=== jamesh_ is now known as jamesh [00:11] anyone familiar with what ubuntu is going to do about nautilus and gio in hardy? gio isn't going to support smb:// or ftp:// according to the gnome folks. can gnome-vfs going to be backported or are we just going to have regressions when going from dapper to hardy? [00:11] * Fujitsu can use SMB in Hardy's Nautilus, and it seems to use GIO... [00:12] oh? [00:12] It's not exavrtly reliable, but... [00:12] well wtf was all that chatter on planet gnome? [00:12] There was something about it not supporting network:/// or similar a while ago. [00:13] aww [00:13] that's right [00:13] http://live.gnome.org/GioToDo says there is no Network: ftp: http/https/webdav: obexftp: burn: or fonts: [00:14] ftp is going to kill me when doing web dev :\ [00:14] Aw, no fonts:/// ? [00:15] I have a burn:/// too. [00:15] * burner shrugs [00:17] You aren’t using FTP with authentication but without encryption, are you? [00:18] * burner is [00:18] * burner does [00:18] * burner has a host that doesn't support sftp [00:18] hi :) [00:19] * Fujitsu would have hoped that such hosts ceased to exist a while ago. [00:19] You aren’t paying anything for the service, are you? [00:19] hi caetano [00:20] hi [00:20] hey, is there someone here responsible for ubuntu's kernel stuff? [00:20] thotypous: #ubuntu-kernel [00:21] thanks [00:21] hey maskd, #ubuntu-kernel [00:27] thotypous: burro [00:27] nao vai falar do oss? [00:31] heno: I have uploaded umenu code to https://code.launchpad.net/umenu [00:31] make prerequisites && make && make test [01:02] aclocal: macro `AM_PROG_MKDIR_P' required but not defined [01:02] * ogra sighs loudly === Martinp24 is now known as Martinp23 [01:31] ogra: automake does tend to make one sigh, doesn't it./ [01:32] yeah [01:33] especially if someone "fixed" your code to use a newer version instead of being compatible with the most common one .... [01:40] any ubuntu-main-sponsors want a quickie? [01:40] (upload, that is) [01:44] ogra: but newer is always better, no? [01:45] jdong: In a few months I might. :p [01:45] but not just yet [01:45] TheMuso: :) that's no help for a feature freeze ;-) [01:45] * jdong thinks transmission should stay in universe until FF! [01:46] jdong: Its called, just got the keys to the world, and don't want to break anything. [01:46] TheMuso: haha well you have my word that it won't break anything :) [01:46] if that means anything :D [01:46] No, I'm not even on the main sponsors team yet [01:47] oh well I'll keep /pounce-ums active then [01:48] lamont, yeah, but some people use changelogs to let other devs know it changed ... ;) [02:01] ogra, i missed the discussion with you and laga on the patches that he had for mythbuntu-diskless against ltsp, what came of that? [02:02] superm1, the plugin is included in ltsp since today, his initramfs changes have to be merged into the mythbuntu-diskless-client package [02:03] there are some changes i''d like to see in the code before final but all in all it looks fine [02:03] ogra, okay so everything should be fine on it then for this cycle? [02:03] yup [02:03] ogra, great, thanks :) [02:16] sigh. jdong did you get taken care of while I was translocating? [02:16] lamont: you mean my call for sponsorship? not yet [02:16] what is it you need sponsored? [02:16] bug 190614 [02:16] Launchpad bug 190614 in transmission "New upstream version: 1.05" [Undecided,New] https://launchpad.net/bugs/190614 [02:16] if it's not too complex, I suppose I could be talked into it... [02:16] hrm... [02:16] it's a trivial and routine one [02:17] I've been the one who's done like the past 4 or 5 [02:18] where does $UPSTREAM_VERSION get used? [02:18] or rather, where should I grab bts from to look at them? [02:18] lamont: it's used in Debian's get-orig-source rule [02:18] lamont: which is redundant and there for historical purposes, as right now a watchfile is used [02:19] (before DFSG needed to do some repacking) [02:19] ah [02:20] * lamont fetches [02:20] * lamont wants to know how to get gutsy's network managler to automatically notice that and switch ssids without me doing "manual config" [02:22] lamont: it doesn't detect your teleportation? [02:24] something