/srv/irclogs.ubuntu.com/2010/05/29/#ubuntu-devel.txt

=== pgraner-afk is now known as pgraner
=== pgraner is now known as pgraner-afk
pwnguinwas libdns53 updated recently?04:25
ScottKProbably.  Launchpad knows for sure.04:27
pwnguinhmm04:28
pwnguinactually, maybe its me and not libdns/lp04:29
pwnguinbut is (.= 5.3.6) a valid versioned dependency?04:29
pwnguindpkg doesn't seem to think so04:30
ScottKCurrent version is 1:9.6.1.dfsg.P1-3ubuntu0.304:30
pwnguini had some system instability today, but it seems odd that i would lose one specific character in a few fields04:30
pwnguinthe number was made up, sorry04:31
pwnguinhere's a concrete example from /var/lib/dpkg/available04:32
pwnguinDepends: libc6 (.= 2.3.4-1), python-2.304:32
ScottKWhat package?04:32
pwnguinpython2.3-egenix-mxtools04:33
pwnguinis there a way to just clear out /var/lib/dpkg and let apt rebuild it? cuz i dont really trust it right now04:34
ScottKWell first, that is an incredibly ancient package.04:35
pwnguini know04:35
ScottKIf you don't trust /var/lib/dpkg, why do you trust the rest of the system?04:35
pwnguinwell, it's running, so there's that04:35
ScottKThat may be just because stuff hasn't been re-read off of disk.04:36
pwnguini just rebooted like twenty minutes ago04:36
pwnguinand looked at the file in question with emacs04:36
pwnguinthis system has been dist-upgraded from 4.10 for the past 5 years04:37
pwnguinso im willing to chalk the ancientness of this particular package to a cache that was never cleaned04:37
ScottKWhy don't you pick a non-ancient package and we'll compare results?04:38
pwnguinbecause that's the one dpkg is failing on04:38
pwnguinok04:39
pwnguinfound the command04:39
pwnguindpkg --clear-avail04:39
pwnguinheh04:40
pwnguinapparently my solution of moving the file was bad, but --clear-avail just makes a new empty file and thats okay04:40
pwnguinanyways, mystery 1 of ten solved04:41
pwnguinok, now ext3 just went to readonly mode. it may be time to retire this machine =(04:43
lokpesthas the idea of not having separate distros for each desktop enviroment but instead let people choose DE at install time been discused?04:54
cody-somervilleYes.04:54
lokpestgeneral conclusion? any links?04:55
cody-somervilleWe want the install to be as simple as possible.04:56
cody-somervilleNo need to bother someone with that question who might not understand it.04:56
lokpestok, but then it should be an advanced option04:56
stentenAlso, how would you get all that onto a LiveCD?04:56
cody-somerville1. Its possible to install different desktop environments very easily for folks who want to. 2. If someone wants a specific desktop environment, they just download the appropriate CD.04:57
cody-somervilleand stenten also makes a very very very good point.04:57
lokpestits not like everybody has to partion there drives, thats the advanced option04:57
cody-somervilleWe agree with you.04:57
lokpestok, but that was an argument against "being as simple as possible"04:58
cody-somervillehow so?04:58
cody-somervilleOh, I see what you mean.04:59
cody-somervilleregardless, we still can't fit multiple desktop environments on one CD04:59
lokpestnot everyone has to partition, but you can if you want to. not everyone should be forced to pick a DE (default for gnome)04:59
stentenIf you wanted that option, you'd have to have a minimal install, and then reboot into a barebones machine and install the DE.05:00
stentenWhy would we do that to users?05:00
stentenAnd it would also mean the LiveCD wouldn't be an option, because you wouldn't have the space for an Ubuntu live environment along with all the Qt packages of Kubuntu.05:01
lokpestthe CD is a dying medium05:01
cody-somervilleoh please, lets not get into that argument05:01
cody-somervillethere are tons of people around the world who still use CDs and who have very poor internet connections05:01
lokpestok, but there could be an alternate medium install?05:02
cody-somervillethere is05:03
cody-somervilleits called xubuntu, kubuntu, edubuntu, etc.05:03
lokpestmeh05:03
lokpestCanonical favors gnome05:04
lokpestthe reason for my question is, if I want to use a distro based on ubuntu that does ony come with gnome preinstalled, how do I remove all gnome and gnome apps? sure I can install kubuntu-deskop, but removing ubuntu-desktop is much more tiresome...05:06
