/srv/irclogs.ubuntu.com/2007/11/24/#ubuntu-devel.txt

Burgundaviaslangasek: just to catch up on a previous conversations, we are going with boring Alpha designations for Hardy?01:28
slangasekBurgundavia: correct01:28
Burgundaviahow very boringly corporate01:28
slangasekdecided before I was involved, I'm just doin' what I'm told01:29
BurgundaviaI liked the names, thought they added character, but they were confusing01:29
slangasekthey also guaranteed url uniqueness on the website, but oh well :)01:30
Burgundaviayes, they did01:30
jcastroUbuntu Alpha Vorlon 1 released!01:30
slangasek...01:30
Burgundaviaany idea who made the decision? did it come from sabdfl himself?01:30
jcastroI wish there would have been a $something Crow so there could have been Murder 1, 2, etc.01:30
Burgundaviagiven the "boringly corporate" theme for Hardy, I wouldn't be shocked01:30
slangasekBurgundavia: dunno, was relayed to me by pitti and cjwatson_01:31
Burgundaviaok01:31
Fujitsujcastro: Hahah.01:31
Burgundaviajcastro: launch a deriv and have a crazy crow release01:31
HobbseeBurgundavia: mdz01:32
slangasekI don't have the impression that LTSness was a significant consideration, since alphas are rather non-user-oriented anyway01:32
BurgundaviaHobbsee: hmm, ahh01:32
Hobbseejcastro: *grin*01:33
Burgundaviaslangasek: but they are tech press oriented, non Linux tech press01:33
slangasekI don't know I'd say they're "oriented" toward the tech press, though the tech press certainly covers them and this isn't a bad thing :)01:34
Burgundaviaright, that is what I meant01:35
GoldenPonynxvl: do you still need it?01:58
* Fujitsu didn't know that ponies were good at eating LongPointySticks.01:59
* GoldenPony is duplicating LongPointyStick02:00
* GoldenPony works well, for flattening people02:00
FujitsuExcept when it gets shot, like a couple of days back.02:00
* GoldenPony cannot be shot.02:00
* persia wonders if GoldenGlue would be useful02:01
=== Hobbsee_ is now known as Hobbsee
Hobbseeoy, network manager - stop dying!03:45
Hobbseesame to my wifi!03:46
somerville32Network manager says I have no network devices03:47
somerville32I'd think my computer was pretty awesome to be able to use the internet without one if only I didn't know better03:48
lamontHobbsee: network managler hasn't run on my laptop in a small eterinity03:53
lamonteternity, even03:53
lamontadmittedly, that's mostly because I purged it.03:53
Hobbseehehe03:53
DoYouKnowhi03:53
Hobbseewell, that tends to have an effect03:54
lamontHobbsee: it makes my network stay up better. :-)03:54
Hobbseewhat's weird is that this isnt the mangler - it's actually the fact tha tmy card randomly stops flashing / being active at all03:54
pwnguinso far i havent found anything terrible with my intel card and network manager03:58
Hobbseewhat's more weird is that this is only happening today03:58
pwnguinHobbsee: maybe you're just being hax'd03:59
Hobbseepwnguin: maybe it's my machine giving me hell for not running gutsy03:59
pwnguinpsh03:59
* Hobbsee glares at mpt_03:59
pwnguinhardy's fine if you know where the landmines are ;)04:00
pwnguinthy name is policyKit04:00
* Hobbsee wonders if this is the real mpt.04:00
* persia recommends chroots04:00
pwnguinchroots dont test nvidia04:00
* somerville32 enjoys multiple machines.04:00
persiapwnguin: Exactly :)04:00
pwnguinor the rest of my laptop hardware04:01
* Hobbsee discovered an edgy install on her other machine.04:01
pwnguinthink i just found a bug in rhythmbox though04:04
* Hobbsee ponders whether -updates can be abused, for kde4 related stuff.04:28
LaserJockHobbsee: wouldn't -backports be the place for that?04:33
HobbseeLaserJock: normally, yes.  but the current version is useless for building kde4 apps - which is it's only use, iirc04:34