like that [02:24] lamont: are you sleeping to and from your destination? [02:30] I removed network mangler and reinstalled it (long story_) [02:30] and right now all i have is 'manual config' [02:33] * lamont dist-upgrades his hardy chroot so he can test build [02:35] does anyone know why libexecdir defaults to /usr/libexec, even that dir usually doesnt exists in debian/ubuntu and is usually pointed to /usr/lib ? [02:35] i would expect it to use a default that actually works with my system [02:36] ogra: historical linux? [02:37] silly ... i would expect tools that have "auto" in their name to set the best defaults for the current system i'm on :) [02:38] ogra: bug keybuk, I guess... [02:38] well, likely asleep :) [02:39] and your excuse? [02:39] freeze tomorrow, and insanely early meeting :) [02:39] oh. meh [02:40] (well, insanely early for my biorhythm ... others like to be up at 8am :) ) [02:42] feature freeze or so? [02:42] yep [02:42] which goes hand in hand with UVF nowadays [02:42] ok. /me has an upload pending for bind9 - but it's more of a bug fix, IMO [02:43] otoh, we're two revs back [02:43] well, it wouldnt be a freeze if theer werent exceptions [02:44] * lamont files a sync request, having noticed that it's out of date [02:57] jdong: you didn't say it build-depends: $WORLD [02:58] lamont: sorry :) [02:58] lamont: you could take my word that it builds fine in a hardy pbuilder [02:58] !jdong [02:58] jdong: yes, but you're FULL OF CRACK! [02:58] ;-) [02:59] If it's transmission though, I'd be inclined to believe him. [03:00] heh [03:00] Get:91 http://us.archive.ubuntu.com hardy/main gettext 0.17-2ubuntu1 [1978kB] [03:24] jdong: transmission_1.05-0ubuntu1 uploaded [03:24] lamont: thanks! [03:24] lamont: now it'll hurt less the next time you do it for me :D [03:24] lol [03:25] (luckily for you there's FF) [03:25] yeah - my build chroots hadn't been set up yet on the laptop [03:25] what's FF? :-) [03:25] I will _not_ upload firefox for you [03:25] * lamont ducks [03:25] Hah [03:26] lol [03:26] :) [03:26] Is jdong wanting to crackport firefox? :-P [03:26] I wonder if anyone has ever done stats on who's sig is most common on new upstream versions in the archive post-feature freeze... [03:26] StevenK: what else could FF possibly mean? :-) [03:27] lamont: "Feature Freeze" [03:27] StevenK: you do spoil trolling, you know... [03:27] lamont: seb128 perhaps? [03:27] lamont: :-) [03:27] LaserJock: almost certainly [03:27] StevenK: ff-3 crackports already done btw ;-) [03:27] jdong: Argh [03:28] * lamont notes that is "just do the right thing, dammit' build-package script is now 250 lines of shell [03:28] StevenK: it was done responsibly and safely though ;-) [03:28] I should really consider rewriting that in something else. [03:29] lamont: Blink? [03:29] 250 lines to do what? [03:31] could one of you many core-devs around sponsor a small patch to acpi-support? bug 123104 [03:31] Launchpad bug 123104 in acpi-support "Suspend using nvidia driver and NVIDIA 8400M GS doesn't work" [Undecided,In progress] https://launchpad.net/bugs/123104 [03:32] * ScottK is pretty certain that's not where I'm going to start doing Main uploads.... [03:32] ScottK: _I_ won't touch an acpi upload [03:33] ScottK: Did you get core-dev? [03:33] StevenK: mostly magic sed commands to smash build-deps around [03:33] StevenK: I did [03:33] StevenK: he's a real person now [03:33] ScottK: Nice! Congrats [03:33] superm1: we still use acpi-support? [03:33] lamont: Heh [03:33] congrats ScottK, i wasnt aware of this either [03:33] lamont: ... Why? [03:33] StevenK: Thanks. [03:33] superm1: Thanks. [03:33] StevenK: he's a real person because he's cool. [03:34] ScottK: congrats! [03:34] acpi? scary - too easy to break random crap elsewhere with what seems like a simple fix. [03:34] lamont: Sorry, that "... Why?" was due to "mostly magic sed commands to smash build-deps around [03:34] jdong, I was under the impression that part of it was still used [03:34] because sometimes I backport things from hardy to dapper. [03:34] to build the modules that need to be taken out [03:34] Crazy [03:34] superm1: that may be. Sounds a bit silly though [03:34] jdong: acpi-support is on my freshly-installed gutsy box [03:34] superm1: I'd talk to mjg59 first. [03:35] superm1: and pm-utils's taking out modules routine actually does absolutely crap [03:35] jdong: THanks [03:35] lamont: hardy's using pm-utils though :) [03:35] ah, ok [03:35] thanks StevenK i'll check with him [03:35] superm1: I filed a bug and a debdiff on that ;-) [03:35] one of these days I should do more playing with hardy... [03:35] lamont: it looks very promising as a solid release :) [03:36] mostly ScottK just tells me to upload stuff, and I do.... [03:36] no one wants a solid hardy release more than me [03:36] backporting from hardy to dapper, um, sometimes really really sucks [03:36] lol [03:38] only 10 sed commands (admittedly, it could be 2, since all but one just tweak debian/control [03:38] I think my favorite has to be: [03:38] sed -ri "$(grep-dctrl -P -sPackage -n python- ${CTL} | [03:38] sed 's/$python$$-.*$/\/^[^ ]*(Pac|Dep|Sug|Pro|Con|Rep)\/s\/\0\/\12.4\2\/g;/')" ${CTL} [03:38] Please. I just ate. [03:39] yikes [03:40] wow [03:40] what's wrong with having sed generate a sed program? [03:40] have I mentioned that I love python, and I _HATE_ python-packaging changes? [03:40] lamont: A few things. Mostly readability and maintainability? [03:41] StevenK: piffle [03:42] sed -i '/^Build-Dep/s/libdb-dev (>=4.6.19-1)/libdb4.4-dev/g' ${CTL} [03:42] that one is nicer. [03:43] see, this way I just say 'build-package -bfeisty postfix_4.7.1-1' and it does the right thing, and generates binaries for testing, so that I can sign and upload the source. [03:43] admittedly, the dapper backports I'm doing have a very targeted uh, target. [03:43] lamont: why don't you just write a pbuilder-satisfydepends that's more lenient? :) [03:43] lamont: then I can steal parts of that for backports usage too [03:44] jdong: because I'm uploading these to a pedantic archive [03:44] lamont: Is the meat of build-package calling sbuild or pbuilder? [03:44] lamont: lovely :) [03:44] sbuild. [03:44] what's pbuilder? [03:44] :-) [03:44] Haha [03:44] * lamont has only been using sbuild for most of a decade... [03:44] I keep saying that I should really take time and learn pbuilder... it never makes it above the cutline though... [03:45] Depends how long your machine takes to untar a 80Mb base tarball [03:45] lamont: pdebuild FTW [03:45] lamont: combined with bzr builddeb, its fun [03:46] a major improvement in the script came recently when I taught it to notice that I'd forgotten the epoch in the version, and just rip that from changelog [03:46] I'm going to stand up for sbuild, and say that keescook is my hero and that mk-sbuild-lv rocks [03:46] lifeless: I've used bzr far more in the last five weeks than ever before. [03:46] lamont: thumbs up or down [03:46] I'm building a list... have filed a bug or 3 [03:47] well, _a_ bug, iirc. [03:47] I think that I have more [03:47] lifeless: I did make my life a bit simpler by making sure that I'm using bzr 1.1-1 everywhere, instead of 0.8.$mumble [03:48] 0.8.2-1ubuntu3 to be specific [03:48] and myriad versions in between... [03:49] bzr uncommit? +1 [03:49] any bzr command that generates copious output not doing the right thing (piping to $PAGER)? -1 [03:50] mostly thumbs up [03:50] lifeless: you guys got a "bzr for git users" doc somewhere? [03:51] LaserJock: the wiki, possibly the user manual now [03:51] lifeless: give me an easy way to bzr import from active git repo, and push commits both ways [03:52] LaserJock: I' [03:52] ve mostly been getting the crash-hands-on course [03:52] I wish there was a "git for bzr users" -- I really like bzr, but can't get my head around git [03:52] see the tech talks [03:52] most notably randalls [03:53] StevenK: I think I might be the opposite [03:53] I don't think I got much out of bzr [03:53] but I find git quite nice [03:53] bzr just works like cvs and svn, so I just clicked. [03:53] git just looks oddball [03:53] StevenK: bzr doesn't work like CVS. CVS doesn't work. [03:54] StevenK: that's what happens when someone designs the plumbing and doesn't care about the porcelain [03:54] Fujitsu: Sure it works, it just doesn't do some things well [03:54] git