[01:10] <Riddell> does blueprint still have the "needs discussion" tickbox?
[01:11] <tfheen> no
[01:13] <cjwatson> Riddell: use the discussion/drafting states to distinguish
[01:13] <Riddell> thanks
[02:03] <Lathiat> pitti: ping?
[02:04] <pitti> Lathiat: pong
[02:08] <Burgundavia> evand: ping
[02:14] <evand> Burgundavia: pong
[02:14] <Burgundavia> pm
[03:00] <lastnode> imbrandon, ping?
[04:54] <_Enchained> hi
[04:54] <_Enchained> anyone can help me in packaging for ubuntu ?
[04:55] <Burgundavia> _Enchained: #ubuntu-motu
[04:55] <_Enchained> ok thx
[04:57] <jdong> is Ubuntu siding with Debian on the whole BitTorrent 4/5 licensing issue?
[04:57] <jdong> or would it be eventually possible for Ubuntu to get a newer version of the mainline bittorrent client?
[04:58] <psusi> what licensing issue?
[04:58] <Burgundavia> jdong: I think so, we don't accept the cddl either
[04:58] <Burgundavia> psusi: choice of venue makes it non-free
[04:59] <psusi> huh?
[04:59] <Burgundavia> read teh debian-legal summary
[04:59] <psusi> where can I find that?
[04:59] <Burgundavia> google it, I don't hvea the url off the top of my head
[05:00] <psusi> google for what?  debian-legal + bittorrent?
[05:00] <Burgundavia> yep
[05:00] <jdong> psusi: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=298814
[05:00] <psusi> ok
[05:00] <Ubugtu> Debian bug 298814 in bittorrent "bittorrent: New upstream version available" [Wishlist,Open]  
[05:00] <jdong> read the first few comments
[05:00] <jdong> and the linked links
[05:01] <jdong> Burgundavia: what about for restricted or multiverse?
[05:01] <jdong> mainline 3.x is starting to get next to useless in comparison to other clients
[05:01] <Burgundavia> why not look at other implementations?
[05:02] <jdong> Burgundavia: I am pretty involved with ktorrent, so I am looking at other implementations
[05:02] <jdong> Burgundavia: but we currently install mainline 3.xx as the default torrent client
[05:02] <Burgundavia> I think there is a C, a C++ and a C# library out there
[05:02] <jdong> and IMO that's not a good choice
[05:02] <jdong> Burgundavia: there is basically no other decent GTK+/GNOME client
[05:03] <jdong> you have BitTornado, which is almost as suckish anything but an ultra ideal swarm
[05:03] <Burgundavia> yep
[05:03] <jdong> and Azureus, which (1) still doesn't work *cough* and (2) eats RAM and CPU for lunch
[05:04] <jdong> and (3) continues to be a bit quirkish due to GCJ instead of Sun JVM
[05:04] <Burgundavia> yes
[05:04] <jdong> Burgundavia: people are starting to run uTorrent inside WINE :D
[05:04] <Burgundavia> plus it is massiively overkill
[05:04] <HrdwrBoB> jdong: it's the best solution
[05:04] <HrdwrBoB> unfortunately 
[05:05] <jdub> we need a nice, simple downloader style client that can tell its own running process to kick off multiple downloads
[05:06] <jdong> HrdwrBoB: ktorrent's almost in a position to overtake that :D
[05:06] <psusi> why does the released edgy versio of grip tell you that it is a development version when you start it up?  did someone slip up?
[05:06] <jdong> HrdwrBoB: we are missing peer exchange support, and that's about it
[05:08] <HrdwrBoB> psusi: it's a universe package
[05:09] <HrdwrBoB> sound juicer is ubuntu default CD ripper
[05:09] <jdong> psusi: the version in Edgy was indeed a development version of grip?