lokpestas removing the meta-packages dont remove all the apps05:07
cody-somervilleyou download the Kubuntu CD05:08
lokpest....05:08
lokpestI dont want kubuntu05:08
cody-somervilleIf you remove gnome and install kubuntu-desktop, you pretty much what you'd get if you installed the Kubuntu CD05:09
lokpestI want the kubuntu-desktop of the ubuntu-derivative05:09
cody-somervilleDo you not understand that Ubuntu, Kubuntu, and Xubuntu are all built out of the same archive and share the same base set of packages?05:09
lokpestyes I do05:10
lokpestbut removing ubuntu-deskop doesnt remove all non kubuntu apps05:10
cody-somervilleit does nothing, its a meta package05:10
lokpestthere will still be software that are not kde05:10
cody-somervillewow, thats even a more corner case05:11
stentenlokpest: http://www.psychocats.net/ubuntu/purekde05:11
cody-somervillestenten, he doesn't want pure kde05:12
cody-somervillestenten, he wants the tainted, lol05:12
* stenten facepalms05:12
lokpeststenten: ty!05:13
stentenrofl05:13
cody-somervillelokpest, That'll give you what you'd get if you installed Kubuntu....05:13
lokpestyes05:13
lokpestand i dont want kubuntu05:13
lokpestI want trisquel05:14
lokpestbut trisquel only comes with gnome05:14
stentenyou can run gtk apps on kubuntu...05:15
lokpestbut I know I could install kubuntu-desktop, but that didnt solve the problem that stentens link solves05:16
* lokpest dislikes where ubuntu is going freedome-wise05:16
stentenCouldn't you just install Kubuntu and then install trisquel?05:17
stenteni'm so confused :(05:17
lokpesthuh?05:17
cody-somervilleAnyhow, this is starting to get off topic for this channel.05:17
lokpesttrisquel is a distro based on ubuntu05:17
lokpestright, mentioning that you dont like Canonicals policies regarding proprietary software is not allowed05:21
* lokpest *notes down*05:21
ScottKstenten: That link doesn't do what it says it does.  The standard Kubuntu install is not pure KDE.05:28
lokpestsadly no05:30
lokpestit installs OO.org and shit05:31
lokpestwhy anyone would want a word processor beats me05:32
stentenMy bad. Thanks for letting me know.05:32
lokpestI dont know why I let other people pick aplications for me05:35
asaclokpest surely can use the alternate installer and start from zero ;)06:52
ionasac: He’d still have Linux, the GNU coreutils, an init daemon etc. chosen for him! Even that’s obviously too little control.06:57
asaclol06:58
wgrantAnybody up for a derivative that doesn't even include coreutils?06:59
wgrantI think it's a useful market to expand into.06:59
asaci can release the "empty" derivate ;)07:01
ionBut... but... by installing it, you’re still picking the empty for me!07:02
aburchion: Debian at least has a FreeBSD kernel as well. :)07:03
asacempty is nothing so i am not picking anything ;)07:03
wgrantIndeed, we'd still be choosing a set of applications :(07:03
wgrantWe cannot win.07:03
ionasac: [] is just as valid a set of stuff as [linux, coreutils, ...] ;-)07:04
asac[] != NULL07:04
asaci go for the NULL07:04
asacbut yeah. i think there is always the DIY derivate07:05
aburchDoes libmediawiki-perl have any rdeps in Ubuntu?07:09
wgrantAt least not in lucid.07:09
mneptokno coreutils? maybe then RMS will let everyone call it just "Linux"07:11
ionHah07:11
wgrantHaha.07:11
mneptok"why anyone would want libc is beyond me. i just want a word processor."07:15
=== cam is now known as Guest71321
=== Guest71321 is now known as bitshifternz
siretartfta: can you please file a bug about that? the issue is more complicated as it looks on the first sight, and I think that should be documented in a bug10:09
ftasiretart, bug 58720311:37
ubottuLaunchpad bug 587203 in mplayer (Ubuntu) "mplayer relocation error with the new ffmpeg" [Undecided,New] https://launchpad.net/bugs/58720311:37
siretartfta: I see. hm. I'm pondering if I should add Breaks to libavcodec on mplayer11:39
siretartwhay do you think?11:40
ftaunsure. A rebuild of mplayer + a deps libavformat52 (>= 4:0.6) should be enough,11:43
ftai see you have a new mplayer to merge11:43
siretartok11:43
siretartalso note my comment on the bug11:44
ionIf a new version of a library breaks existing stuff that uses it, shouldn’t there be a soname bump?11:45
