01:18] there's a separate bug about adding a command line option to set config flags [01:18] so, i wouldn't worry about that right here [01:18] ok [01:27] ok, so i'll work something up and push it early to a branch for review, like you mentioned. [01:27] cool, thanks [01:27] don't feel it has be perfect before you ask for review [01:28] i have to go out in a bit but i'll be back within about 2 hours [01:28] i'm going to do lots of reviews today [01:29] ah well it's night for me, i wouldn't have anything before tomorrow at least === thomi_ is now known as thomi === krow_ is now known as krow [06:29] hi thomi [06:29] hi poolie [07:21] poolie: I notice that the diff on my lp merge proposal has merge conflict markers in it, but they're not in my local branch. LP bug perhaps? [07:22] also, if you're happy with my latest suggestion I'll go ahead and push that up now [07:58] thomi: i guess those conflicts are because the trunk changed since you started [07:59] poolie: oh, of course. Sorry, it's been a long week ;) [08:00] np, thanks for the patches [08:07] hmm, in trunk there's a half-written docstring in config.LocationStack that's causing the conflict. [08:08] poolie: are you happy with the approach I took in the config stuff? I'm done for the evening, but if you'd like me to change them let me know and I'll do final tweaks tomorrow. [08:18] i haven't reread the patch yet but it sounded good [08:18] * poolie looks now [09:04] jelmer: hi. do you think we should care about shipping subvertpy 0.8 and bzr-svn 1.1 for hardy? [09:06] hi Max [09:07] maxb, I guess so, is there anything in particular that makes it hard to build subvertpy 0.8 on hardy? [09:16] jelmer: debhelper 8.1 :-) [09:22] hi jelmer! [09:23] hi maxb === Mkaysi_ is now known as Mkaysi [09:40] hi poolie [09:40] maxb: ah, argh [09:40] * jelmer tries to remember what we need 8.1 for [09:41] wow [09:41] ok, that's confusing [09:41] oneiric's unity shows the wrong title bar text [09:41] or, slightly out of date [09:42] that was nearly embarassing [09:42] ouch [09:43] I'm finding unity fairly bleeding edge in oneiric too [09:43] it still says #canonical when i've clicked in to #bzr [09:44] i'm just going to file about that then perhaps you can help me with something small [09:45] Are we expecting a 2.4 RC sometime, or is .0 final the next up? [09:45] .0 is next unless there's a showstopper [09:49] vila should be back from his break in a day or two [09:49] did you want to put something in it in particular? [09:49] https://bugs.launchpad.net/ubuntu/+source/unity/+bug/820836 [09:49] Ubuntu bug 820836 in unity (Ubuntu) "window title gets out of date" [Undecided,New] [09:49] Nah, just curious. [10:02] hi all ! [10:03] Speaking of the devil... [10:09] hmm, should I read the logs or can I get a summary ? ;) [10:10] mgz: welcome back [10:11] vila: Oh, nothing too interesting really. [10:11] 04:49 vila should be back from his break in a day or two [10:11] mgz: I hope my fix for testtools actually fixed it, I don't have a windows install to test against [10:11] Apparently, 13 minutes is a day. Or maybe two ;) [10:12] jml: thanks for the quick fix, I can pull again [10:12] hehe, I'm back online but working from a friend's place, network needs to be tweaked, should be done soon [10:13] mgz: cool. [10:13] poolie: still around ? [10:13] mgz: fwiw, I want to get the unicode bugs fixed, and then release asap [10:13] hi vila, welcome back [10:13] jml: same here (but different project ;) [10:14] otp with jelmer atm [10:14] ok [10:15] jml: that sounds good. I've got some ugly things I can do to make doctest work, can propose those and see how horrible you think they are [10:15] * mgz waves at vila [10:15] mgz: cool. [10:15] mgz: _o/ [10:16] mgz: now that I have an actual reliable way to run the tests across lots of Pythons, I feel much better about the whole thing [10:16] oh [10:16] never mind [10:17] well, that's also a good thing ;) [10:46] hi [10:46] how do i set up a bzr workflow where i can do local commits? [10:47] bzr unbind ? [10:47] Or just ci --local [10:47] I usually opt for the latter on my laptop when I'm on the train [10:47] then just push it later online [10:48] cheater: Isn't that the default? [10:48] With bzr branch? [10:53] the easy answer is to just branch and then you can commit locally [10:53] hi vila? [10:53] jelmer: For building subvertpy-dbg, and not having the #! lines sometimes point to the dbg python? [10:54] Anyway, I'm not too concerned, given I've already backported dh 7.3.15 to hardy .... 