[05:09] <psusi> that is what it claims
[05:10] <psusi> get a big fat popup the first time you start it saying this is a development version, if you have truoble, revert to stable
[05:10] <jdong> psusi: well, it's true... :)
[05:10] <jdong> it is a development version of grip
[05:11] <psusi> shouldn't we not be releaseing development versions in the stable release of ubuntu? ;)
[05:12] <Hobbsee> psusi: not necessarily
[05:12] <Hobbsee> psusi: seems it was a sync from debian anyway, with a couple of rebuilds
[05:13] <jdong> psusi: the MOTU team must've thought it was a better choice than the stable version of grip
[05:15] <Hobbsee> mpt: indeed
[05:15] <psusi> ok.. but shouldn't they have disabled the obnoxious warning at least? ;)
[05:16] <jdong> psusi: perhaps, yes
[05:16] <jdong> psusi: guess nobody bothered to complain before edgy's release
[05:18] <Burgundavia> psusi: if you care about an app, watch it through the dev cycle and make certain ti gets to where you need it to be
[05:18] <psusi> didn't realize it was universe... just installed it from the add programs applet
[05:18] <psusi> because sound juicder can't correctly rip this cd I got today
[05:26] <psusi> hrm... is this with just the main bittorrent client, or derived ones as well?   like bit tornado?
[06:23] <zOrK> where can I begin developing? is there some webpage?
[06:24] <crimsun> https://wiki.ubuntu.com/UbuntuDevelopers
[07:03] <cge> What is the process for requesting that something be removed from main?
[07:03] <crimsun> removed from Ubuntu or demoted to universe?
[07:03] <cge> reportbug - demoted to universe, probably. 
[07:04] <crimsun> any way it can be fixed?
[07:04] <cge> The only people I've spoken to about it seemed to think that it would be easier to rewrite it.
[07:06] <cge> It currently sends bug reports to ubuntu-users, since it is based on Debian's reporting system.
[07:07] <cge> So some users, who see it in main and assume that it is supported by Ubuntu, install it and use it, resulting in them being admonished for spamming ubuntu-users.
[08:40] <asabil> hi all
[08:41] <asabil> https://features.launchpad.net/distros/ubuntu/+spec/usb-adsl-modems
[08:41] <asabil> don't you think that this is important enough to require a higher priority ?
[08:41] <LaserJock> it was discussed today
[08:42] <asabil> and ?
[08:42] <LaserJock> it has been marked as read for review
[08:43] <asabil> hmm okey
[08:43] <LaserJock> *ready
[08:44] <asabil> because this is one reason why people in my country don't use ubuntu :/
[08:44] <asabil> given that the eagle-usb drivers are broken :D and ueagle-atm is not in dapper nor edgy
[08:45] <LaserJock> yeah
[08:45] <LaserJock> so it was proposed, discussed, drafted, and up for review
[08:45] <LaserJock> that's pretty good for the first day of the summit :-)
[08:45] <asabil> let me just hope that it will get accepted
[08:46] <asabil> cool then ;)
[09:39] <sivang> morning
[10:48] <PecisDarbs> hi devs, question to Ubuntu packagers - how frequently happens translations updates packages release and how to initialise one, if there is need to fix faulty translations. For example, there is lot of buggy translations in particular lanugage and it would be nice to issue new translation update. Is that possible?
[10:49] <pygi> PecisDarbs: there will be translations build somewhere in the future I think
[10:49] <pygi> PecisDarbs: pitti is your man
[10:50] <PecisDarbs> thanks for info :)
[10:50] <PecisDarbs> pygi: he is frequent guest here?
[10:50] <PecisDarbs> :)
[10:51] <pygi> PecisDarbs: I doubt you'll find most of folks here right now, since they are at conference
[10:51] <crimsun> it's 1:31 AM where they're at.
[10:52] <PecisDarbs> ahhh
[10:52] <PecisDarbs> I don't need it ASAP, just a little info :)
[10:52] <PecisDarbs> it/him/s
[10:52] <pygi> PecisDarbs: I think every first monday in month or something like that is update
[10:53] <PecisDarbs> I see
[10:53] <PecisDarbs> good
[10:53] <PecisDarbs> thanks for info :)
[10:53] <pygi> yw ^_^
[12:39] <Administrator__> I want to build my own recovery dvd. If I could get the source of the ubuntu installer cd, or if there was a way of adjusting the init scripts that would be great. The init script is going to do the following, restore partition table, mbr, mkfs, mount, restore biggbackup tarball, grub install. reboot. If there is a way to use the current ubuntu cd that would be great.
