=== Snova_ is now known as Snova [00:01] kirkland: do you remember what's the state of acpi in jaunty? Is acpid supposed to be running? [00:02] kirkland: it seems that the jaunty installer doesn't install acpid by default anymore [00:02] kirkland: which explains why virsh shutdown vm-id doesn't work [00:02] mathiaz: i have it installed on my desktop [00:02] mathiaz: perhaps it got dropped as a recommends somwhere? [00:02] mathiaz: and we need to re-add it to the server seed? [00:03] kirkland: IIRC there is some code in the installer that checks if /proc/acpi/ is available and if so installs acpid. [00:03] yes, in hw-detect. It only works if it's on the CD. [00:03] there was substantial conversation recently about acpid being obsoleted by other packages [00:03] slangasek might remember more [00:04] cjwatson: oh. And acpi is not the -server isos. [00:04] strictly speaking, what's the difference between a "package archive" and a "repository"? [00:06] cr3: do you believe that there is a difference? [00:06] i thought older packages are kept in the archive [00:06] mathiaz: acpi != acpid [00:06] cjwatson: not to my knowledge, but I was wondering if these were strictly synonyms or whether there might be a slight difference [00:07] thiebaude: that would be a pretty confusing term to use for Ubuntu given that it's archive.ubuntu.com [00:07] cr3: they're both somewhat fuzzy terms but I would regard them as essentially synonyms, unless perhaps there's some clarifying context [00:07] cjwatson: for example, maybe archive.ubuntu.com/ is the repository and maybe ubuntu/ is the archive. just thinking out loud [00:07] no [00:07] cjwatson: well - neither acpi, nor acpid is on the -server iso. [00:07] then current packages are not kept in a archive :) [00:08] thiebaude: current packages are, as it happens, kept in an archive. [00:08] "archive" and "repository" are two more-or-less synonymous terms for something you point apt at. [00:08] kirkland: I think acpid should be put back on the -server iso. [00:09] kirkland: IIUC it's an important component of the suspend/resume for server spec. [00:09] except that isn't quite accurate because you actually point apt at a single distribution and a set of components within an archive, in each sources.list line [00:09] cjwatson: might it make a difference when you point apt to a series (.../ubuntu jaunty main) and when you point apt to a directory (.../ubuntu ./)? [00:09] cr3: man sources.list [00:09] mathiaz: yeah, i concur [00:09] cjwatson: i think i'am starting to understand [00:10] mathiaz: you wanna do that? or did you want me to? [00:11] kirkland: you can do it - I was wondering why acpi stopped working in my freshly installed jaunty kvm guests. [00:11] kirkland: you may wanna wait afte beta though. [00:11] cr3: which uses "archive" consistently for both meanings, but AFAIK you could replace "archive" with "repository" throughout and get the same meaning [00:11] kirkland: I don't think it's a beta blocker. [00:11] slangasek: when would it be safe for us to add acpid back to the server-ship seed? [00:12] cjwatson: ok, thanks for the clarification [00:13] thiebaude: Debian has archive.debian.org, which holds archived versions of no-longer-current Debian releases. The name clash with archive.ubuntu.com is perhaps slightly unfortunate, but we were aware of it at the time and decided that it was not terribly important [00:14] thiebaude: the equivalent for Ubuntu is old-releases.ubuntu.com [00:14] and old versions of packages that weren't necessarily final versions in a given release are kept on Launchpad [00:14] mathiaz: it only needs to be in server-ship, right? [00:14] Sigh. Bugs should not have a user-visible importance. [00:14] kirkland: yes [00:15] It just leads to people making irrelevant complaints about their bug having the "wrong" importance. [00:15] (I'm trolling slightly ...) [00:27] kirkland: "back"? [00:28] kirkland: ah, so it's available for apt-install. Mmm, do it now and I can get it included for beta. [00:29] mathiaz: ^^ [00:29] it's not a beta blocker, but I've just now rerolled ubuntu-server and can do so again if you want this in [00:35] slangasek: committing ... [00:36] mathiaz: slangasek: committed and pushed [00:36] Committed revision 1449. [00:36] ok [00:37] starting the CD rerun [00:37] is notify-reboot-required still the right way to request a reboot? [00:41] seems so [01:09] So, if one wants