/srv/irclogs.ubuntu.com/2007/10/30/#ubuntu-devel.txt

chowmeinedare there plans to adopt debian's new graphical installer for the alternate CDs?00:39
Keybukmjg59: dial-up-support ... NM 0.7 ... win00:41
Keybukchowmeined: that seems to defeat the *point* of the alternate CDs00:41
chowmeinedreally?00:42
chowmeinedi dont use the alternate cd because its non-graphical.. i use it because it lets me setup raid and lvm00:42
Keybukthe alternate CD is intended for situations where the ordinary desktop CD isn't suitable00:42
chowmeinedand of course it would still be an option to use the text installer.. just like the debian installer provides00:42
chowmeinedexactly00:43
chowmeinedlike when i want raid and lvm00:43
Keybukbut making it graphical would make it unsuitable for "when graphical doesn't work"00:43
chowmeinedhence the text mode option00:43
Keybukthe option would make it twice as hard to support00:43
Keybuksince we'd have to test it twice00:43
Keybuktext mode only means it works for both groups of users00:44
Keybukyou can still set up lvm and raid when it's text mode00:44
chowmeinedwell 2 times 0 is still 000:44
Keybukwhy zero?00:44
chowmeinedby continuing the policy of not testing00:44
Keybukwhat policy?00:44
Keybukwe *heavily* test all the CD images we release00:44
chowmeinedwell thats a different topic00:45
Keybukwe spend an extraordinary amount of time doing it00:45
chowmeinedand most of the normal things work just fine00:45
chowmeinedits excellent what is done00:46
Keybukadding yet another permutation would make that job harder00:46
Keybukespecially when that permutation has limited value00:46
chowmeinedbut i still managed to rack up about 10 bugs since gutsy was released00:46
Keybuksure, you get bugs00:46
Keybukthere's always bugs00:46
chowmeinedyes00:46
Keybukreleasing bug-free software is impossible00:46
chowmeinedit is00:46
Keybukand, in fact, we don't even attempt to fix all the bugs we find00:46
chowmeinedwhy not?00:46
Keybukbecause sometimes fixing the bug is harder than releasing it00:47
Keybukknown bugs are better than unknown bugs00:47
chowmeinedfair enough00:47
pwnguinhopefully the automated testing stuff will bring up some nice regression testing etc00:47
chowmeinedthen it would be good to remove the options which expose the bugs00:48
chowmeinedlike "Suspend" from the menu in gdm00:48
Keybuka graphical installer on the alternate CD doesn't really have a use case00:48
Keybukchowmeined: hmm, that works?00:48
chowmeinednot on any computers ive tried it on (6)00:48
Keybukworks for me00:48
Keybuksuspend/resume is notorious00:48
chowmeinedit works when im logged in, it works with pmi action suspend00:48
chowmeinedbut not from the gdm login screen00:48
pwnguinwhich is why its not supposed to show up if it doesnt think it can do it...00:49
Keybukthat's not a question about the installer though00:49
Keybuksince the installer doesn't use gdm00:49
chowmeinedi changed topics00:49
Keybukah, I was yanking us back to the topic00:49
chowmeinedi dont want to talk about the installer anymore00:49
Keybukbugs go in http://bugs.launchpad.net/ubuntu00:49
chowmeinedbecause you made your point and it seems fai00:49
chowmeinedfair*00:49
chowmeinedi know where bugs go :) ... anyways its too late, ill save it for hardy00:50
Keybukerrr00:50
Keybuksaving the bug is a biiiit late :p00:50
Keybukwe're developing hardy *now*00:50
pwnguinhardy's open -- no CDs out yet of course00:50
chowmeinedi realize this00:50
chowmeinedno ISOs though00:50
Keybukif you file the bug in six months time, after hardy is released, then it definitely won't get fixed for hardy00:50
chowmeinedill file it when alpha 1 comes out00:50
Keybukwhy?00:51
Keybukif you know the bug is already there, why not file it now?00:51