7.3 -> 8.1 can't introduce that many more requirements, right? :-) [10:55] mgz: I feel like I'm still missing something re bug 804127 [10:55] Launchpad bug 804127 in testtools "assertThat(..., verbose=True) sometimes generates unprintable AssertionErrors" [Medium,Triaged] https://launchpad.net/bugs/804127 [10:55] maxb, :) [10:55] hullo jml! [10:55] mgz: I guess Mismatch.describe() has a poorly defined interface, since it's not clear if it returns bytes or unicode [10:55] poolie: hi :) [10:55] ... "sometimes generates unprintable comments from developers" [10:56] poolie: hah [10:56] poolie: you'll like the new release actually, it's much less noisy. much better errors (when they work) [10:56] oh excellent [10:56] mgz and i were working on a branch to suppress some of the bzr-side noise [10:58] mgz: but perhaps you mean something else by not having a clear interface [10:58] gahr. I should reply to the comment. I always hate it when people take things to IRC & then don't update. [10:59] vila, hi? [11:00] jml: right, if it's not clear what describe is allowed to return, you'll risk mixing non-ascii byte strings and unicode strings [11:01] okay, I need to nip out for a bit now, will be around again later [11:01] mgz: cool. I'll update the bug report. Am keen to get this one nailed. [11:01] jpds, oh, i did bzr checkout. [11:01] thanks guys. [11:01] mgz: thank you again for helping! [11:02] cheater: bzr checkout creates a branch that's bound to the master branch, so every commit automatically goes there as well. bzr branch creates an independent branch. [11:07] night all [11:10] jelmer, gotcha :) [11:20] oops, looks like bzr-hg plugin is not happy with beta bzr (or maybe I need beta bzr-hg to go with it) [11:21] What's the quickest way to toggle a bzr plugin on and off? [11:22] moving it out of the plugin directory I guess [11:24] exarkun: yeah, either that or setting BZR_DISABLE_PLUGINS=hg as environment variable [11:49] jelmer: what happened to the idea of making tag fetching optional for 2.4 ? [11:50] jelmer: and more to the point should I delay the 2.4.0 release because of it ? [11:51] vila: hey, welcome back! [11:51] :) [11:51] vila: Did you have a good vacation? [11:51] yup, tons of hot weather [11:51] \o/ [11:52] I'm in Val's baggages visiting a friend in Bretagne where the weather is.... different ;) [11:52] Bretagne, nice :) [11:53] Well, if the weather is nice :) [11:53] vila: there hasn't been much progress on making tag fetching optional, as it's been pretty quiet around here [11:53] yeah, it rains only once a year [11:53] vila: I do think we really should make it optional before 2.4. [11:53] but it almost never stops ;) [11:53] vila: hehe [11:54] jelmer: ok, so I should postpone the release for a week then [11:55] jelmer: do we have a bug number for that ? [11:55] I don't think we do yet [11:55] ok [11:56] anything else I should be aware of ? (I haven't read all backlogged mail yet) [11:56] (but noticed babune being red in anger all over the place ;) [11:57] vila: Yeah :-( [11:57] vila: The other big news is that haproxy is now enable for codehosting \o/ [11:57] \o/ [11:58] there is one bzr bug related to that, but it's been present for a while so I'm not sure how important it is to fix before 2.4.0 [11:58] bug 819604 [11:58] Launchpad bug 819604 in Bazaar "when an idle ssh transport is interrupted, bzrlib errors" [Critical,Confirmed] https://launchpad.net/bugs/819604 === vila_ is now known as vila [13:07] vila: I vaguely recall you had some plans for running the tests against bzr.dev on babune, is that correct? [13:27] jelmer: err, babune is running all tests against bzr.dev :) Which tests are you reffering [13:27] referring to ? [13:27] vila: sorry, I meant plugin tests [13:28] ha, right, yeah, still planned, no ETA ;-} === gerard_1 is now known as gerard_ [15:19] hey all, I have a quick question about push-and-update plugin [15:20] I'm running bzr 3.2.1, and I moved the __init__.py and the push_and_update.py to my .bazaar/plugins [15:21] and no matter which transport I use, either bzr+ssh or sftp, I get This transport does not update the working tree of bzr+ssh://matt@192.168.12.164/etc/puppet/ [15:21] can anyone explain how to use this plugin? [15:26] ... please? [15:35] hi abetterlie [15:36] abetterlie: You need to add the plugin directory as a whole to ~/.bazaar/plugins [15:41] ah [15:41] ok [15:42] and it also needs to be renamed push_and_update, because - is not allowed for some reason [15:42] weird, becuase I have some precommit hooks that are bare .py files in .bazaar/plugins and they work just fine [15:43] That only works if $pluginname is entirely contained in $pluginname.py [15:44] I see. [15:46] when I push, bzr is still spitting errors, but the plugin seems to be working fine. [15:46] should I file a bug? [15:47] It'd be an enhancement, maybe. [15:47] bzr's push still doesn't do anything with the remote working tree. The plugin just makes it so when you run 'push', it does bzr's push and then _also_ does something else that does the update. [15:48] I could see arguments both ways about supressing the message. Not sure how easy it would be to do. [15:49] Hm. [15:49] https://bugs.launchpad.net/bzr-push-and-update/+bug/360521 [15:49] Ubuntu bug 360521 in Plugin to Update Remote Trees "Warning about working trees not updating not suppressed" [Medium,Fix committed] [15:49] That sounds like it should already be there. [15:49] ah [15:49] that's the one [15:49] http://pastebin.com/XU9ybkka [15:50] I don't see why it was ever changed from a seprate command push-and-update === beuno is now known as beuno-lunch [16:36] is there any way to disable SSL certificate checks in pycurl, or to enable authentication caching in urllib2? === beuno-lunch is now known as beuno === med_out is now known as med === yofel_ is now known as yofel [23:17] hi all [23:27] mørning poolie [23:30] moin [23:42] hello [23:44] hi grid