siretartion: a) the symbol in question was an unpublished, private symbol which mplayer should not have referenced in the first place, and b) the soname was bumped for libavutil49 -> libavutil5011:47
ftayep, it should. libavutil just got one (49->50)11:47
ftaoh11:47
siretartin order to adress the transitive dependency problem, I've introduced symbol versioning, both upstream and in the package11:48
loolmdeslaur: Sorry, missed your note on virt-manager; just uploaded my own merge12:01
loolmdeslaur: Please replace with yours if I missed anything12:01
ftasiretart, can't you just backport the ftbfs fix from rc4 to rc3?12:15
siretartfta: a) - there is not even a rc4 branch yet and b) there has been over a year of active development in libswscale, and we are talking about implementation internals that are used in the more obscure parts of mplayer. I'm currently discussing them upstream for exactly this reason.12:49
ftasiretart, do you mean it's also broken in debian/unstable?12:50
siretartfta: debian/squeeze will ship ffmpeg 0.5, so I won't upload ffmpeg 0.6 to unstable. period.12:51
ftaoh, I didn't know that12:51
siretartfta: if you would install ffmpeg from experimental, then yes, mplayer will break12:52
=== jtechidna is now known as JontheEchidna
siretarthowever maverick has just opened, so its a perfect time to see how many and what packages break with ffmpeg 0.6 :-)12:53
siretartmplayer was pretty obvious to me12:53
siretartfta: btw, I really lot of folks have demanded to include ffmpeg 0.6 for lucid. in fact, the package was even ready in time, but now you know why I've rejected that :-)12:57
Chipzzion: soname versioning? the mplayer project? surely you are kidding? :)13:01
ftasiretart, i'm a heavy user of mplayer since ~2001, i've built it from cvs/svn for many years, i know how painful it is13:02
Chipzziirc the best known project for NOT using sonames properly is mplayer and assorted side-projects13:02
siretartplease stop spreading FUD. a) mplayer doesn't provide any shared libs. at all. and b) I've joined ffmpeg upstream exactly for this reason13:03
ftasiretart, yesterday, i was caught in that bug so i just downgraded ffmpeg to finish what i was doing. now i'm willing to give a try at whatever solution you may have13:04
siretartfta: as indicated, I'd drop the build dependencies on libavcodec-dev, libavformat-dev, etc. and build statically against the included copy as interim hack13:05
siretartin parallel, work upstream on getting rc4 in shape13:05
Chipzzsiretart: good to know you're working with upstream. not wanting to sound sarcastic, but any success in what you're trying to do (which I assume is trying to convince them of the merrits of a more stable API)?13:11
=== aronxu is now known as happyaron
siretartChipzz: depends on what you mean with stable API. if you mean 'no changes' - that would also mean no development and more features.13:17
siretartChipzz: if you mean 'no backwards incompatible changes', that's fairly a non-issue since years13:17
siretartChipzz: the fact that many projects use unpublished headers internals that are subject to change cannot be solved from within ffmpeg. those projects needs to work with upstream to get that functionality public or stop using it13:18
=== pgraner-afk is now known as pgraner
blue_annais the system strftime the one in gnulib package? just a bit confused because I was expecting a something -dev15:14
StevenKblue_anna: No, you want libc6-dev for strftime15:17
blue_annaStevenK: thank you15:17
blue_annaStevenK: I have the libc6-dev installed instead of the libc6-dev-ppc64, and libc6 instead of libc6-ppc -- would that cause problems? :) I was just about to try to debug why strftime's %P and %p weren't working15:20
blue_annaI'm on a powerpc architecture :)15:20
blue_annayou know, I installed gnulib and %p started working15:24
blue_annalol what a crazy system15:24
* wise_crypt is away: need a rest, tired looking a fast channel :)15:24
blue_annaI just noticed that my system does not have installed the libc6-ppc64 library but does have the libc6 standard. I've been noticing plenty of odd little behaviors and a general slowdown of applications and I'm starting to think it is that the powerpc architechture was built against the wrong set of libs15:36
