[00:25]  * jdong ponders the laziest way to replace this encrypted Hardy netbook image with Jaunty.....
[00:26] <lifeless> update-manager -d -c
[00:26] <jdong> lifeless: do you think that will go smoothly from Dell's modified LPIA Harddy port up to Jaunty?
[00:27] <lifeless> no idea
[00:27] <lifeless> but it would be fun to find out
[00:27] <jdong> yeah that was my worry :)
[00:27] <jdong> regular Hardy, that's what I would've done in a heartbeat
[00:28] <jdong> ok I'm just not gonna be lazy and download an alternate CD and repack it into a USB.
[01:09] <mneptok> jdong: you'll do as you are ordered, and like it, soldier!
[01:10] <jdong> mneptok: wow... either the first heat wave is taking its toll on me or we need to take this conversation into private quarters :D
[01:12] <mneptok> jdong: let me get my "Ursula, Demon Mistress Of The SS" costume
[01:13] <jdong> mneptok: my therapist said those only exist in nightmares
[01:14] <mneptok> jdong: http://cfs7.tistory.com/image/11/tistory/2008/08/05/15/27/4897f2d819627
[02:46] <pipegeek> I think there may be a bug in the version of kvm that shipped with jaunty (kvm-84).  It's currently very slowly and meticulously hosing a qcow2 image (instead of snapshotting it).  If I find it, I'll submit a bug report.  It looks like there's already a bug report open for a similar problem with the version of qemu-img that shipped with jaunty (352785), and that it was reported nearly a month ago.
[02:46] <pipegeek> err, s/kvm/kvm-img/
[05:03] <doctormo> I'd like to create the simplest kind of debian package, one of a binary blob, which only really needs to put some files in the right places.
[05:04] <StevenK> doctormo: Then your debian/rules contains a bunch of boiler plate and a few calls to cp
[05:05] <doctormo> StevenK: Yes my next line was going to bem shoudl I use rules or just an install file?
[05:05] <StevenK> Or mv, if you're feeling particularly evil
[05:05] <doctormo> because the install file seems to cover this use case too
[05:05] <lifeless> an install file would be better I think
[05:05] <StevenK> doctormo: Sounds like a perfect case for debian/install and dh_install, but it is your choice
[05:08] <jdong> doctormo: IMO install file would be the more "aesthetically pleasing" way of doing it though nothing's wrong with the cp-in-install-target approach either
[05:09] <lifeless> data driven++
[05:10] <doctormo> Also, since so many of you are here,
[05:11] <doctormo> I'm packaging up a binary driver which has a license which allows redistribtion, what's the rules on putting in on my ppa?
[05:11] <jdong> that's more of a #launchpad question
[05:12] <doctormo> jdong: aye aye
[05:12] <doctormo> thanks for your help
[05:25] <doctormo> Ah, now it's interesting, some of the files are symbolic links, using just the install file, the debuild process just creates copies of the files instead of preserving the links
[05:31] <jdong> there's a special install.links file or something like that for symlinks IIRC
[05:35] <doctormo> jdong: do you know of documentation where the format of that file might be available?
[05:38] <ScottK> man dh_link is a likely place to start.
[05:39] <LaserJock> doctormo: it's just links or <binary name>.links
[05:39] <LaserJock> doctormo: but that's to create new links, not preserve old ones
[05:40] <doctormo> LaserJock: It's ok to create new links, it's just linking a lib to it's versioned versions
[05:42] <LaserJock> doctormo: hmm, I'd think that should have worked
[05:42] <doctormo> LaserJock: I still don't know the format of the file debian/links
[05:44] <LaserJock> source destination
[05:44] <LaserJock> doctormo: like ln -s just without the ln -s :-)
[06:09]  * LaserJock just realizes that it's impossible to do a search for "canonical" LP teams :(
[06:10] <LaserJock> that is, searching for the term "canonical" gives a permission denied
[06:50] <pitti> Good morning
[06:56] <pitti> cjwatson, infinity: I uploaded debhelper on Friday, waiting in unapproved now
[07:21] <dholbach> good morning
[08:54] <jussi01> err, who wrote the release notes?