[12:40] <Hobbsee> Administrator__: the people who know are mostly in mountain view at the moment, and still asleep
[12:44] <rigidus> hi all
[12:45] <rigidus> could anybody help me guys, how to give -i option to postgresql's postmaster process? Thanks :-)
[12:46] <Hobbsee> rigidus: try #ubuntu for support
[12:46] <rigidus> thanks I tried
[12:52] <cge> rigidus: Try on the mailing list, or the forums, if you don't get an answer in #ubuntu.
[12:52] <rigidus> I tried the forum but there was no answer
[12:53] <rigidus> I guess the startup script is not completed
[12:53] <cge> unfortunately, #ubuntu and the forums are declining in usefulness for people with more technical or nontrivial problems because of the huge number of users that flood them with obvious questions. The mailing list is probably best. But please don't bring them here.
[12:53] <rigidus> ok, thanks
[12:55] <Hobbsee> rigidus: likely the people who know are mostly in mountain view at the moment, and still asleep - try ubuntu-users mailing list for support
[12:57] <rigidus> thanks Hobbsee
[01:39] <JohnFlux_> Hey all
[01:39] <Hobbsee> heya
[01:40] <JohnFlux_> is there a channel for the summit
[01:40] <Hobbsee> JohnFlux_: yeah, #udsmtv
[01:40] <Hobbsee> JohnFlux_: but it's still very early morning in MV
[01:40] <Hobbsee> hence you probably wont find people awake
[01:48] <JohnFlux_> Hobbsee: i know
[01:48] <JohnFlux_> Hobbsee: i cant sleep ;)
[01:49] <Hobbsee> JohnFlux_: heh.  jetlag?
[01:49] <JohnFlux_> yeah heh
[01:49] <tepsipakki> is security-repos from archive.u.c now obsolete?
[01:49] <tepsipakki> are ^
[01:49] <JohnFlux_> Hobbsee: i wonder too heh.  im a kde coder
[01:49] <Hobbsee> JohnFlux_: ahh :)
[01:50] <Hobbsee> JohnFlux_: yet you're in everywhere but #kubuntu-devel it seems.  (which is also silent, due to everyone being asleep)
[01:50] <JohnFlux_> Hobbsee: its really good though - im having fun
[01:50] <Hobbsee> what are you coding?
[01:50] <Hobbsee> :)
[01:50] <JohnFlux_> ah there we go
[01:50] <JohnFlux_> Hobbsee: im the maintainer of ksysguard
[01:51] <JohnFlux_> and one of the coders of konversation
[01:51] <JohnFlux_> and miscillanous kde stuff
[01:52] <Hobbsee> ahhh....so that's where i recognise the name from
[01:52] <JohnFlux_> ;)
[03:52] <bddebian> Heya
[04:21] <LTjake> hello. I was wondering, as a core developer for Catalyst (perl MVC framework) and an ubuntu user (server and desktop), I'd like to see these packages as up-to-date as possible; is there any way i can help?
[04:35] <sivang> LTjake: first check if they are in universe,
[04:35] <sivang> LTjake: if they are, then you can work on them and upload your packages to REVU
[04:36] <sivang> LTjake: after prooving your skills and after working with folks in the MOTU team, you can get approved for uploads and then you could take care of them yourself
[04:36] <LTjake> okay. checking...
[04:38] <LTjake> do things get sync'ed from debian packages at all? the latest catalyst seems to be in debian unstable...
[04:38] <LTjake> the packages in universe are out-of-date, btw.
[04:40] <grndslm> this might not be the place to ask, but i have been asking in the lvm && ubuntu channels with no help....
[04:40] <grndslm> why on earth would pvscan list my device as an unknown device??
[04:43] <Chipzz> grndslm: not getting an answer elsewhere is not an excuse to ask here ;)
[04:43] <Chipzz> but this may be a valid question :P
[04:43] <Chipzz> just pointing out in general
[04:50] <grndslm> Chipzz, somebody in the lvm channel tells me that lvm version 2.02.06 is buggy...but this is obviously version *-2ubuntu3
[04:50] <grndslm> so i dunno if this would be the same bug(s) or what
[04:51] <Chipzz> apt-get source and look at what's patched
[04:52] <_ion> Wow, looks like you get to write twice the amount of code in order to achieve the same thing with Catalyst, compared to Rails.