blue_annaany idea how I can scan for libraries that have a -ppc64 alternate that is not installed, when the base library is?15:36
Q-FUNKhttps://bugs.launchpad.net/ubuntu/+source/eglibc/+bug/58718616:00
ubottuUbuntu bug 587186 in eglibc (Ubuntu) "libc6 upgrade fails: illegal instruction" [Undecided,New]16:00
Q-FUNKis this for real?16:00
blue_annaQ-FUNK: what are you doing?16:04
Q-FUNKblue_anna: come again?16:04
blue_annaQ-FUNK: the page you linked wasn't loading :) I see it now16:05
blue_annaQ-FUNK: you've got the wrong architecture package for libc for some reason16:06
blue_annaQ-FUNK: i386 instead of i586 for your system, looks like16:06
Q-FUNKblue_anna: no, I don't.  see doko's answer to the bug.16:06
blue_annaQ-FUNK: the only responces I see in there are from Martin16:07
Q-FUNKblue_anna: we never had i586 packages.  i386 has always implied i386 and newer x86 arch.16:07
blue_annaok, but it says in the log Uname: Linux 2.6.32-10-generic i58616:07
blue_annaQ-FUNK: what's uname -r say? just to verify -- I mean, if that still works :S16:07
Q-FUNKthat's the cpu architecture on the hardware as reported by uname.16:08
blue_annaQ-FUNK: can you link me the specific responce you meant? I don't see any responces from any "doko"16:08
azeem_blue_anna: Matthias16:08
azeem_Q-FUNK: so your question is whether 10.10 will only support i686 or not?16:09
blue_annaI need a link -- there's no Matthius on the page either, taht I see16:09
azeem_"Matthias Klose wrote 2 hours ago:"16:09
azeem_blue_anna: maybe you opened the wrong link16:09
Q-FUNKazeem_: if that's the case, it should be advertised in a more blatant way.16:09
blue_annaAh no, its that Matthias and Martin look very much alike :)16:09
azeem_Q-FUNK: it will be in the release notes I guess16:10
blue_annaQ-FUNK: can you roll it back from live cd?16:10
Q-FUNKazeem_: you'll notice that this decision will essentially kill LTSP on Ubuntu, btw.16:10
chrysnwhen modifying a debian /contrib package for ubuntu, should it go to universe or multiverse?16:11
azeem_blue_anna: I don't Q-FUNK ask for support16:11
azeem_asks*16:11
Q-FUNKblue_anna: yeah, that's not a problem.  the real problem is ubuntu distributing packages marked as i386 that are really meant for i686.16:11
azeem_eh, that's not the problem16:11
blue_annaok, yea that's a problem16:11
azeem_i386 is the processor architecture16:11
blue_annaglad to hear you're system's not fried :)16:11
azeem_that 10.04 already requires i586 I guess, 10.10 would just bump it to i68616:12
chrysn(it's in contrib as it depends on libcgal, which is qpl and therefore in nonfree)16:12
Q-FUNKazeem_: libc6 was built with -m686, if doko's reply means anything.16:12
bluefoxicythis is strange16:36
bluefoxicyI have trouble connecting to a wifi access point16:36
bluefoxicywhat happens is when I connect to it properly, iwconfig shows the correct ESSID16:36
bluefoxicywlan0     IEEE 802.11bg  ESSID:"Public Hotspot - BIHS"16:36
bluefoxicyHowever, in other times when it refuses to connect...16:36
bluefoxicybluefox@icebox:~$ sudo iwconfig wlan0 essid dogs16:37
bluefoxicybluefox@icebox:~$ sudo iwconfig16:37
bluefoxicywlan0     IEEE 802.11bg  ESSID:"dicks"16:37
bluefoxicybluefox@icebox:~$ sudo iwconfig wlan0 essid "Public Hotspot - BIHS"16:37
bluefoxicybluefox@icebox:~$ sudo iwconfig16:37
bluefoxicywlan0     IEEE 802.11bg  ESSID:"%\xCF\x08\xF5\xE9\xE2^S`\xAA\xD2\xB2\xD0\x85\xFAT\xD85\xE8\xD4f\x82d\x98\xD9\xA8\x87uepZ\x8A"16:37
bluefoxicyer.  Censor fail <_<;16:37
bluefoxicyIgnore the output of my turrets.16:37
bluefoxicyat any rate a second try does this16:38
bluefoxicywlan0     IEEE 802.11bg  ESSID:"?b\x80)D\xDE|\xA5\x89NWY\xD3Q\xAD\xAC\x86\x95\x80\xEC\x17\xE4\x85\xF1\x8C\x0Cf\xF1|\xC0|\xBB"16:38
bluefoxicyand so on.  Constant garbage16:38
crimsunQ-FUNK: I'm presuming that doko didn't follow the link from ProcVersionSignature/Uname in the bug report. Otherwise, were you aware of https://blueprints.launchpad.net/ubuntu/+spec/foundations-m-686-compile ?17:01
Q-FUNKcrimsun: I wasn't aware of it and now that I am I cannot agree with it.17:02
ScottKUntil you convince someone to change the design decision, it's pretty clearly not a bug.17:11
Q-FUNKScottK: bedian went through this one when they bumped up the minimal requirement on sparc.  it would be a good idea to follow what they did, to prevent upgrades from taking place on older hardware.17:12