[08:55] <StevenK> Many people
[08:55] <jussi01> perhaps we should suggest using "gksudo gedit" instead of "sudo gedit" ?
[08:56] <liw> why?
[08:57] <jussi01> !gksudo
[08:58] <liw> "sudo gedit" seems to work for me (I had nothing to do with the release notes, though)
[08:58] <jussi01> liw: it works, but isnt the right way to do it. (read the link there)
[09:00] <jussi01> brb
[09:01] <geofft> out of curiosity, does 'sudo -H' solve these issues? (i.e., does gksudo just happen to do what -H does, by default?)
[09:01] <liw> sudo seems to preserve HOME already
[09:03] <liw> I have nothing against people recommending gksudo instead of sudo for GUI apps, but as long as sudo works perfectly fine, except for a couple of broken apps, I'm not sure I care much :)
[09:03] <geofft> right, -H means to set HOME=/root
[09:03] <geofft> $ sudo env | grep HOME     HOME=/home/geofft
[09:04] <geofft> The issues that web page discusses seem to be caused when root creates files in your homedir
[09:07] <liw> few GUI programs should be run as root (firefox certainly not)
[09:41] <geofft> Apparently screen's initscript cleans sessions I started at the emergency password prompt.
[09:41] <geofft> Where on the "bug" to "you're doing something weird, stop it" scale is that?
[09:43] <liw> "obviously a bug"
[09:44] <liw> /tmp cleaning might have similar issues
[09:48] <directhex> it looks like i need an SRU after all.
[09:50] <dholbach> asac: thanks for signing up for a Packaging Training session :)
[09:53] <asac> dholbach: welcome
[10:10] <jussi01> sorry, had to disappear for a while there - work ftw!
[10:12] <jussi01> liw: My issue is we have always had the recommendation for using gksudo as put forward in that factoid. If we are going to have that there, does it not come across silly to do the contrary on the release notes?
[10:19] <liw> jussi01, I don't know. Since plain sudo works for all relevant apps, I'm personally fine with having it in the release notes, but if you would like it fixed, suggest it to the release notes people (or edit the wiki)
[10:21] <dholbach> bryce, tjaalton: does bug 352708 need any more info to make it more useful?
[10:21] <slangasek> jussi01, liw: FWIW, this was discussed and we're using 'sudo' for consistency, so please don't edit the wiki until you've persuaded the release team why this is wrong :)
[10:25] <jussi01> slangasek: exactly the reason I hadnt edited it yet. thanks for clarifying
[10:49]  * slangasek looks in dismay at the set of upgrade bugs at https://bugs.launchpad.net/~milkthemilk
[10:49] <slangasek> I don't think that user's system is coming back
[10:53] <StevenK> The description in #366326 is fairly tasty
[10:54] <pitti> ScottK, slangasek: sru-accept updated to add missing space to bug comment, FYI
[10:54] <slangasek> pitti: ah yes, thanks
[11:05] <pitti> NCommander: so, I set up an apport retracer on rimu for armel, but I get timeouts from LP
[11:06] <pitti> seb128: ^ (FYI)
[11:06] <seb128> pitti: ok
[11:06]  * pitti blames firewall rules
[11:07] <ogra> pitti, doesnt matter, armel has no bugs anyway :P
[11:07] <pitti> ogra: https://bugs.edge.launchpad.net/ubuntu/+bugs?field.tag=need-armel-retrace :)
[11:07] <ogra> its the totally bugfree arch :)
[11:08] <ogra> lies all lies !
[11:08] <directhex> sabdfl, a word of advice? mail from roy schestowitz -> /dev/null. the guy's a compulsive liar.
[11:08] <pitti> ogra: got an Android G1 last week, so now I have arm in the house as well ;)
[11:08] <ogra> pitti, none we support though ...