chowmeinedbecause i dont know00:51
chowmeinedits in gutsy.. but00:51
Keybukif you don't file the bug ...00:51
chowmeinedmaybe they fixed the infrastructure issue00:51
Keybuk... then how will anyone know to fix it?00:51
pwnguinchowmeined: you might appreciate the stuff liw is doing with automated testing. https://wiki.ubuntu.com/MeetingLogs/openweekgutsy/AutoTests00:51
chowmeinedsomebody already reported the bug for gutsy.. i said my "me too".. and theres still no response00:52
chowmeinedso ill wait for hardy00:52
chowmeinedalphas00:52
Keybukyou didn't say it had been already reported :)00:52
chowmeinedpwnguin, yea automated testing is nice, i tried using dogtail.. it was a pain00:52
chowmeinedbut i mean.. the theory is nice00:53
chowmeinedgetting a 'low' priority is a death sentence for bugs ..00:53
chowmeinedso is it possible for me to test software before it is released in an ISO? can i just install single debs in gutsy from the hardy repo?00:58
chowmeinednot core ones of course..00:59
Keybukyes00:59
KeybukBurgundavia: around?01:00
BurgundaviaKeybuk: pong01:14
KeybukBurgundavia: whatever happened to ubuntu demon and spamming planet about his killer hard drive of doom ?01:14
sladenKeybuk: maybe his hard drive died...?01:25
Keybuksladen: I'm going to buy new shoes this week01:26
Keybukyou should come too01:26
Keybukor you could have my old ones01:26
sladenKeybuk: touché01:31
KeybukBurgundavia: does easy-ldap-server need another session?01:44
BurgundaviaKeybuk: no01:47
KeybukBurgundavia: let me reach for my figlet01:47
KeybukCOULD YOU SET IT AS DRAFTING THEN PLEASE01:47
Keybuk;)01:47
BurgundaviaI have no idea how it ended up on the uds-schedule01:48
KeybukThis topic was proposed by  nijaba  on 2007-10-01. It was previously marked "Accepted" for the agenda by  Rick Clark  11 hours ago.01:48
Burgundaviaright01:48
Burgundaviadeclared informational and superceded01:50
thullyhi - I'm trying to track down a bug by rebuilding git snapshots of my kernel, and after one build I'm getting a "debian/rules not found" error03:56
thullyHow do I get debian/rules back?  I need it to build proper kernel .debs...03:57
imbrandonis it not found or not executable ? iirc it needs to be set +x and the error looks similar04:10
imbrandonother than that do a new git co04:10
imbrandonKeybuk, you get stuck with the scheduling this time ? heh fun fun04:11
mekiusbryce: did you happen to receive my message yesterday?04:11
thullyimbrandon: it's not there at all, the last build simply ate the file...04:18
=== simon__ is now known as frostburn
=== lucas__ is now known as lucas
=== Gman is now known as GmanAFK
=== asac_ is now known as asac
sebastian^good morning folks08:46
gaspadoko: ping10:34
=== Adri2000_ is now known as Adri2000
Kopfgeldjaegerhi12:21
=== norsetto is now known as norsetto_limbo
* Hobbsee waves to fabbione12:46
simiramorning fabbione12:46
fabbionehi ladies12:46
simirais Tollef still asleep or what?12:47
fabbionesimira: dunno.. i am not at UDS12:47
zulhey fabbione12:48
fabbionehi zul12:48
simirafabbione: you're not? Why?12:48
fabbionesimira: because i am not distro team anymore12:49
fabbionehaven't been for the past 3 months or so12:49
simirafabbione: oh, I thought everyone was at the uds... so you're not going until next week, then?12:49
fabbionesimira: i am flying out tomorrow for 2 days of vacation in Boston12:50
fabbionewill meet the others next week12:50
* fabbione needs to visit his grandma too12:50
simirafabbione: that sounds nice. Hmm, maybe I should get Tollef out of distro... :p12:50
fabbionesimira: well he was not in San Fran a month ago :)12:51
simirafabbione: true, he was in Boston a month ago :p12:51