persiaHobbsee: The issue is verification: you'd need to show a regression (which requires tricky wording), and demonstrate no impact (which likely means rebuilding all the rdepends and testing them agressively, which would take weeks, if not months).04:35
LaserJockHobbsee: but what's the functional difference between -backports and -updates?04:35
HobbseeLaserJock: backprots is for random crack, to use selectively04:35
Hobbseesarah@LongPointyStick:~% rbuildepend libqca2-dev | wc -l                 3:35PM04:36
Hobbsee1204:36
persiaErr.  Not "random crack", but stuff from the new dev release that compiles on the old release04:36
Hobbseemost of which is already kde4 stuff, or qca2 stuff.04:36
Hobbseeoh, and psi04:36
persiaheh04:36
* Hobbsee thought psi had to build with theold version04:36
Hobbseethat can still go back to libqca-dev.  well, that's what it's shlibdeps have come up with.04:37
Hobbseeweird.04:38
Hobbseepersia: mmm.04:38
* Hobbsee just hopes hardy gets sorted soon04:39
persiaHobbsee: You could fix all the NBS stuff, and it'd be in better shape :)04:39
Hobbseepersia: wouldnt help the intel breakage and such :)04:40
Hobbseebesides, its' hard to do such a thing, when your machine hardlocks twice in 25 mins.04:40
persiaHobbsee: Ah: video drivers.  No, not as such.  I'd wait for the new kernel04:40
Hobbseejust hoping it comes soon :)04:40
Hobbseegot caught only upgrading parts of hardy, causing more random freezes04:40
* persia suspects they weren't random, and suggests aptitude's advanced dependencies resolution algorithm to detect breakage before installation.04:43
Hobbseewell, it was update-manager - then other bits would blow up04:44
Hobbseeso then i said "screw it, i'll just upgrade the rest, and hope it works better than it did"04:44
Hobbseewhich it does.  in some ways.04:44
Hobbseeit works fine, until it crashes.04:44
Hobbseerather than working slow and crap, until it crashes :)04:44
Hobbseealthough the crashes are increasing04:45
persiaHobbsee: update-manager does a really smart dist-upgrade, but it's less smart about incremental daily uploads: it tends to grab the new stuff, and hope there aren't any library collisions.04:45
Hobbseeyeah, true04:45
persiaAlthough, if your performance is that bad, I should really upgrade: my system will probably stop crashing :)04:45
Hobbseehehe04:45
HobbseeFujitsu's got it as well04:45
Hobbseeanyway... /me --> vote04:46
* persia dist-upgrades on the basis of this information04:46
Hobbseeright.  have voted.05:06
* mdomsch is going to like dpkg triggers05:06
Hobbseemdomsch!05:06
mdomschgood evening Hobbsee05:07
Hobbseeheya mdomsch05:07
Hobbseemdomsch: how do you like ppa?05:07
mdomsch Hobbsee just got it activated for dell-team, haven't uploaded anything yet05:07
Hobbseeget to it :P05:08
mdomschbeen too busy fixing my packages and playing wii with my kids05:08
mdomsch:-)05:08
Hobbseehehe05:08
mdomschuploading firmware-tools now05:12
Hobbseehrm.  what's the factoid for customising the ubuntu cds....05:12
mdomschand firmware-addon-dell05:14
mdomschHobbsee, https://launchpad.net/~dell-team/+archive has them now05:22
Hobbseemdomsch: \o/05:23
Hobbseemdomsch: i thought you were going for all releases05:23
Hobbseenot just hardy05:23
mdomschHobbsee, I need to change and re-upload for gutsy05:23
mdomschand because I need triggers, I can't release for feisty :-(05:24
Hobbseemdomsch: you know that you cant use the same version number for gutsy and hardy?05:24
mdomscheww05:24
mdomschfwiw, fedora uses a "dist tag" appended to the -revision part of the version field05:25
Hobbsee(due to archive structure, where all the binaries get put in pool/)05:25
mdomschwhich solves this cleanly05:25