[11:09] <pitti> ogra: we need to change that!
[11:09] <ogra> pitti, we likely do but in the other direction ...
[11:09]  * amitk hides his android ports tree
[11:09] <pitti> ogra: (just kidding)
[11:09] <ogra> karmic migth only support armv7
[11:09] <ogra> which pretty much makes us lose all community
[11:11] <pitti> :(
[11:11] <ogra> yeah
[11:33] <geofft> where should I report HTTPS on brainstorm.ubuntu.com being broken?
[11:33] <directhex> i'd start a brainstorm vote on it. options include "ignore it" and "fix it". i'd totally vote up the "fix it" option though
[11:34] <geofft> hahaha
[11:34] <geofft>  SSL received a record that exceeded the maximum permissible length. (Error code: ssl_error_rx_record_too_long)
[11:35] <directhex> https://bugs.launchpad.net/ideatorrent/ i think
[11:35] <Hobbsee> there's #ubuntu-brainstorm as well, fwiw
[11:37] <geofft> ah, thanks
[11:46] <ScottK> pitti: Thanks.
[11:52] <ScottK> pitti: I need to do another clamav SRU I think rather urgently for Bug #365823 - I was hoping you'd be willing to accept the one that's currenlty pending and verified first.
[11:52] <Ampelbein> ScottK: i attached a new debdiff to bug 351017
[11:53] <pitti> ScottK: 'accept first' -> move to -updates before accepting another SRU?
[11:53] <ScottK> Ampelbein: I'm not in a position to sponsor it right now, I'd approach the person that sponsored the last upload.
[11:53] <ScottK> pitti: Yes.
[11:53] <pitti> ScottK: or is your new upload fixing something that got broken in the previous SRU?
[11:53] <Ampelbein> ok
[11:54] <ScottK> pitti: No, I confirmed the problem exists in -release too.
[11:54] <pitti> ScottK: no problem from my side
[11:54] <ScottK> Great.  I'll have the new on uploaded shortly.
[11:54] <pitti> ScottK: the intrepid SRU isn't verified yet, though, so I only copy the jaunty one
[11:55] <ScottK> Correct.
[11:55] <ScottK> This new bug is only for Jaunty anyway.
[11:55] <pitti> copied to j-u and karmic
[11:57] <ScottK> Thanks.
[12:08] <ScottK> pitti: clamav uploaded to jaunty-proposed again.  The test case is also being written.
[12:40] <Keybuk> pitti: could you push your udev changes somewhere?
[12:40] <Keybuk> lp:~ubuntu-core-dev/udev/ubuntu-jaunty maybe?
[13:17] <pitti> Keybuk: oh, sure
[13:18] <seb128> evand: hi, do you need any extra info on this usb key usb-creator bug?
[13:18] <seb128> evand: I will probably format the key or something so I can use it
[13:21] <evand> seb128: no, I have enough.  Please do go ahead and format the key.
[13:21] <seb128> evand: ok thanks
[13:26] <pitti> Keybuk: done, lp:~ubuntu-core-dev/udev/jaunty (btw, you might want to consider to upgrade --1.6 for go-faster stripes)
[13:28] <Keybuk> pitti: but I only have bzr 1.13 ?
[13:29] <pitti> Keybuk: it's just the format version (there's an --1.9 too :) )
[13:29] <Keybuk> pitti: what I mean is that's the version I init'd with
[13:29] <Keybuk> so surely that's the latest format?
[13:29] <pitti> for some reason unknown to me it's not the default
[13:29] <pitti> perhaps to keep it compatible with some ancient dapper bzr or so
[13:29] <Keybuk> *blink*
[13:29] <Keybuk> so what format are we supposed to be using?
[13:30] <pitti> I don't think we are "supposed" to any
[13:30] <Keybuk> and how are we supposed to make an informed decision?