loolasac: Hey; I just prepared pango 1.19.0 and it does a good job at uglifiyng firefox' font rendering; we should take a look together13:02
=== norsetto_limbo is now known as norsetto
=== elkbuntu is now known as elkbirthday
sladensimira: haven't seen tollef yet13:05
simirasladen: he's online now. :)13:06
simirasladen: so, you got there in time, then13:06
sladensimira: just made it, even with the adventure involved :)13:07
asaclool: oh .. ok13:08
pittihttps://lists.ubuntu.com/archives/ubuntu-devel/2007-August/024101.html13:08
joejaxxok13:08
=== pedro is now known as pedro_
calcHobbsee: good morning13:09
Hobbseehiya calc!13:11
=== Pici- is now known as Pici
=== sladen changed the topic of #ubuntu-devel to: Archive: OPEN | Development of Ubuntu (not support, even with hardy; not application development on Ubuntu) | #ubuntu for support and general discussion for dapper/edgy/feisty/gutsy | #ubuntu-motu for getting involved in development | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://wiki.ubuntu.com/HelpingWithBugs | Hardy opened, go wild! | UDS in #uds-boston
=== pitti_ is now known as pitti
warp10pitti: May I query you?13:48
jdstrandmdz: fyi, rick just added ubuntu-firewall as the next session (9:50)13:52
mdzhe emailed me, thanks13:53
=== bryce_ is now known as bryyce
jdstrandslangasek: fyi ubuntu-firewall is scheduled now (it just changed)13:54
warp10pitti: May I query you?13:56
pittiwarp10: hey!13:58
pittiwarp10: a bit unstable internet here, but go ahead13:58
=== norsetto is now known as norsetto_limbo
wasabiIs there anyway to use dpkg to declare a Depends to a version number of some sort which is fixed for ubuntu release? For instance Depends: something (= 7.10)14:20
wasabiActually, let me reverse that. I want to declare a Conflicts: distro (> 7.10)14:21
mjg59No14:22
mjg59Conflict on the package that provides the functionality you conflict with14:22
wasabiProblems is that isn't known yet.14:22
wasabiI think I'd have to explain the entire problem domain for this to make sense.14:22
wasabimjg59: I'm thinking about the case of how third party software (closed source perhaps) might distribute .deb files for use on Ubuntu in a way as to ensure their customers have the best experience.14:23
wasabiThat is, their software has been tested on feisty, but they aren't willing to commit to running it on gutsy, yet.14:24
mjg59wasabi: No, there's no mechanism for doing that, and we won't add one14:25
mjg59You can check /etc/lsb-release at runtime and warn the user that it's unsupported14:26
wasabiHmm.14:27
wasabiYeah, that's probably a decent way. Do you see what I'm trying to accomplish though? I'd like the update manager to tell the user he can't update to gutsy until he addresses the closed source package.14:27
wasabiAs in, either updates it to a version that supports gutsy, or removes it.14:27
mjg59No, that's the wrong solution14:28
mjg59There's no plausible way for an application to only work on feisty unless it fails to declare its depends properly or unless we've screwed up14:28
wasabiSure there is.14:28
wasabiConsider vmware.14:28
mjg59That's not a feisty/gutsy issue.14:28
wasabiHmm perhaps.14:29
wasabiYou're right.14:29
mjg59I can replace the kernel on feisty14:29
Amaranthyou can use the feisty kernel on gutsy too14:29
wasabiYeah, okay, you win there.14:29
mjg59Yeah14:29
wasabiI guess it'd be fine if I could believe Ubuntu would never break ABI or something. =)14:29
mjg59One thing we could do would be to have the kernel do Provides: vmware-modules14:29
mjg59Or something14:30
mjg59And then have vmware depend on that14:30
wasabiEh. Vmware's software would not be able to alter your packages.14:30
wasabiAt least, that's my assumption.14:30
mjg59Correct14:30