persiaWell, actually the same number can be used for both, but it's not ideal for new uploads.05:25
Hobbseemdomsch: you can do the same, or similar.  use 1.4.9-0ubuntu2~<targetrelease>105:25
mdomschworksforme05:25
Hobbseepersia: not in ppa, surely.05:25
Hobbseepersia: http://ppa.launchpad.net/dell-team/ubuntu/pool/main/f/ doesn't do releases05:26
persiaHobbsee: Right.  Context.  Same binary == same version, but dist copies only happen to the main archive.05:26
* persia shuts up again05:26
mdomschHobbsee, must be different, not gutsy < hardy ?05:26
Hobbseemdomsch: it so happens that gutsy < hardy anyway, but yes, they must be different.  and one must be above the other, if you want to stop warnings about "help, you're downgrading"05:26
Hobbseepersia: that being said...we cant upload teh same version to gutsy and hardy, can we?05:27
Hobbseein the main archive05:28
Hobbseemdomsch: thanks, that's part of what i needed to stick in the all new and shiny ppa FAQ05:28
persiaHobbsee: Err.  From a dpkg perspective it works fine, as long as it's the same binary.  The trick is that the Packages,gz files need to have the right contents, which depends on the archive system being used.05:28
Hobbseepersia: true.  it's an archive limitation, not a dpkg one05:29
persiaHobbsee: I'm not sure "limitation" is the right word, but yes.05:29
Hobbseewell, it's on purpose.  but yes :)05:29
persiaNote that same binary means same compilation: always copy from older to newer, rather than the other way, or things are likely to break, as forward compatibility is a difficult problem.05:30
Hobbseepersia: this is true05:33
=== Shely is now known as Huahua
=== Huahua is now known as Shely
mdomschpersia, in my case, the package is a python package, so compilation happens at install time05:49
mdomschbut if it were C I would agree05:50
persiamdomsch: Even python packages have a build process.  If python-central | python-support changed significantly, you'd prefer to build against the old one, to ensure that the install-time compilation worked for both releases.05:51
persiaAlthough I admit that "compilation" isn't the right word to use for the python package build process.05:51
Hobbseepersia: he left, and it also contains debian/ in the source stuff05:53
persiaHobbsee: All true.  I type too slowly :)05:53
FujitsuHobbsee: Did I see a mention of a non-slow -intel before?06:17
HobbseeFujitsu: yeah.  the latest non-particularly-slow intel in the archives isnt too slow06:18
Hobbseelike, it's not noticibly laggy06:18
Hobbseecrashes too often, though06:18
Hobbseeit's less laggy than when it first came out, though06:18
FujitsuHobbsee: Oh, good. I might upgrade, then.06:18
FujitsuScrolling is just painful.06:18
warp10Hi all!06:19
Hobbseeoh yeah, scrolling still sucks06:19
Hobbseebut the rest of it is fine06:19
FujitsuAh.06:19
=== LaserJock is now known as LaserRock
=== sarah is now known as Hobbsee
mpt__Hobbsee, what did I do?07:31
=== mpt__ is now known as mpt
Hobbseempt: suspect IP address07:31
Hobbseesorry, hostname07:32
Hobbseempt: thought people might have turned from impersonating mark to impersonating you.07:32
mptI'm not that popular, surely07:33
Hobbseewho knows...07:34
StevenKmpt: AboutThisComputer got Approved, but it has the all of the hardware bits back again07:34
mpthttps://wiki.ubuntu.com/AboutThisComputer07:34
mptwhoops07:34
* mpt forgot to press Ctrl+Space first07:35
mptStevenK, there's no wiki page by that name, and <https://blueprints.edge.launchpad.net/ubuntu?searchtext=about> returns zero results07:37
StevenKHum. Where did it go?07:39
poningrumpt: its probably asa ;)07:39
* persia finds https://blueprints.launchpad.net/ubuntu/+spec/about-this-computer07:39