[13:30] <pitti> I just converted mine to 1.6, so did a lot of other people
[13:30] <slangasek> 'bzr help formats'
[13:30] <pitti> to make use of stacked branches and much faster pushes/commits
[13:31] <Keybuk> slangasek: that says "use the default format"
[13:31] <pitti> Keybuk: it's not an error to use the default one, I just wanted to mention that a newer one might particularly benefit udev, with its throusands of commits
[13:31] <Keybuk> pitti: the bzr repository is converted from git
[13:31] <Keybuk> so I don't really want to muck around, since it breaks pretty much monthly anyway
[13:31] <Keybuk> and I have to push --overwrite it
[13:32] <pitti> ugh
[13:32] <pitti> Keybuk: I recently tried to get bzr-git working, but it failed spectacularly; did you manage to get this to work?
[13:32]  * Keybuk just works on udev in git
[13:32] <Keybuk> pitti: nope, I'm just fast export | fast import
[13:32] <pitti> ah
[13:34] <Keybuk> I started off trying to work in bzr, with a branch based off git
[13:34] <Keybuk> but I repeatedly suffered from that age-old problem where I needed to know what git was doing, and what bzr-git was doing, in order to use bzr
[13:34] <Keybuk> so it was more effort
[13:34] <Keybuk> so I reverted to just using git, with the packaging in bzr
[13:34] <pitti> I just use git for importing postgresql as well
[13:34] <Keybuk> for util-linux-ng, the packaging is in git too
[13:35] <mvo> I'm considering doing the same for compiz
[14:01] <pitti> Keybuk: any idea why https://merges.ubuntu.com/main.html still considers most merges as "updated"?
[14:02] <Keybuk> pitti: MoM is still running
[14:02] <Keybuk> it hasn't got as far as updating the web pages yet
[14:02] <pitti> ah, so it'll fix itself then?
[14:02] <pitti> thanks
[14:02] <Keybuk> did I not understand that we weren't to use MoM for the merge this time anyway?
[14:02] <pitti> oh, we aren't?
[14:03] <Keybuk> cjwatson: ?
[14:03] <Keybuk> Mark was saying we were to use bzr and james_w's imports?
[14:03] <pitti> OIC
[14:03] <pitti> did I miss an announcement for that?
[14:05] <ogra> Keybuk, damned and i was hoping we actually did such a good job with upstreaming stuff that we only had 52 updated packages :P
[14:05] <cjwatson> james_w's imports aren't quite going to make it for the start of the merge, unfortunately, but will hopefully be available by part-way through (before UDS)
[14:05] <pitti> ogra: remember that in jaunty, debian was still in freeze
[14:05] <ogra> yeah
[14:05] <ogra> scary ...
[14:05] <pitti> most of my patches forwarded to Debian weren't applied then
[14:06] <cjwatson> they need a bit of rejiggering to have the right ancestry relative to Debian, and to upgrade to a new branch format per the Bazaar team
[14:06] <pitti> and many still aren't :( (but many were)
[14:15] <ScottK> I seem to have fallen out of favor with ubuntu-devel-discuss (just had a post moderated and it looks to me like I sent from the correct address).  I'd appreciate it if some list admin would have a look ....
[14:17] <slangasek> pitti: could you comment on the correctness of the patch in bug #366119?  (cf. my comments about invoking hal vs. pm-suspend)
[14:18] <pitti> slangasek: gosh, I thought we entirely got rid of acpi-support suspend quirks in jaunty
[14:19] <slangasek> well, we've gotten rid of the quirks
[14:19] <slangasek> but not eliminated all the cases where acpi-support will bypass pm-utils
[14:19]  * pitti reads the completl bug trail first
[14:19]  * slangasek shakes his fist at gdm
[14:23] <pitti> slangasek: replied and subscribed
[14:29] <pitti> cody-somerville: do you still need cdbs' xfce.mk?
[14:29] <pitti> cody-somerville: last merge log said it's being phased out
[14:29] <cody-somerville> pitti, We do not use it anymore however I'm wondering about how the archive reorg will affect that
[14:30] <pitti> cody-somerville: what does the reorg have to do with it?