prior to 1.5? sucky UI that is not recommendable for anyone other than kernel gods [03:54] Heh [03:54] Well, even Rusty can't work out git [03:55] rusty? [03:55] I really like git-cvs [03:55] LaserJock: parsecvs for the win [03:56] although you need sane ,v files for that [03:56] lamont: Rusty Russell [03:56] lifeless: yes, postfix broke parsecvs too === ember_ is now known as ember [03:57] merge speed in git is rather incredible, I must say [03:57] I think the part of bzr that's gives me problems, and perhaps what I like about git, is the "branch as dir" concept [03:57] ?? [03:58] I just don't like having directories all over so I tend to not branch very much in bzr [03:58] whereas in git my dir structure stays clean even if I have branches all over [03:58] * ogra was proposing a branch as pizza function since ages, but noone heard [03:58] * lamont tends to bounce between branches in one working directory... if I really really need more than one dir, I can clone the branch I want into a sisiter dir [03:59] LaserJock: Where as I'm completely the opposite. git's hidden branch approach smacks of code-simplicity over interface tradeoffs to me. [03:59] StevenK: you dont want to get your head around git [03:59] * lamont isn't sure about this "branch as dir" thing... maybe I need to actually read the users manual... [03:59] RAOF: hidden branch? it's a file. [03:59] you want it in a separate dir, you clone [03:59] lamont: problem is, plumbing sets constraints on porcelain [04:00] lamont: And then checkout. Using strange and arcane syntax. [04:00] yeah... that's why it wants to be pretty static from the beginning [04:01] RAOF: of course. clone is well, clone [04:02] debian #464876 [04:02] Debian bug 464876 in drupal5 "drupal5: New upstream version 5.7 available, fixes several bugs" [Normal,Fixed] http://bugs.debian.org/464876 [04:03] git clone postfix test; cd test; git rebase origin/stable/v2.4 [04:03] it's not _so_ arcane... [04:04] of course the downside is that 'master' in test is then stable/v2.4 in the original (postfix) tree [04:04] just remember: source management systems all suck. differently. [04:05] lamont: branches in a file in .git -> not discoverable; complex rules to edit, can't use standard shell commands to manage branches, urls are tricky to define well and usefully. [04:07] actually, .git/refs/head/$branchname is a file... containing the sha1 of the head of that branch [04:08] and git-branch -a will show all of them [04:08] hrm.. does dapper dpkg have anything like source:Upstream-Version? [04:13] nope. [04:13] sigh [04:15] lifeless: moblin are using git, so I have to at least learn a bit [04:15] StevenK: seriously, watch randall's tech talk [04:16] once you understand git's "index" (which every source management system has, and almost none export to the user's view), then it all falls into place. [04:16] until then, it's confusion [04:17] randall does a good job of giving a tech overview of git. [04:17] lamont: Google for "randall tech git talk" or so? [04:17] lifeless: bzr bisect would be cool to have. [04:17] StevenK: yep [04:17] google for "git tech talk" (with quotes) and (last time i did it) there were two results: linus and randal [04:17] randall [04:17] even [04:18] lamont: https://launchpad.net/bzr-rebase [04:19] lamont: exposing the index is not done by most systems for a reason [04:19] lamont: its like exposing inodes to the user (and I don't mean C programs) on a file system. [04:21] lifeless: yeah [04:22] lifeless: You know, it seems like all the features I want in bzr are already plugin projects on LP. Are they going to get an actual release sometime soon? Please? :) [04:22] RAOF: what do you mean [04:23] lifeless: yeah... the index does tend to confuse the hell out of people without providing them the benefit that understanding it would given them. [04:23] s/given/give/ [04:23] Well, there seem to be a lot of useful bzr plugins on LP; things like bzr-git, bzr-rebase, bzr-bisect, etc. But none of them seem to have released a tarball. [04:24] Or been packaged in some other form useful to me. [04:24] lamont: s/would/might/ [04:24] lamont: most users won't get any benefit from the index [04:26] yeah [04:26] they don't /want/ a current state that differs from the working tree in any wa. [04:26] the index being exposed is one of those things you get when a kernel jock designs an SCM. [04:26] * lamont uses it as a 1/2 commit :0) [04:29] RAOF: so package and upload them [04:29] RAOF: join the bzr-packaging team [04:29] RAOF: tarballs are so 1980's [04:29] lifeless: is there such a team? [04:30] Oh, but that's *work*. Worse, it's work that I have to do! :P Ok, added to my TODO. [04:30] LaserJock: its on alioth [04:30] lamont: Googling only shows Linus' talk [04:30] StevenK: the first one is the one you want [04:30] StevenK: it just says "git" [04:31] http://www.youtube.com/watch?v=8dhZ9BXQgc4 [04:31] gah [04:31] gimme a second [04:32] http://video.google.com/videoplay?docid=-3999952944619245780 [04:33] * StevenK clicks the yummy download link [04:33] RAOF: basically, with a vcs every push is a release; folk really stop bothering with tarballs in smaller projects. too much overhead, and too slow. [04:33] StevenK: any of the october versions are Randall. Linus was a few months earlier [04:34] lifeless: overall, I'm very impressed with bzr [04:34] lifeless: except it's hard to distribute and is much harder to find a "stable" release [04:34] lifeless: Fair enough. Bzr builddeb should make packaging from bzr more enjoyable. [04:34] * backport to dapper [04:34] * source:Upstream-Version expanded inline [04:34] there. now build-package deals with packages that use source:Upstream-Version [04:34] lamont: good :) [04:35] wc /home/lamont/bin/build-package [04:35] 254 802 7371 /home/lamont/bin/build-package [04:35] LaserJock: uhhh, its trivial to distribute, and tarballs have no magic equality with stability. [04:37] lifeless: it's not trivial to distribute until we get rid of source packages, and no, there is no "magic", but it's more "conventional" === calc_ is now known as calc [04:41] lifeless: for anything that's going into a debian/ubuntu/whatever release, trivial packaging of tar+diff is kinda a requirement [04:41] or at least a tarball [04:41] lamont: 'bzr export' [04:41] yeah - it's trivial for us'ns [04:42] I was aluding back to the bzr builddeb [04:42] which I freely admit to never having used [04:42] bzr-builddeb is great. [04:44] "which I freely admit to never having used" [04:44] that's coolest sentence construct of the day! [04:45] jdong: cool sentence construction is just one of my more flashy skills [04:45] I haz gud wordspeak [04:45] in ur channelz hacking ur brainz [04:46] 'zactly [04:46] lifeless: so when are we all switching to RCS? :-) [04:46] my kids go to RCS. [04:46] hrm. [04:46] I think that's different [04:47] [04:47] lamont: you you have to pick them up and drop them off individually? [04:47] :) [04:47] And woe betide if your children change names! [04:47] actually, their start/stop times _are_ staggered 15 min... so the one just gets dropped off early and picked up late [04:48] www.ridgeviewclassical.com [04:48] A classical school. [04:48] StevenK: _very_. [04:48] Damn, I was going somewhere with that, and then my thought train derailed [04:49] Oh yes. [04:49] A classical school, as opposed to you know, *regular* boring school. [04:50] to make everyone feel better.... [04:51] at the MIT software engineering course, I am being FORCED to write JAVA CODE in a PAPER NOTEBOOK. [04:51] of course, 0.5 pages of code later, it turned into a 3 page argument why this is nonsensical :) [04:51] I hate doing that in exams. [04:52] take a different course [04:52] StevenK: it's more around changes in the modern education ideas [04:52] which the school kinda, uh, rejects. [04:52] lamont: Like computers? [04:53] like student-desire-drivin curriculum. [04:53] actually, the school teaches openoffice to everyone === asac_ is now known as asac [04:53] Pah, we never had that when I went to school [04:53] it's a mindset, not specific curriculum. [04:53] curricula? [04:54] curriculae [04:54] curriculum [04:54] tardy? no problem. sit right there on the bench until the period is over. you will not interrupt the class and waste 1+ student-hours (30 students, 2 min or so disturbance) [04:55] little things like