to get their feet wet in Ubuntu-devel, what's the language of choice to learn? [01:10] "Makefiles" [01:12] ebroder: Was that meant for me? === Snova_ is now known as Snova [01:14] TwoToneSpirit: Yes, but somewhat jokingly [01:15] TwoToneSpirit: Makefiles are the core of Debian/Ubuntu packages, although there are usualy several layers abstracting that away [01:15] Ubuntu uses Python fairly extensively for a lot of the custom development that's done [01:16] I currently regard PHP as my primary language [01:16] That being said, there are packages that use just about every language imaginable - I doubt there are many languages that wouldn't be useful [01:16] TwoToneSpirit: I find Ubuntu development is a heck of a lot easier if you're prepared to work in several different languages [01:16] I hear that Python is a natural next step from PHP for many people [01:16] people do ask this question a lot here, but there is no good answer to it [01:16] cjwatson: Well that's why I asked about getting my feet wet :-) [01:17] Python is used a lot for many things in Ubuntu, but in order to work with packaging you'll also need the rudiments of GNU Make and the Unix shell [01:17] i finally have all my upstream bugs linked :) [01:17] and familiarity with C, Perl, and maybe C++ is unlikely to hurt either [01:17] it's much more important to have the skill of picking up languages than it is to know any particular language, I find :) [01:18] as ebroder says, distributions are made up of a wide variety of software written in nearly every language under the sun [01:18] aiui some languages are more evil (lisp) than others, heh [01:18] The best language to learn for packaging, is Telepathy :) [01:18] Aww! I <3 lisp. Well, Scheme [01:19] there are mini-languages in Ubuntu that exist for a single purpose === gmaculo_ is now known as gmaculo === Snova_ is now known as Snova [02:50] yippee, 0 new, 1 confirmed (can't test yet), and only 10 targeted bugs for 9.04 :) [02:50] for OOo [02:52] calc: Cool. [02:55] TheMuso: i can't believe i managed to completely catch up on triaging :) [02:56] calc: That's unpossible [02:56] * Amaranth files 40 more bugs targeted to 9.04 [02:56] Amaranth: :-P [02:57] * Amaranth thinks calc is just closing bugs [02:58] Amaranth: nope, been working around the clock :\ [02:59] Amaranth: https://edge.launchpad.net/ubuntu/+upstreamreport <- also shows the amount of bugs ;-) [03:00] only 426? that's not bad [03:00] and most of them are upstream, nice [03:00] for some reason the number of bugs not triaged shows as 36 but when you click on it you can only see 35 [03:00] i wonder if one of the bugs is hidden from me for 'security' reasons [03:01] you aren't in the ubuntu-bugs team? [03:01] yea, but even then you can't see security bugs until subscribed [03:01] there is a higher level elite launchpad bugs access level [03:01] calc: need me to find something? [03:02] kees: yea look at https://bugs.edge.launchpad.net/ubuntu/+source/openoffice.org/+bugs?search=Search&field.status=CONFIRMED&field.status=INCOMPLETE_WITHOUT_RESPONSE&field.status=INCOMPLETE_WITH_RESPONSE&field.status=NEW do you see 36 bugs or 35? [03:02] kees: upstream bug report claims there are 36 but it only shows me 35 [03:02] calc: I used to be in some team that had access to security bugs [03:02] kees: if there is something that is marked special i probably need to be subscribed to it [03:02] calc: I see 35. Perhaps upstream report is out of date? [03:02] thought it was ubuntu-bugs [03:02] kees: maybe so, i thought it was always up to date [03:03] hm, dunno [03:03] i can see if i can find someone on the launchpad team to check it out for me [03:04] kees: thanks for checking for me :) [03:05] calc: sure, no problem. [03:05] Amaranth: bug-control can see private-not-CVE bugs [03:05] dunno about CVEs though [03:05] that was it [03:05] I stupidly let it expire [03:05] maco: aiui bug-control can't see private bugs marked as security, not just cve bugs [03:05] along with motu [03:06] by definition private bugs are only visible to their subscribers [03:06] oh ok [03:06] ive never seen a security one that didnt say cve [03:06] cjwatson: ah so it could be a bug on OOo is private that does not have bug-control or security team on it? [03:06] this is not to say that ubuntu-bugcontrol or some related team is not subscribed to some private bugs [03:06] cjwatson: bug control can see them too to check