[14:30] <cody-somerville> translations
[14:30] <cody-somerville> Currently only main get the magic translation love
[14:30] <mr_pouit> pitti: I filed a bug report long ago asking for its removal
[14:30] <pitti> mr_pouit: oh
[14:31] <pitti> ah, bug 254048
[14:31] <cody-somerville> mr_pouit, Do you think its a possibility we might use it again if we get the translation magic stuff back with the archive reorg?
[14:31] <pitti> but then you could just as well include langpack.mk
[14:32] <mr_pouit> cody-somerville: no, I would prefer to stick with upstream translations
[14:32] <cody-somerville> mr_pouit, okay.
[14:36] <superm1> cody-somerville, with archive reorg, universe and multiverse are getting translated too then?
[14:36] <cody-somerville> I dunno
[14:36] <cody-somerville> Thats what I was wondering
[14:36] <superm1> won't that make the lang packs too big for DVDs?
[14:38] <superm1> well should the first phases of archive reorg be happening during the karmic cycle then, or still in planning mode and (karmic +x)?
[14:54] <cjwatson> superm1: we'll only be doing the permissions reorganisation in the karmic cycle. Once we do the more sweeping component reorganisation (still in planning mode and so only "later"), we'll be able to have finer-grained control over language pack generation, so we'll be able to say "Ubuntu desktop", "Kubuntu", etc. rather than "main".
[14:54] <cjwatson> superm1: I don't think we'll be putting universe and multiverse in language packs, no
[14:54] <superm1> cjwatson, ah i see.  that makes much more sense
[15:01] <slangasek> Riddell: follow-ups on bug #339313 point out that the release notes only mention problems with WPA2, whereas users are reporting that plasma-widget-network-manager also has problems with WEP; can you confirm, and do you know if anything more should be added to the RN besides s/WPA2/& and WEP/?
[15:06] <Riddell> slangasek: I've not heard of it being broken with WEP, but that doesn't mean to say it never is
[15:07] <Riddell> slangasek: I'll look into it
[15:07] <liw> evand, are you planning to add a "remaining time estimate" to usb-creator?
[15:08] <evand> liw: I am now :)
[15:08] <liw> evand, I'm wanting to write a little Python library to do the actual estimation, would you like to include that? :)
[15:08] <evand> surely
[15:09] <slangasek> Riddell: ok, thanks
[15:09] <liw> cool, I'll hit you with a patch some day soon
[15:15] <liw> evand, hmm, also, usb-creator seems to be really, really slow on this laptop/with this usb stick
[15:23] <evand> liw: significantly slower than using dd directly?
[15:24] <liw> evand, it's been running for half an hour now...
[15:24] <liw> or something like that
[15:24]  * Keybuk actually LOLs at the "Paging Jeff Waugh" on Planet GNOME
[15:25] <evand> liw: What size persistence file did you create?
[15:25] <liw> evand, um, the default -- 128 mb?
[15:25] <evand> very odd
[15:26] <evand> Did it crash?  (~/.usb-creator.log)
[15:26] <liw> evand, correction, it's been running for one hour
[15:26] <liw> evand, hasn't crashed, and occasionally it progresses a bit
[15:27] <liw> perhaps it's a crappy usb stick, I'll try with another (on another computer, though, so lots of things will change)
[15:28] <liw> how big a stick do I need, anyway?
[15:28] <evand> anything larger than 700MB should work
[15:38] <liw> evand, hmm, on the other machine, with another stick, usb-creator already finished; I've had trouble with Transcend sticks before, so I'll stick with the theory that that's the problem
[15:38] <evand> ok
[15:38] <evand> thanks for the heads up anyway
[15:39] <ogra> liw, on a sidenote, usb 1.1 vs 2.0 issue ?