Hobbseewhere did what go?07:39
StevenKAh, sneaky07:40
StevenKhttps://wiki.ubuntu.com/DesktopTeam/Specs/AboutThisComputer07:40
=== stu1 is now known as stub
mptthanks07:41
mpthum, no mockup?07:41
mpt"The icon of the window will be a visually attractive computer icon"07:42
StevenKThe mockup is in text only.07:42
mptThat's going to match what the computer actually looks like about 1% of the time07:42
Rabbitbunnyis there some sort of 'noob guide' for updating the repository? it only has gtkpod 99.2 and liggpod 0.3.2 when 99.10 and 0.6 are current. if I have to make from source I might as well do it for everybody.07:42
StevenK  libgpod |    0.6.0-2 |         hardy | source07:42
Rabbitbunnyfor dapper07:43
mptAt least it will be an improvement07:43
StevenKWe aren't going to update those for Dapper.07:43
persiaRabbitbunny: You could request a backport, but Dapper is considered pretty stable: it's better to upgrade to a newer release.07:43
Rabbitbunnyrequest a backport? i have source..07:44
persiaStevenK: Can all of AboutSystem, AboutUbuntu, and AboutUbuntuRevisited go away now?07:44
RabbitbunnyStevenK: What do you mean 'we'?07:44
persia!backports07:44
ubotuIf new updated Ubuntu packages are built for an application, then they go into Ubuntu Backports. See https://help.ubuntu.com/community/UbuntuBackports - See also !packaging07:44
StevenKpersia: AboutSystem is unrelated. AboutUbuntu and AboutUbuntuRevisited are refered to by AboutThisComputer07:45
StevenKRabbitbunny: We meaning Ubuntu developers07:46
persiaStevenK: AboutSystem looks earlier than AboutUbuntu, and same topic.  I'll deprecate that, but leave the rest.  Thanks.07:46
=== nixternal is now known as _nixternal
poptoneshello, could i get some assistance regarding a memory leak in nautilus?10:37
Hobbseepoptones: better to do so on irc.gnome.org, i expect10:38
Hobbseebesides, its' a weekend.10:38
poptonesthx10:45
=== luka74 is now known as Lure
Riddell~/win 1312:54
Riddellhmm12:54
Hobbseefail :)12:55
Riddellah, Hobbsee, could you give back kdebluetooth and kdesdk13:00
HobbseeRiddell: where?13:00
Riddellhardy13:00
HobbseeRiddell: given back13:09
=== cprov-away is now known as cprov-lunch
=== Shely is now known as MJ086
=== \sh_away is now known as \sh
\shmoins14:31
=== \sh is now known as \sh_away
=== \sh_away is now known as \sh
=== \sh is now known as \sh_away
=== \sh_away is now known as \sh
lamontslangasek: http://conflictchecker.ubuntu.com/possible-conflicts/ lacks hardy...15:17
=== \sh is now known as \sh_away
=== DrKranz is now known as DktrKRanz
=== DktrKRanz is now known as DktrKranz
=== MJ086 is now known as Shely
=== \sh_away is now known as \sh
=== \sh is now known as \sh_away
=== \sh_away is now known as \sh
=== asac_ is now known as asac
=== LaserRock is now known as LaserJock
=== _nixternal is now known as nixternal
sarahlit would appear that the ubuntu log_sql package hasn't been updated in a while.18:14
persiasarahl: Which package?  I don't even see one by that name.18:17
lamontsarahl: package names don't have underscores in them...   what package do you mean when you say 'log_sql'?18:49
\shlamont, i think sarahl means libapache2-mod-log-sql-mysql19:32
=== Kopfgeldjaeger is now known as Sahil_test_von_k
=== Sahil_test_von_k is now known as Kopfgeldjaeger
tunais ther anyone here responsible for http://cdimage.ubuntu.com/custom/ ?20:58
\shtuna, at this time? 21:00 UTC? not likely21:19
slangaseklamont: hmm, alright. who runs conflictchecker.ubuntu.com? :)21:37
lamontslangasek: nfc.  elmo should know??21:41
lamontor any release manager... :-)21:41
lamontslangasek: dunno... '21:44
lamont'twas in your announcement. :-)21:44
slangaseklamont: only indirectly :)22:14
slangaseklifeless: was the conflict checker something you maintained?22:14