mjg59It's a change we'd have to make14:30
wasabiHmm. So I guess VMware would need runtime logic to detect the missing modules and install the appropiate package.14:31
wasabiThat's really the only solution I can see.14:31
mjg59No?14:31
wasabiWell, how would it then?14:31
mjg59Well, sure, for the current versions we're boned14:31
wasabiWhat would you suggest going forward then?14:31
mjg59Enhance our kernel packages to provide: the functionality they provide that other things may depend on14:32
wasabiI don't see how that helps.14:32
mjg59Oh, wait. We do that already.14:32
wasabiUh huh.14:32
mjg59wasabi: So vmware does depends: vmware-modules14:32
wasabiOkay, let me rephrase my question.14:32
wasabiI'm talking from teh POV of Joe Blow ISV.14:33
wasabiI have some software I want to distribute for Ubuntu.14:33
wasabiI'm not going to be able to convinc eyou guys to add stuff to your kernel packages.14:33
mjg59If you depend on a specific kernel module ABI, then yes, it is14:33
wasabiWell, VMware again is the best case.14:33
wasabiI doubt you'll want to add their kernel modules to your packages.14:34
wasabiI doubt they'd want you to... as it would mean they couldn't deploy updates.14:34
mjg59Right. So, where do the modules come from?14:34
wasabiVMware's own apt repository.14:34
mjg59So I'm struggling to see the issue14:34
mjg59If the user has the correct modules and kernel installed, they can run vmware14:34
mjg59So vmware should just depend on them14:35
wasabiUh huh. Which correct ones?14:35
mjg59Whichever ones the package requires14:35
wasabiIf feisty has 3 kernel versions available. I need 3 module packages available.14:35
mjg59No. The user could boot any one of them.14:35
mjg59You only need one of them.14:35
wasabiYeah, exactly.14:35
wasabiEh?14:35
mjg59If I have 2.6.20 and 2.6.22 installed, I only need vmware modules for one of them14:35
mjg59Because if I want to run vmware, I can reboot into 2.6.2014:35
wasabiNo, you need vmware modules for whichever you want to use at the given time.14:36
mjg59No.14:36
wasabiWait you're suggesting we force users to reboot into old feisty kernel versions?14:36
wasabiThat's silly.14:36
mjg59No, I'm saying they can reboot into an old version14:36
mjg59So it's a runtime issue, not a package dependency14:36
wasabiI know that.14:36
wasabiWhich is why I originally suggested that the vmware program At Runtime, detect that you're missing the modules for whatever kernel you're running.14:37
mjg59Ah, right. That's an orthogonal issue.14:37
mjg59You should still have the dependency in the packaging system14:37
wasabiOh, yes. The specific Vmware-modules packages should depend on the specific Ubuntu kernel package.14:37
mjg59But yes, beyond that it's a runtime concern14:37
wasabiAnd VMware should depend on something all 3 provide.14:38
wasabiBut at runtime a bit of checking has to go into making sure they are actually installed.14:38
wasabiOkay, kernel modules probably aren't the best example to start this off with.14:38
mjg59Well, whether or not /dev/vmware exists14:38
mjg59Or vmmon, or whatever it is14:39
wasabiYeah. Whatever.14:39
* mjg59 goes to change session14:39
wasabiMy main concern is how to properly instruct the user that he should not update to gutsy until he updates some program.14:39
wasabiOr simply not to, and let his program break.14:39
wasabiThere is sort of a dependency between a piece of software and 'gutsy'. The dependency exists because Ubuntu has committed to 6 month stable releases. It's provided something to depend on.14:40
sladenthere's the kernel abi-versioning dependances that red-hat use14:47
traeneHi. I am looking for help for file locking.15:00
Chipzztraene: please read topic15:02
traenesry15:03
gribouillehi15:21