[15:41] <mdz> ara:  http://qa.ubuntu.com/ says "ubuntu brainstorm" :-)
[15:41] <ion_> Cool, the legacy lcdfilter started working again in jaunty after updating. I’m unable to find any mention of it in the changelogs. When it broke (a couple of weeks ago, i think), i didn’t find anything in the changelogs either.
[15:41] <liw> ogra, remind me how I verify?
[15:41] <ara> mdz: yes. it said that a long long time ago. schwuk is working in a new landing page
[15:42] <ogra> liw, hmm, good question :) dmesg might be able to tell something ... i used to check with lsmod in the past but nowadays i think we compile in the usb host drivers
[15:43] <dholbach> https://wiki.ubuntu.com/UbuntuOpenWeek starting in 18 minutes on #ubuntu-classroom
[15:43] <ogra> dholbach, will there be canapes, tea and coffee ?
[15:43] <liw> ogra, alas, I see nothing obvious in the output of lsusb or dmesg
[15:44] <dholbach> ogra: sure :-)
[15:44] <ogra> liw, i have one of these cheap intel give away usb hubs here that switch the connection to 1.1 *behind* the port with no indication at all ... writing a 2G stick with that thing easily takes 2h for me
[15:46] <liw> evand, usb-creator still wants a desktop ISO, right?
[15:46] <mdz> liw: I believe it accepts server, alternate and DVD as well
[15:47] <evand> indeed, what mdz said
[15:47] <liw> excellent
[15:50] <pitti> slangasek, cjwatson: cdbs merged, tested, uploaded; *ugh*
[15:56] <mnemo> pitti: I've been asking a lot of people to run apport-collect lately and it seems that this script itself actually crashes for a significant number of users --> https://bugs.launchpad.net/ubuntu/+source/apport/+bug/368004
[15:58] <mnemo> pitti: within the hour I had two such crashes and this makes triage of all the xorg bugs pretty difficult
[16:15] <Keybuk> 418 outstanding merges
[16:15] <Keybuk> 2 updated merges
[16:15] <Keybuk> eep
[16:18] <pitti> yay lenny release
[16:22] <seb128> karmic is going to be "fun"
[16:23]  * ogra wonders if seb128 has that as IRC macro anywhere ... you seem to say that sentence every time the archive opens :P
[16:23] <highvoltage> heh, how so?
[16:23] <seb128> ogra: every cycle you mean?
[16:23] <ogra> yeah :)
[16:24] <ogra> its a traditional thing, isnt it ? :)
[16:24] <seb128> I probably didn't for jaunty
[16:24] <seb128> that was quiet cycle
[16:24] <directhex> not for me
[16:24] <seb128> but karmic seems to shape to be crazy
[16:24] <ogra> yeah, compared to whats ahead :)
[16:24] <directhex> jaunty was ke-raaazy for some of us
[16:24] <seb128> directhex: I didn't say it was not busy
[16:25] <seb128> directhex: but we didn't land thousand of new packages or any major refactoring in upstream code
[16:25]  * ogra notices that jaunty was one of his busiest ... thanks to ARM
[16:25] <directhex> lenny freeze didn't help
[16:25] <seb128> I found it helpful to have a bug fix cycle
[16:25] <seb128> we got bug fixes updates and not thousand of new versions
[16:26] <directhex> ogra, and the intersection of our work? http://www2.apebox.org/wordpress/wp-content/gallery/00-single/armless.png ;)
[16:27] <ogra> heh
[16:28] <directhex> seb128, on the subject of bug fixes, how do you feel about bug 367810?
[16:28] <directhex> wait, trust me to pick the one archive admin who isn't in the SRU team -_-
[16:29] <ScottK> There's more than one.
[16:30] <directhex> okay, one of the ten, then.
[16:30] <pitti> cjwatson, slangasek: I took the liberty to accept cdbs myself, since we previously talked about it
[16:41] <cjwatson> pitti: np
[16:43] <cjwatson> anyone know of a sync request they need to do anyway for which bug 246307 would fire?