\shCOMPLAIN: http://archive.ubuntu.com/ubuntu/pool/main/t/tix8.1/ is totally empty but it's needed for python2.2 in dapper, please readd those missing packages, kthxhappysunday ,)22:54
\shI wonder what happend to tix8.1 that it totally disappeared ... hopefully "you know who" was not the one who cursed it22:57
slangasek\sh: needed how?22:57
slangasekpython2.2 itself doesn't appear to depend on it22:58
\shslangasek, tix8.1-dev is a build-dep for python2.2 which needs a sec fix now...22:58
slangasekah22:58
\shwell then some other build-dep of py2.222:58
\shslangasek, Build-Depends: debhelper (>= 3), autoconf2.13, libreadline5-dev, libncurses5-dev (>= 4.2-3.1), tk8.4-dev, libdb4.3-dev, zlib1g-dev (>= 1:1.1.3), libexpat1-dev, libgmp3-dev(>= 4.1.4-8), libgdbm-dev, blt-dev (>= 2.4z), tix8.1-dev (>= 8.1.3.93), libssl-dev23:00
slangasekwell, drescher only has the warty/hoary/breezy versions available, which were in main; I imagine tix8.1 would need to be reuploaded to dapper?23:00
\shslangasek, looks like...question is what happened?23:01
slangasek\sh: it was probably removed prior to dapper release and no one noticed, since python2.2 wasn't touched at all between breezy and dapper and is in universe?23:02
\shslangasek, dunno..hopefully it was in universe...I'll have to setup a dapper chroot to check the reverse dependencies23:02
slangasekhmm?23:03
slangasekI'm saying that python2.2 is in universe23:03
\shslangasek, oh...sure..pythons2.2. was universe23:03
slangasekwhich by extension implies that tix8.1 was in universe, yes, or else it would've been caught by germinate23:03
\shhopefully23:04
slangaseker, n/m23:04
slangasekI'm thinking backwards23:04
\shwell, we should think about a complete rebuild of the archives, at least a couple of weeks before release23:07
\shwhen I think about the bash <-> dash change it's really a mess to find out, that when you need to fix some vulnerabilities , you have to fix some crappy bashisms too23:07
lamont\sh: we do rebuild the archive (well, main at least), a couple weeks before release.23:12
lamontwe just don't replace the debs, since that would be, um, bad23:12
\shlamont, hmm...na...I just had a problem with perl last week..I attached a patch to fix a cve...and suddenly, I ran into this Makedepends.sh problem...which was fixed in gutsy but not in feisty, but feisty had the bash dash change so it never rebuild after this change23:13
* Fujitsu notes we're trying to collect resources for universe rebuilds.23:13
lamont\sh: see http://people.ubuntu.com/~lamont/buildLogs/Test/byDate/all.html23:13
lamontFujitsu: we could possibly do a rebuild of gutsy/universe now that gutsy is released...23:14
* lamont adds it to his ubuntu-datacenter wishlist23:14
Fujitsulamont: So we have resources to do rebuilds of old releases?23:15
\shlamont, man, those infos should be at a central place, where everyone can see it directly ,-)23:15
lamont\sh: that was the best we could do for the gutsy-autotest rebuild23:15
lamontwhich only did main23:15
lamontFujitsu: the machines that get used to do the autotest (re)builds are actually the security/livecd/etc machines... and there's only one per architecture... hence they don't like to get lots of strain23:16
lamontand the builds are done outside of launchpad23:16
Fujitsulamont: Ah, so it is only that one per arch...23:16
FujitsuI thought there might be some that weren't registered in Soyuz.23:16
\shFujitsu, if we get a sponsor for around 2x2500 euros...we could get 14TB of raid6 diskspace with 2x (2x amd64 dualcore + 2x16GB ram)23:16
lamontFujitsu: soyuz knows all23:16
FujitsuDo we know when Soyuz is getting the rebuild feature? It has been RSN for a while now. Like, a couple of years.23:16