gribouillewhy isn't there a Contents-i386.gz file in http://archive.canonical.com/ubuntu/dists/gutsy/ ?15:22
Riddellmjg59: about?15:43
mjg59Riddell: Yo15:43
Riddellable to join us in room A briefly?15:43
mjg59Sure15:43
mjg59Give me a sec15:43
=== pedro is now known as pedro_
Adri2000soren: now pbuilder builds successfully16:19
=== gnomefre1k is now known as gnomefreak
sorenAdri2000: You are right. I'll upload later today.16:32
Burgundaviakeybuk_: I don;t seem to be included ont eh agenda. Can you pull a new csv from LP?16:53
Burgundaviathe attendees16:53
=== pedro is now known as pedro_
=== ryu2 is now known as ryu
=== norsetto is now known as norsetto_limbo
tuxmaster1988Does the ATI X1300 have hardware acceleration under ubuntu?18:09
sharkphi18:15
sharkpMaybe I've found a bug in the kernelù18:15
sharkpI've a laptop with two sata hd+18:16
sharkpbut I can't see (on gutsy) them18:16
sharkpI think there's a problem between kernel and sata controller18:16
sharkpand I need to know how module I have to load manually18:17
sharkpwould you see my lshw?18:17
sharkpoh, guys, here we are in 24418:18
sharkpisn't there anyone who reply me?18:18
slangasekthis is not a user support channel; see the topic and ask in #ubuntu?18:20
sharkpslangasek:  I can read18:20
sharkpIMO it's a bug18:20
Burgundaviathis is also not a bug support channel18:20
Burgundaviaif it si bug, it goes int eh bug tracker18:21
sharkpand, always IMO, this is a developement chanmnelù18:21
Burgundaviaif you have a solution for the bug, please include that in the bug18:21
sharkpBurgundavia: is it so expensive for you to reply me?18:21
sharkpand to know my problem?18:21
slangasek"how do I load a module" is not a development question, nor is "what information do you need for a bug report"18:21
sharkpok, I've understood this is a nerd-chan18:22
sharkpExcuse me for my noob opinion18:22
sharkpgood work18:22
Burgundaviathat was fun18:22
Burgundaviahey jo18:22
Burgundaviajono:18:22
slangasekI wonder if nerd-chan is anything like Super Milk-Chan18:23
Burgundaviacan I be part of that?18:23
BurgundaviaI feel very left out18:23
PiciUgh.  I told him to ask about the issue in #ubuntu. Not come bother people in #ubuntu-devel.18:24
slangasekPici: well we're not particularly bothered, but everyone's at UDS and he's not going to /find/ the help he needs in here right now no matter how much he insists :)18:25
loolAre recommends honored when germinating?18:35
Skiessican someone update libtorrent10 and rtorrent packages of the hardy repository? or should I ask in #ubuntu-motu?18:36
* ogra1 wonders if hardy is actually in a usable state yet18:39
SkiessiI'm using feisty, gutsy and hardy repositories and it's working great18:40
* ogra1 shudders18:41
BurgundaviaSkiessi: please don't recommend such ideas18:43
SkiessiI don't recommend such ideas18:43
Skiessiat least not for developers18:44
=== Pici- is now known as Pici
MithrandirStevenK: can you give me a sloccount of modest?19:07
StevenKMithrandir: Er, sure.19:10
StevenKMithrandir: Total Physical Source Lines of Code (SLOC)                = 53,32919:12
wasabiFrickin'. /.ing launchpad.19:13
Keybuk;)19:13
KeybukI blame ubuntu-demon19:13
superm1Keybuk, is it too late to regenerate the schedule for the rest of the night?  bryce updated his spec to be 'drafting' now19:14
wasabiAnd finding somebody to blame is the most satisfying solution!19:14
Keybuksuperm1: yes19:15
superm1:(19:15
MithrandirStevenK: ok, thanks.  That's not that bad..19:18
coderhi19:21
=== pedro is now known as pedro_
sorenAdri2000: I haven't forgotten, by the way... :)20:21
Mithrandirhm, why does bzr say 'No loggers found for logger "bzr"' when I commit something?20:35