Q-FUNKdebian, even17:12
ScottKI think that's a valid point.17:13
ScottKIf you reframed the bug that way, I think it's reasonable.17:13
Q-FUNKScottK: well, there is a larger issue at stake:  building with anything newer than -m586 essentially kills the whole LTSP project's ability to run on Ubuntu, since most terminal hardware is 586 generic.17:16
Q-FUNKnoticing that the blueprint for that m686 idea is a blank template also makes the proposal rather suspicious.17:18
crimsunI trying (and failing miserably) to find a recent mailing list post regarding the UDS-Maverick Foundations outcomes, but there's http://oubiwann.blogspot.com/2010/05/ubuntu-foundations-and-maverick-meerkat.html17:19
jpdsQ-FUNK: http://pastebin.ubuntu.com/441398/17:22
Q-FUNKjpds: interesting.17:24
crimsunoh good, I was having problems connecting to the gobby doc17:24
Q-FUNKall this shows is tha choosing m686 was more of a me-too choice than a properly planned change.  m586 wasn't properly investigated, according to that.17:26
Q-FUNKit also shows that whoever is gonna upgrade from lucid to lucid+lts is gonna have a nasty surprise.17:27
* ogra_tb is pretty sure update-manager will have proper warning mechanisms builtin by then17:27
Q-FUNKogra_tb: back when debian bumped the sparc requirement, they installed a check in libc6's preinst script.   if the hardware was no longer supported, the package refused to be updated.17:31
blue_annawhat provides all the fstab functionality? getmntent(), etc17:31
ogra_tbwell, there are more elegant ways in ubuntuu17:31
Q-FUNKupdate-manager is not one of them.  it wouldn't be foolproof17:33
ogra_tbsolutely foolproof17:37
ogra_tb*it's absolutely17:38
ogra_tband already has such checks for ARM17:38
=== edson is now known as ecanto
blue_annais 10.10 on eglibc2.12 or still 2.11?17:57
Q-FUNKtoday was the first 2.12 snapshot upload17:58
aburchI have one backport request which was ack'ed a month ago but hasn't been uploaded since (LP #550880). Is there anything left to do?18:07
ubottuLaunchpad bug 550880 in Karmic Backports "Please backport simutrans-pak64/102.2.2-1 from Lucid" [Wishlist,In progress] https://launchpad.net/bugs/55088018:07
blue_annaanyone know what handles the UUID /LABEL translation in /etc/fstab? from the manpages you would think it is mntent.c but I just read that and mntent_r.c and its not there.18:13
blue_annaso I grepped the entire eglibc tree for UUID and it's nowhere in the code18:14
blue_annajust trying to find a foothold18:19
geserblue_anna: try looking at libblkid1 as its description seems to match what you look for (and mount depends on it)18:20
blue_annageser: thank you18:22
blue_annaanyone here use yaml in C?20:00
ScottKaburch: No.  It just waits an archive admin to process it.20:18
stefanlsdnigelb: heys. give me a shout when your around. exams finished, so can spend some time on cleansweep21:19
blue_annastefanlsd: you had exams last week?21:19
blue_annait's almost june21:19
stefanlsdblue_anna: mm. isnt that normal exam time? (uni)21:21
blue_annadunno, been so long since I was in college :) my brother just finished his about 4 weeks ago21:21
stefanlsdblue_anna: im also in africa, stuff happens slower here :)21:25
blue_annalol I doubt that -- but maybe classes21:26
Sarvattis cifs-utils going to brought over from debian for maverick? smbfs hasn't been installable for about a month, had to grab cifs-utils which replaced it from debian to mount some windows shares21:29
geserSarvatt: most likely but the archive admins have to review the list of NEW packages before syncing them which takes time. You might want to poke an archive admin (e.g. cjwatson) to sync it faster.21:32
* wise_crypt is away: sleeping 21:41
chervacan someone help me with to command to get the current source for gnome-screenshot from svn.gnome.org ( http://svn.gnome.org/viewvc/gnome-utils/trunk/gnome-screenshot/ )21:46
* wise_crypt is back (gone 00:05:37)21:46
chervaso I can try to tweak it a little for 10.1021:46
* wise_crypt is away: 21:47
arandcherva: "GNOME has changed to using Git for version control. Current GNOME sources can be found on git.gnome.org. All content on this site is obsolete"21:52
chervaarand, 10x21:52

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!