lamontFujitsu: I think it's further down the list than the current cut-line... dunno when it might clear that line23:17
lamontthe biggest issue for LP is that it needs to not keep the new debs23:17
FujitsuI guess that we can't expect too much progress from two whole developers.23:17
lamontand that's kinda contrary to what the librarian thinks it should do23:18
* \sh grabs a beer23:18
Fujitsulamont: I guess it would be hard to model the correct behaviour.23:18
lamontFujitsu: the other challenge for autotest is that it requires an import of the archive into DAK, which isn't exactly the fastest operation in the world.23:19
Fujitsulamont: And that hits various DAK constraints that Soyuz doesn't enforce, right?23:19
lamontdepends on what you mean by constraints23:19
FujitsuSanity checks and the like.23:19
lamontthose pass, actually23:20
FujitsuAh.23:20
lamont\sh: only the red files are failures...23:21
lamontand I believe I filed bugs against all of the failures23:21
lamontsome of the "successes" have implicit pointer conversions in them --> failure23:22
* lamont makes a note to put together the implicit-conversion patch for launchpad's buildds23:22
\shlamont, well, it was just one problem i ran into...and the problem was fixed in gutsy...but I guess, the change inside the chroots from bash to dash came after release of feisty these days23:22
lamontcould be.  OTOH, feisty-security should be building with the same chroot that feisty release did...23:23
* Fujitsu likes the large amount of documentation on how everything is set up.23:23
\shlamont, can be...but pbuilder at home builds with feisty chroot whats provided from the archives :)23:23
\shlamont, but you agree with me, that a missing package which is needed by another package to build is a real problem and needs to be fixed and should be avoided all the time, it doesn't matter if it's main or universe23:24
lamontFujitsu: was that maybe sarcasm?23:30
Fujitsulamont: Unless you can point me to some.23:30
lifelessslangasek: as wiki.ubuntu.com says; mvo or I are contacts for the conflict checker23:30
lifelessslangasek: I wrote it23:31
lifelessslangasek: and have handed it over to mvo23:31
lamontFujitsu: yeah... well, it was all temporary when I was setting up the buildds originally, and so that part never got documented so much.23:32
lamontand yeah, it'd be nice to have it better documented.23:32
Fujitsulamont: I meant about everything, including the Soyuz bits, but that's a good point.23:32
* lamont can't claim any responsibility for the Soyuz bits being documentation-light23:33
* Fujitsu holds lamont responsible anyway.23:34
LaserJockheh23:34
lamontFujitsu: I do admit to being involved in the design23:35
FujitsuThere we go!23:35
* Fujitsu burns.23:35
* lamont completes level 3430 of kobo deluxe.23:38
FujitsuThat's a few levels.23:38
lamontsomehow, I think that's more sad than wonderful.23:38
FujitsuProbably.23:38
lamontthere are only templates for 50 levels... then it repeats, only with more crap23:39
\shwhatever kobo is23:39
Fujitsu\sh: see the kobodeluxe package.23:39
\shhmmm...23:40
\shna...23:40
\shI'*m more in playing anarchy online or eve online23:40
slangaseklifeless: why, so the wiki does say, thanks :)23:42
slangaseklifeless: can you take care of getting hardy on there then, or should I poke mvo?23:42
\shkeescook, bug #163845 ready for review23:45
ubotuLaunchpad bug 163845 in python2.5 "[python] Multiple integer overflow vulnerabilities possibly resulting in the execution of arbitrary code or DoS" [Undecided,In progress] https://launchpad.net/bugs/16384523:45
\shgood night guys23:57
=== \sh is now known as \sh_away

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