davf_Can anyone tell me how I can patch just one kernel module without having to recompile the whole kernel?21:21
Keybukmjg59: still hasn't gone up ;)21:23
davf_?21:24
jonmasterssladen: ping21:24
davf_Can anyone tell me how I can patch just one kernel module without having to recompile the whole kernel?21:24
sladenjonmasters: hi capt fat21:24
sladenhile:21:24
davf_Please?21:25
jonmasterssladen: you in town?21:25
sladenjonmasters: ABI versioning came up earlier, if BenC is aorund, that might be something to discuss21:25
jonmastersdavf_: do you have a kernel source tree prebuilt, or not?21:25
sladenjonmasters: yes, one street away from you MUWAHAHAHAH21:25
warsocketI promised this to the head of PackageKit to ask you guys this question21:25
warsocketso here it goues: What do you guys think about checkinstall;21:25
jonmasterssladen: I have some stuff to get done tonight debugging something, but it'd be nice to say hi. Are you around now?21:26
sladenjonmasters: however, after just 5 meals at this hotel, I can tell why all americans I somewhat large21:26
davf_jonmasters: I just installed gutsy on this machine and started d/l kernel build packages21:26
sladenjonmasters: for some value of round.21:26
jonmasterssladen: I'm 164lbs now.21:26
jonmasterssladen: size 29". I was size 38" when I moved here and when you last saw me.21:26
jonmasterssladen: shall I head over and say hello now?21:27
davf_jonmasters: last time I did this for Feisty I rebuilt the kernel but I don't remember what I did and I want to avoid doing the same thing again if pos.21:27
jonmastersdavf_: you can hack the source for a given module, make sure the .config is in place (do an make oldconfig if needed) then run make against just a given directory.21:27
jonmastersmake drivers/path/to/foo/foo.ko21:28
jonmasterssladen: are you at the hotel now?21:28
sladenjonmasters: we should finished at 18:0021:29
sladenjonmasters: if you waddle your 82kg into a car and drive down the block you should have found somewhere to park by that time21:29
davf_jonmasters: ok so I got build-essentials linux-source and kernel-package. Any thing else?21:30
davf_jonmasters: btw what is the source dir?21:30
davf_jonmasters: nevermind found it.21:31
davf_jonmasters: so I need to untar linux-source where it is yea?21:32
bddebianHeya21:39
lamegohey bd21:39
=== superm1_ is now known as superm1
bddebianHi lam21:39
bddebianGrr, lamego21:39
lamego:P21:40
davf_jonmasters: so to do the make on a patched module only when I type make in that dir it says no targets. what should the target name be (recompilig zd1211rw)22:01
davf_Or if anyone else can advise?22:02
davf_Anyone can help me. I just want to "make" one kernel module that I needed to patch how do I do it?22:13
davf_I did make oldconfig22:13
davf_went to the dir and tried make and got no target22:14
mekiusHi, working on a custom live cd, I updated the kernel that sits in the squashfs and attempted to copy the kernel and initrd to the live cd in /casper/vmlinuz and /casper/initrd.gz respectively, but the initrd doesn't seem to contain the squashfs module.  I looked in /lib/module/2.6.22-14-generic/ and there is an ubuntu dir and that has the squashfs module in it, but this doesn't get pulled into the initrd image, any pointers?22:43
=== norsetto_ is now known as norsetto
tekteencan someone help me with my preseed file? It does not answer the questions but I do not get an error that it can't find the preseed. The file is pasted at http://paste.ubuntu-nl.org/42757/23:27
tekteencan someone help me with my preseed file? It does not answer the questions but I do not get an error that it can't find the preseed. The file is pasted at http://paste.ubuntu-nl.org/42757/23:33

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