the crash reports [03:06] calc: entirely possible [03:07] cjwatson: you don't happen to have master access or something like that do you? :) [03:07] private bugs are private -- only subscribers can see them. most private bugs start their life as private security bugs, and as such, security is subscribed to them, even after they're unmarked security and only the security team can unsubscribe the security team, so we tend to be able to see any private bug. [03:07] maco: only because apport explicitly subscribes "crash bug triagers for main packages" or whatever it is to those bugs [03:07] lag [03:07] calc: no [03:07] cjwatson: ok [03:07] cjwatson: ooo ok [03:07] calc: launchpad.net/~admins [03:07] cjwatson: ok [03:08] calc: but actually if you want to know, answers.launchpad.net/malone is probably a better way to ask [03:08] ok === Mez_ is now known as Mez [06:41] good morning [06:46] good morning dholbach [06:46] hiya highvoltage [06:46] morning [06:46] hiya lool [06:46] Hi dholbach, how is it going? [06:46] I'm waking up - how 'bout you? :) [06:46] It's about the opposite: I was up so I came to the computer :-) [06:46] dholbach sometimes sounds like a lyrics generator :) [06:47] highvoltage: really? I didn't know that was one of my strengths :) [06:47] learn something new every day :) [06:47] how are you doing highvoltage? [06:47] oh it is. I might ask you if I can use some of it one day :) [06:50] dholbach: I'm just about to finish a second big deadline that's been a denial-of-service on me for the last 2 months or so [06:50] then I can return back to a state where I can actually relax and do normal things again :) [06:50] dholbach: but it's going well thanks [06:50] dholbach: I guess I'm slightly jealous that summer is leaving us and going over to your hemisphere now :) [06:50] highvoltage: slowly, very slowly - it's 2°C here [07:04] Good morning [07:04] mdz: hal-info> indeed; I commit them upstream right away, and every now and then I package a git snapshot [07:04] mdz: right now we are frozen anyway [07:09] does the apport-retracer process a bug if it's already been marked as a duplicate of another bug? [07:16] sbeattie: no, those are ignored [07:18] pitti: okay, thanks. [07:19] pitti: BTW, it'd be useful to extend apport-collect to take a crash file as an argument. [07:20] sbeattie: that would attach the info in the .crash file to a bug report? [07:20] sbeattie: I'd rather have apport-{gtk,qt,cli} do that, or a separate script [07:21] sbeattie: there's an existing bug report for that, too [07:21] pitti: have apport-{gtk,qt,cli} attach to an existing report rather than always create a new one? [07:21] something like that [07:22] well, I don't particularly mind which does which [07:23] pitti: sure, I'm not picky about the solution; I just dislike currently creating a duplicate bug so I can do the apport collection when I've recreated a bug on something that didn't get one to begin with for whatever reason (in this case, failure of mdadm while booting) === bluesmoke_ is now known as Amaranth === tkamppeter_ is now known as tkamppeter === bigon_ is now known as bigon [08:18] cjwatson: why do we now use LABEL by default? [08:19] uh? [08:19] setting labels, or referencing them? [08:20] I assume he means in /etc/fstab [08:20] mounting-by-LABEL instead of mounting-by-UUID [08:22] I'm entirely unaware of this design decision having been revisited [08:22] Keybuk|slangasek: that reminds me, is bug 347370 actually a bug or just an expected change in hal behavior? [08:22] Error: Could not parse data returned by Launchpad: The read operation timed out (https://launchpad.net/bugs/347370/+text) === smb_tp_ is now known as smb_tp [08:22] last I knew, Debian was transitioning to also mount by UUID in squeeze [08:22] slangasek: my reference is bug #347685 [08:22] Launchpad bug 347685 in initramfs-tools "breaks with labels containing slashes" [High,Triaged] https://launchpad.net/bugs/347685 [08:23] Colin Watson said "Since we now use LABEL= by default if filesystem labels are set" [08:23] ah, "if filesystem labels are set" [08:23] Sorry, try that again; https://bugs.launchpad.net/ubuntu/+source/hal/+bug/347370 [08:23] Ubuntu bug 347370 in hal "REGRESSION: volume labels have underscores instead of spaces" [Low,Confirmed] [08:23] so they're not set by default, but referenced if someone sets them, hmm [08:24] right, but the reason we never did that