[16:44] <cjwatson> i.e. the Debian changelog delta contains non-ASCII characters
[16:44] <cjwatson> if you do, I'd appreciate if you could try http://paste.ubuntu.com/159460/
[16:52] <pitti> mnemo: thanks, I'll take a look at it and followup in the bug
[17:40]  * hyperair grumbles about jaunty not using 1GB of RAM, and swapping too damn much
[18:14] <Adri2000> mvo_: can you take a look at bug #338419 please? why does it want to install filezilla-locales (which doesn't exist anymore by the way) instead of filezilla?
[18:18] <mvo_> Adri2000: thanks, will do (but tomorrow morning)
[18:59] <chieh> Hello, If i wrote an app I would like to add into the ubuntu repository, how do I go about doing it?
[19:00] <johnm> chieh: you can easily get your own launchpad ppa to begin with :)
[19:00] <chieh> i'm not sure what you mean?
[19:01] <chieh> oh i found a site on it
[19:01] <chieh> but is there a way to incoporate it with ubuntu apt-get?
[19:01] <maxb> chieh: Hop over to #ubuntu-motu for new package contributions
[19:03] <chieh> ok, thanks
[19:19] <shaya> unsure where to file a bug for cpu governor choice.  i.e. I think if a laptop (or a desktop) is plugged in, there's no reason to use the ondemand governor, and one ends up with a degraded experience by using it.
[19:19] <shaya> any ideas?
[19:29] <Commander1024> heyho, found a bug in ubuntu 8.04 LTS stable 32bit (at least)
[19:30] <Commander1024> where shall I report that?
[19:33] <Commander1024> package: openvpn
[19:33] <Commander1024> file /etc/init.d/openvpn line 58
[19:33] <ion_> http://launchpad.net/
[19:35] <Commander1024> hmm, registration requred :-( well I'll file it there -.-
[19:38] <ion_> Yeah, Launchpad isn’t an OpenID consumer yet. :-(
[19:39] <Commander1024> just mentioned, that I already got an account because of earlier incidents so I should have remembered launchpad -.-
[19:39] <Commander1024> just found out ...
[20:02] <Commander1024> filed as Bug #368141
[20:32] <olmari> Hello
[20:33] <olmari> There is small problem when installing new ubuntu 9.04 trough "netboot" or mini.iso
[20:33] <olmari> culprit is AFAIK package libmbca0
[20:35] <cjwatson> olmari: it would help if you said *what* the problem was - but really, bugs like this are better reported in Launchpad
[20:35] <olmari> I'll try to be short: if I choose only to install most basic CLI system and then spt-ge install ubuntu-desktop, it downloads everyfile, apparently libmbca0 last, then there reads [99% working] and it sits there forever
[20:35] <cjwatson> that's not libmbca0's fault
[20:35] <cjwatson> that's a rarely-reproducible bug in apt
[20:36] <olmari> hmm
[20:36] <cjwatson> every so often, the http method gets stuck right at the end
[20:36] <cjwatson> if you try it again, it'll probably work
[20:36] <olmari> nope
[20:36] <olmari> tried 3 times allways same file / spot
[20:36] <cjwatson> unfortunately we've never quite managed to track it down, because (at least on the systems we've tried) it doesn't happen every time
[20:36] <cjwatson> mvo would probably like to work with you on it, but he's not around right now
[20:36] <olmari> ok
[20:37] <cjwatson> please file a bug on https://bugs.launchpad.net/ubuntu/+source/apt/+filebug in the meantime
[20:37] <cjwatson> mention that I sent you
[20:37] <PatrickBic> hi
[20:38] <PatrickBic> my old code (which worked in ubuntu8.10) still compiles but crashes on execution => http://pastebin.be/18092 ... is this usual/wanted/subject-to-change?