that. very teacher-centered education [04:55] lamont: do they get their drugs from CVS? [04:55] we actually don't have any CVS stores here. iz very sad. [04:55] I nearly fell over when I saw the CVS store in Boston [04:56] And then on Friday, pitti and I saw a CVS import [04:57] StevenK: that just hurts [04:57] hrm... closing time... I guess I should start wrapping up [04:58] dear kde. 70 MB is not a deb. it's several. [04:58] kthx [04:58] OK, not sure if this is some dev issue, but here's a weird problem. In 2.6.22-14-generic my usb based rtl8187 works fine but kernel of course can't see my full 8Gb. In 2.6.22-14-server the RAM gets seen but the nic is no good. Even tried compiling from svn from rtl-wifi but getting indecipherable compiler errors on all svn versions 61 thru 67.The exact same code compiles without a groan under -generic [04:59] Some weird PAE thing perhaps? [05:01] * genii hands SNuxoll a coffee [05:02] yay caffiene [05:10] * lamont relocates homeward [05:35] hey guys, does anyone know if the sqlite3 in the gutsy repos is compiled with the --threadsafe option? [05:38] talcite, it would seem from the build log that yes: http://launchpadlibrarian.net/9193582/buildlog_ubuntu-gutsy-i386.sqlite_2.8.17-2.1build1_FULLYBUILT.txt.gz [05:39] Any core-devs involved with Mono here? [05:39] it has the "-DTHREADSAFE=1" bit in it [05:41] beuno: thanks [06:10] Good morning [06:22] * lamont gets home, sleeps. night all [06:23] Night lamont. [06:27] Fujitsu, did you check PM? [06:45] superm1: ping? [06:45] yo [06:46] superm1: hey.. i think you should plan to change the Depends: mysql-server to Predepends [06:46] on which? [06:46] er which package particularly at least [06:46] superm1: problem on upgrade is simple.. mythtv wants to backup and update the db, but if mysql is upgraded too, it's not running.. failing to backup and update [06:46] superm1: dunno.. you can figure that out yourself :) [06:47] ah for folks going from dapper->hardy or gutsy->hardy [06:47] at the moment i am just doing hardy->hardy regular updates [06:47] and running into that particular issue i suppose? [06:47] problem occours when there is a major myth db update and mysql update at the same time [06:47] so that will be the case also on upgrades from release to release [06:47] or anything that upgrades to hardy [06:48] superm1: yes.. right now [06:48] thegodfather, okay i'll see what i can do, you should see it in the next hardy update. let me know if that works out better for you then [06:49] superm1: well it can be simulated if you need to [06:49] superm1: just prepare a fake mysql update in your local repo and a mythtv update at the same time [06:50] yeah good point [06:50] superm1: make sure that the mythtv update simulate a mysql DB update and zack [06:50] well i'm fairly confident that you're right on this too, pre-depends should solve it. === calc_ is now known as calc === hunger_t_ is now known as hunger [07:48] * Mithrandir grumbles at his /dev/sda being renamed to /dev/hda [07:49] lol, really ? [07:50] * ogra has never seen it this way around [07:51] happened between -5 and -7 in hardy. [07:51] unsurprising, this breaks libpam-mount [08:01] good morning [08:10] moin [08:12] Stefan Bader is not hanging around here, is he? [08:17] are there any formal critera to get a patch into the ubuntu kernal? [08:18] anybody? [08:20] alright, I'll just post to launchpad [08:20] quit [09:06] Do MIRs need to be done by FF? === \sh_away is now known as \sh [09:07] https://bugs.edge.launchpad.net/ubuntu/+source/linux-source-2.6.24/+bug/190587 whats up with this security risk [09:07] Launchpad bug 190587 in linux "Local root exploit in kernel 2.6.17 - 2.6.24 (vmsplice)" [High,Fix committed] [09:08] hi, [09:08] for the sake of 'history', i'd like to have all ubuntu revisions ever of my packages in debian. [09:09] is there something like snapshot.d.n for ubuntu? === \sh is now known as \sh_away === \sh_away is now known as \sh [09:15] panthera: you should be able to fish them out of links in Launchpad, at least from dapper onwards [09:16] start from https://launchpad.net/ubuntu/+source/PACKAGENAME [09:18] ionstorm: seems to be in pretty good progress [09:24]