originally is because filesystem labels are trivial to clash [08:24] if you have two disks with LABEL=root [08:24] yes [08:24] it's not just that the wrong one might be mounted [08:24] I don't know why we're doing that, then [08:24] but that the one that is mounted is picked AT RANDOM [08:24] but it worries me a bit less than the idea that we were doing this on all new installs [08:24] it's fine if the user uses LABEL= themselves, and puts that in fstab themselves (which is why we've always supported it) [08:25] but any automatically generated file imo should never use it [08:25] sbeattie: no clue if it's expected, that's pitti's department :) [08:25] even if the filesystem has one [08:27] anyway, you'll probably have to wait a few hours to get an answer from cjwatson, he was up late last night handholding a beta-critical debconf fix [08:41] TheMuso: around still? [08:41] TheMuso: I have an issue with pulseaudio, and I'm trying to find out who's the culprit [08:42] TheMuso: It regularly exits with "E: cpulimit.c: Received request to terminate due to CPU overload." particularly on sites with flash; there's a wiki page on the pulseaudio wiki to debug such issues which suggest it might be driver related; http://pulseaudio.org/wiki/BrokenSoundDrivers however doesn't work for me as the proposed test case spawns pulseaudio for me [08:43] slangasek: X didn't start with the valgrind madness, so had to revert. [08:43] In fact even "aplay" spawns pulseaudio by default when sending to hw:0; this seems to be a bug in /usr/share/alsa/pulse.conf [08:43] aw [08:43] Nafallo: do you have an error message? I should fix up the recipe if it's wrong [08:43] (well, otherwise I'll find out here next time my X crashes, since I've already done the same) [08:44] ==3295== [08:44] ==3295== Warning: Can't execute setuid/setgid executable: /usr/bin/X.valgrind-madness [08:44] kirkland: oh, btw, was /home encryption with the alpha5 alternate cd supposed to work? [08:44] ==3295== Possible workaround: remove --trace-children=yes, if in effect [08:45] ==3295== [08:45] valgrind: /usr/bin/X.valgrind-madness: Permission denied [08:45] 3023 2832 [08:45] giving up. [08:45] ^ slangasek [08:45] slangasek: thank god for /var/log/gdm/ :-) [08:45] s/god/kees/ [08:46] Nafallo: ah - should be fixable, I'll dig into that some more, thanks [08:46] :-) [09:18] How does grub figure out on which partition to look for its configuration file? Does it have the partition ID hardwired into the stage 1.5 loader? [09:20] soren, I believe it does [09:23] liw: hey, do you still work on computer-janitor? [09:24] seb128, sure [09:25] liw: you seem to not be subscribed to its bugs on launchpad, could you consider the patch on bug #344704 or at least comment? the application is one of the few poorly translated in jaunty right now due to that [09:25] Launchpad bug 344704 in computer-janitor "[I18N] Wrong translation domain in ui_gtk.py, some translations not shown" [Undecided,In progress] https://launchpad.net/bugs/344704 [09:25] Keybuk: the installer now uses them if they're explicitly set (but otherwise uses UUIDs, as obviously labels aren't suitable for automatic assignment); furthermore, it will not allow you to proceed past the partitioner if you have any duplicate labels [09:25] seb128, I am getting all its bug mail, certainly; I'll apply the patch as soon as I get around to it [09:27] liw: ok, launchpad is still weird to me your name is not on the list on the side of this page for example, anyway thanks for considering that one [09:27] liw: would be nice to get it fixed just after beta so translators have time to test it and make sure it works good and update their translations if required [09:29] cjwatson: why does it do that though? [09:29] when we originally discussed this back on Montréal, we explicitly ruled out ever writing LABEL= by default [09:29] Keybuk: because labels are a lot more readable, and if the user explicitly set some, then it's probably because they wanted to use it [09:30] since labels are most likely to be the filesystem's purpose [09:30] ie. "/" or "root" or "home" or "/home" [09:30] they are much more likely to clash if you put an extra disk in your machine [09:30] or even just a usb key [09:30] and then you end up with random mount behaviour [09:30] we agreed that if a user wanted to use a label in /etc/fstab, they could put it their themselves [09:30] sigh. [09:30] but that