[20:38] <PatrickBic> concerning libnotifymm
[20:38] <olmari> if it does matter anything, I managed to install mine ubuntu by choosing ubuntu-desktop in installer, then when it "jams" I press esc and then choose form main menu "select packages to install" or to that effect, ubuntu-desktop was selecer already, then chose next and it started to install straight on :)
[20:39] <olmari> but I'll ofcourse do the bugreport too
[20:39] <cjwatson> olmari: right, that would work around it because it doesn't need to use the http method at all the second time round
[20:40] <olmari> cjwatson: well it apparetnly didn't download anything the second time as everything was on the comp already, including that lib
[20:40] <olmari> cjwatson: if that also matter
[20:40] <cjwatson> olmari: that's a restatement of what I just said, yes :-)
[20:40] <cjwatson> olmari: it doesn't matter
[20:40] <olmari> cjwatson: well I'm not native english speaker so you have to just bear me ;)
[20:41] <cjwatson> the problem you're encountering is specific to downloading stuff, so if apt-get doesn't have to download anything because it's all cached locally then that won't hit the problem
[20:42] <olmari> cjwatson: so the problem is in the apt download, it finishes downloading but not the process, I assume :)
[20:42] <olmari> now in retrospect
[20:42] <cjwatson> right
[20:43] <cjwatson> we encounter it from time to time in our live filesystem builds, but unfortunately only occasionally (once every week or two)
[20:43] <cjwatson> the symptoms suggest a race condition, so it could well depend on e.g. relative speed of your CPU and network connection, or things like that
[20:44] <Commander1024> cjwatson: considering your fist problem (http download with apt) do you have an nforce board with integrated "hardware" firewall?
[20:44] <olmari> hmm... I hope I could reproduce the bug when/if someone want's to try squash it... as I said it happened exact same spot each 3 times until I figured a way to finish installation (without do old ubuntu)
[20:45] <cjwatson> Commander1024: firstly, it's not my problem; secondly, it's vanishingly unlikely to be specific to certain motherboards. We've seen it on a variety of systems. It's almost certainly an apt bug
[20:45] <olmari> Commander1024: The problem is with ME this time, I have ibm t42p laptop, it is all intel stuff
[20:46] <cjwatson> it's very rare for problems above the level of the kernel and X (i.e. stuff that talks to the hardware directly) to be hardware-specific, and it should not normally be the first assumption
[20:46] <Commander1024> hmm, was just an idea, I am maintaining over 40 servers and workstations with debian / ubuntu and I have never seen such behaviour
[20:47] <cjwatson> it displays all the symptoms of a race condition. Race conditions are like this. Some people will see them all the time, some people will see them occasionally, some people will never see them.
[20:47] <cjwatson> it makes them very frustrating to debug
[20:47] <cjwatson> they also tend to go away when you try to apply debugging tools to them
[20:48] <Commander1024> that's right, it's just that this "nforce firewall thingy" created crippeled and wasted downloads of any kind
[20:48] <olmari> I can imagine... again I haven't even seen this problem before but for this one it was a constant
[20:48] <cjwatson> that's not the problem then.
[20:48] <Commander1024> k, got that
[20:48] <cjwatson> if the download were just corrupt, then apt would fail with an error message; it wouldn't just sit there forever.
[20:48] <cjwatson> (and if it did sit there forever, then that would *still* be an apt bug ...)
[20:56]  * cjwatson fixes rmadison to know about karmic
[21:10] <Adri2000> I wonder why we do not have a file somewhere with the list of known releases so that all the programs who need it can just use that file instead of managing their own list
[21:11] <olmari> cjwatson: https://bugs.launchpad.net/ubuntu/+source/apt/+bug/368178 if interested :)
[21:17] <doctormo> Is there a way to re-set home after of while someone is logging in?
[21:18] <doctormo> I'm trying to create a useful way to use usb-pam while allowing the user's home directory (thus files and settings) to be safely stored on the usb stick too.
[22:32] <jablko> where is cron job output sent by default in ubuntu?
[22:35] <ion_> To the appropriate user via /usr/sbin/sendmail probably. Now please read the topic.
[23:51] <EagleScreen> what is the difference between ubuntu-laptop-mode y laptop-mode-tools packages?