we would never put them there automatically [09:30] ok, I'll revert it post-beta. [09:30] thanks [09:31] and try to explain to debian-boot@ what's going on [09:31] because this was explicitly requested there upon trying to get the UUID changes into Debian [09:31] I really just think seeding mount-by-label automatically is dangerous [09:31] if I plug a usb key in, it's generally slightly more likely to win than the sata disk controller [09:31] but not always [09:31] so the random behaviour is just too much of a risk [09:34] I'd only remembered the duplicate label issue [09:34] TheMuso: do we really need this orca menu item? it adds a menu category for something which opens a command line tool ... [09:34] Keybuk: I think I'll add a partman/mount_by_label preseed or similar [09:34] defaulting to false [09:35] that'd be ok [09:35] I don't mind if a user takes an action to use them, then it's their fault [09:35] and when they file a bug, we can go "aha! but you did foo, so don't do that then" [09:35] do we need to unwind this for beta to avoid having to support upgraded systems? [09:35] I'd just release note it [09:35] THE RIGHT ANSWER [09:36] when did the installer start using labels? [09:36] 24 Feb [09:37] yeah, I'd just release note that - say that systems installed may use label in /etc/fstab - and that this could cause unexpected behaviour if another disk with the same label is added later [09:37] with instruction on how to find the uuid of the device [09:37] can you file a bug on partman-target and ubuntu-release-notes? [09:38] I got five hours sleep last night due to this debconf bug, and need to go and get a bit more ... [09:38] sure [09:38] thanks [09:40] pitti: excellent, thank you [09:43] lool: dtchen is working on this, give me a sec and I'll get you a bug numbetr. [09:43] seb128: I can address this post beta, its probably not needed there anyway. It opens a terminal because orca tries to set itself up if orca and accessibility is not set up. [09:44] TheMuso: Ok; concerning the autostarting of pulse is it a bug or it's just me reading the description wrong? [09:45] lool: description of what? [09:45] TheMuso: the pulse.conf machinery [09:45] TheMuso: ok thanks [09:45] # PulseAudio alsa plugin configuration file to set the pulseaudio plugin as [09:45] # default output for applications using alsa when pulseaudio is running. [09:45] TheMuso: But what this plugin really does is to start pulseaudio if it isn't running [09:45] lool: dtchen decided to re-enable auto spawning of pulse if it wasn't running [09:45] lool: no the plugin doesn't do that, see my comment re auto spawn [09:46] TheMuso: Well however you turn it, it's hard to avoid using pulseaudio in this case [09:46] lool: bug 343254 [09:46] Launchpad bug 343254 in linux "pulseaudio: alsa-util.c: snd_pcm_avail_update() returned a value that is exceptionally large" [High,In progress] https://launchpad.net/bugs/343254 [09:46] TheMuso: I wanted to debug pulseaudio per the upstream instructions and couldn't until I disabled this snippet [09:46] lool: I suggest you talk to dtchen about his reasoning. [09:46] TheMuso: Ok [09:47] TheMuso: The bug you point at is fixed in jaunty; it's certainly not fixed for me [09:48] Ah I guess I need new kernels [09:48] lool: hang on it may be the wrong one [09:48] ah ok [09:48] lool: I think dtchen points to test kernels in that bug [09:48] Yes, I found these kernels in multiple ways already, ok thanks [09:59] lool: where are the installation instructions for UNR? [10:01] mdz: http://cdimage.ubuntu.com/ubuntu-netbook-remix/daily-live/current/ > https://wiki.ubuntu.com/MobileTeam/Mobile/HowTo/ImageWriting [10:01] lool: thanks [10:01] mdz: For the installation media creation; or were you looking for instructions on running ubiquity etc.? [10:04] lool: imagewriter looks nice; how come it isn't in universe? [10:05] mdz: I think at some point we wanted to merge usb-creator and imagewriter rather than having two tools [10:05] lool: I was looking for the analogue of https://help.ubuntu.com/community/Installation [10:05] lool: that sounds like a good idea, but if it won't happen for 9.04, I think having both in the archive would be useful [10:05] I'll raise it with ogra then [10:13] <\sh> something is strange regarding keyboard focus with unr cursor keys always focuses on the menu sidebar but not inside the selection... [10:18] cjwatson, if i have a bug with a number of tasks on the wrong packages [10:18] should i mark them invalid and make new ones, or does it make more sense to convert them to the right ones [10:18] apw: i would just change the package assigned, not invalidating. the reason is that even if a task is invalid, the subscribers of that task still get bugmail. [10:18] cool thanks was leaning that way too as they added no information [10:23] Riddell: new update-manager is uploaded === Trewas666 is now known as Trewas [10:28] dpkg-gensymbols: warning: debian/libblkid1/DEBIAN/symbols doesn't match completely debian/libblkid1.symbols [10:28] meh [10:28] what does _that_ mean?! [10:32] apw: yeah, what he said [10:32] pitti: any idea why bug 347452 hasn't been retraced yet? [10:32] Launchpad bug 347452 in network-manager-applet "nm-applet crashed with SIGSEGV in g_main_context_dispatch()" [Undecided,New] https://launchpad.net/bugs/347452 [10:32] reported 16 hours ago [10:33] meh, no iso re-roll over night ? [10:33] cjwatson: the retracers got stuck, they got restarted now [10:33] s/now/an hour ago/ roughly [10:33] ah, thanks [10:33] I am officially worried about bug 328035 [10:33] Launchpad bug 328035 in xserver-xorg-video-intel "*** glibc detected *** free(): invalid next size (fast) for xf86Wakeup() call" [High,Triaged] https://launchpad.net/bugs/328035 [10:33] * ogra wanted to see if the kernels end up on the armel CDs ... indeed i did forget that the automatic builders are off during freeze [10:34] cjwatson: do you know in which package is the "Restart Now" or "Continue Testing" dialog after ubiquity install? [10:34] if any of you have seen an unexplained X server crash, please check the gdm logs to see if it is this bug [10:34] ogra: I can do a manual one [10:34] seb128: that's in ubiquity itself [10:34] cjwatson, i would really appreciate that [10:34] ogra: running [10:34] thanks [10:34] cjwatson: there is lot of strings not displayed in french there but ubiquity on launchpad seems to not have changed for ages and to be translated in french correctly [10:35] weird [10:35] https://translations.edge.launchpad.net/ubuntu/jaunty/+source/ubiquity/+pots/ubiquity/fr/+translate?batch=10&show=all&search=Restart+Now [10:35] -> not listed [10:35] cjwatson, i see you and steve are both diung a full run every time, just armel would be fine ... [10:35] no "Restart Now" in the template apparently [10:35] *doing [10:35] ogra: *shrug* [10:35] ogra: I just copied and pasted it out of the crontab [10:35] * seb128 apt-get source ubiquity [10:36] ah [10:36] seb128: well, ubiquity's translations do have to be updated by hand, but let me check [10:36] I did that on 18 March [10:36] oh, you only do ports, i see, steve did all desktops yesterday [10:36] cjwatson: ubiquity has been translated 100% in french for ages, did all those strings changed recently? [10:36] just checking [10:37] https://translations.edge.launchpad.net/ubuntu/jaunty/+sources/ubiquity/+translate indicates it didn't change since 2008-04-23 [10:37] ah, yes, they did, 2009-03-02 [10:37] seb128: that's the wrong place to look though [10:37] https://translations.launchpad.net/ubuntu/jaunty/+source/debian-installer/+pots/debian-installer [10:38] Current French: (*) Redémarrer maintenant [10:38] Translated by FredBezies on 2009-03-10 [10:38] Reviewed by Bruno Patri on 2009-03-21 [10:38] right [10:38] so just waiting for us to suck the updated translations into the package [10:38] ok thanks [10:38] I'll do that again afterbeta [10:38] with extra spaces [10:39] I was just making sure it's only pending translations and not a bug in ubiquity or an outdated template in rosetta [10:39] seb128: the ubiquity .pot file is just the .desktop files [10:39] seb128: I'd probably better check that the current installer .pot matches the one I generate [10:41] Who would be responsible for gnome-screensaver translations? I reported to the launchpad team a couple weeks ago that over 100 translations were attributed as uploaded by me but I have nothing to do with it [10:45] TheMuso: there is no audio theme selected by default in the sound capplet do you know if that's a known issue with ubuntu-sounds or something? [10:46] seb128: Its the first I've heard of it actually. [10:46] I'll have a look tomorrow and get it sorted post beta. [10:46] thanks [10:51] whee, out of the blue, someone e-mails me and asks if they may help develop computer-janitor [10:51] this must be what it feels like to author successful free software [10:55] liw, i thought that involved making legal threats against anyone making patches for your software... or was that only a specific free software dev... [10:57] directhex, a developer of formerly free software, even :) [10:58] TheMuso: bug #347537 if you need a bug number [10:58] Error: Could not parse data returned by Launchpad: timed out (https://launchpad.net/bugs/347537/+text) [10:58] liw, it's perfectly free if someone at a conference winks at you when telling you the licenses are incompatible! that's like a CONTRACT! [10:58] seb128: thanks [10:59] seb128: there were a few strings not in the template, so I'm uploading new versions now [10:59] cjwatson: thanks [11:03] cjwatson: can you offer any advice regarding bug 346589? [11:03] Error: Could not parse data returned by Launchpad: timed out (https://launchpad.net/bugs/346589/+text) [11:13] mdz: well, I agree with Evan. There's no point in offering a disk for partitioning when partman is going to fall over when it tries to do anything non-trivial with it. [11:14] mdz: or am I missing your point? you don't give context of your specific situation [11:16] mdz: I think it is a bit tenuous for you to say "both of the issues identified here", since the original reporter actually said that he was pleased to see a warning displayed [11:16] mdz: so if you think that part is a problem, it should definitely be a fresh bug [11:31] mvo: https://bugs.edge.launchpad.net/ubuntu/+source/glibc/+bug/339589 there are a lot of upgrade logs like this one. apparently /usr/share/debconf/confmodule is missing when the glibc.preinst is run [11:31] Ubuntu bug 339589 in glibc "package libc6 2.9-4ubuntu2 failed to install/upgrade: subprocess pre-installation script returned error exit status 1" [Undecided,New] [11:36] doko: I don't think that can be the case. In this case the terminal log shows debconf output [11:36] which doesn't point to /usr/share/debconf/confmodule being missing, quite the opposite ... [11:38] looks more like the dialog frontend getting used when the terminal is not really usable (i.e. will fail when you try to read from it) [11:38] cjwatson: yes, seen now as well. the log has all the escape sequences, so it's looks rather like an upgrade with RELEASE_UPGRADE_MODE != desktop, or the debconf priority changed on the system [11:39] right, though ideally debconf ought to notice that dialog isn't usable and try something else ... === azeem_ is now known as azeem === ziroday` is now known as ziroday [11:55] doko: I have no good solution for this, there is no easy way to detect if the upgrade was aborted by hand manually or if its a real error [11:56] doko: the user will end up with a system that has jaunty in the sources.list and no packages update, it will require manual intervention [11:56] ok [12:26] Keybuk: wtf do I have to do to make a package that wants debhelper 7 survive with hardy's debhelper (6) ?? [12:28] lamont: hardy-backports has dh7 [12:28] not an option [12:39] * lamont ponders dh_prep and how to stab _that_ in the face, too [12:46] lamont: no idea [12:46] dh_prep used to be spelled dh_clean -k [12:47] cjwatson: yeah - man pages for the face-stabbing win [12:47] sed -i 's/$debhelper *$( *>. *7[^)]*/\1(>= 6/' ${CTL} [12:47] sed -i 's/dh_prep/dh_clean -k/' ${RULES} [12:47] somedays, I feel _dirty_ [12:56] hello, if there are any Ubuntu devs around, I would like to report that the "New Broadband Connection" wizard thing is not available for translation, it's in main, but I have no idea what package provides this wizard [12:56] thus I'm asking here if anyone knows what should I report this against [12:56] networkmanager? [12:56] I'll be on and off around here, so if anyone knows just contact me [12:56] I suspected libmbca, but I'm not sure [12:59] evand: auto login doesn't seem to make sense with oem install, should it be removed from the UI for oem install? [12:59] Why doesn't it make sense? [12:59] oh [13:00] you mean auto login in ubiquity when in oem mode, not in oem-config, right? [13:00] Riddell: it makes sense for netbooks [13:00] evand: well it logs in automatically as oem but once the user is set up it still tries to log in as oem (which doesn't exist any more) and gives an error [13:00] ah [13:01] evand: right, it could be added to oem-config [13:01] Riddell: can you file a bug for this? [13:01] yep [13:02]