[04:56] <pygi> hey sabdfl 
[04:57] <sladen> iwj: what exactly is /var/lib/dpkg/updates/ for?  The Changelog says it was added in 0.93.18 (12 years ago), but I can't find a reference to it actually being used 
[04:59] <sladen> iwj: for stackable file-systems, dpkg/available needs updating somehow
[05:11] <ianm_> has there been any discussion regarding prevention of swap-deathin ubuntu?
[05:15] <ianm_> it seems that any app can kill or at least DoS a linux desktop just by allocating lots of memory
[05:27] <psusi> don't run such apps.... and if you run a multiuser server, then there is ulimit
[05:29] <simira> heh... problems in google hq?
[05:30] <ivoks> ?
[05:31] <ianm_> psusi: do you think normal users don't deserve such protection?  how do you know ahead of time which programs are bad (and in which situations)?
[05:31] <psusi> no... normal users are going to be the only person on the system and they aren't going to intentionally write and run a fork bomb
[05:32] <ianm_> I'm not talking about fork bombs, just memory exhaustion
[05:32] <psusi> same thing
[05:33] <ianm_> normal users aren't going to run apps that ask for lots of memory? 
[05:33] <psusi> intentionally hogging all resources to cripple the system.... it's not a concern to single user systems
[05:33] <psusi> they aren't goign to run apps that intentionally ask for more and more memory until it is all used up
[05:33] <ianm_> again not talking about malicious behavior
[05:34] <psusi> if they try to open a 3 gig image in gimp with only 128 megs of ram, then they can put up with the lag
[05:35] <ianm_> it's not lag though, it's endless HD churning until they hard-reboot.  maybe not technically a crash, but effectively it is
[05:35] <ianm_> memory exhaustion is the only problem I've seen, and I see it several times from normal programs under normal conditions.  sometimes firefox, other times evolution
[05:36] <psusi> hard reboot?
[05:36] <psusi> sounds like a hardware bug
[05:37] <psusi> it should just get real swappy until memory demand goes down
[05:37] <ianm_> what if it doesn't go down?
[05:37] <psusi> then the system stays swappy
[05:37] <psusi> until you get fed up with it and start killing stuff ;)
[05:38] <grndslm> how can i find out if a certain patch was applied in an ubuntu release version, specifically lvm?
[05:38] <ianm_> how do you kill stuff when the GUI is unresponsive?  (yeah, terminal, kill -9, but this is linux for normal people right?)
[05:38] <grndslm> killall <packagename>
[05:38] <grndslm> that works well enough for me
[05:38] <ianm_> psusi: or did you mean, like, your cat and your girlfriend? :)
[05:39] <psusi> lol
[05:39] <psusi> yea, it would be nice if gnome-system-monitor locked itself into ram
[05:39] <psusi> so you could stil bring that up and kill stuff
[05:39] <Keybuk> ianm_: click the [X]  button in the corner
[05:40] <Keybuk> metacity will notice that the app is unresponsive to the request, and offer to force quit it (ie send KILL)
[05:40] <grndslm> what's wrong with the kill applet?
[05:40] <ianm_> Keybuk: how do they even know which program is misbehaving?
[05:40] <Keybuk> ianm_: the one that's unresponsive
[05:40] <grndslm> or "force quit" actually
[05:40] <ianm_> all are unresponsive when your system is swap-dying
[05:40] <ianm_> using 'top' is the only way I've figured out which is to blame
[05:40] <Keybuk> yes, and when your system is doing that, killing things won't help
[05:41] <Keybuk> System -> Administration -> System Monitor
[05:41] <Keybuk> the one at the top is using all the CPU
[05:41] <Keybuk> click "End Process"
[05:41] <ianm_> but you understand that navigating menus / GUIs during a swap-death is next to impossible..?
[05:41] <Keybuk> ianm_: so?
[05:41] <Keybuk> what do you want?
[05:42] <psusi> when thrashing no task is using much cpu
[05:42] <Keybuk> novice users can't deal with non-gui
[05:42] <psusi> you just have one that is using all the ram... that's what needs to die
[05:42] <grndslm> is there such a thing as too much swap?
[05:42] <psusi> yes
[05:42] <ianm_> Keybuk: right, so I'm proposing that we don't let a process DO that to the system
[05:42] <Keybuk> ianm_: how do you prevent a process doing doing that?
[05:42] <psusi> how do you propose that?
[05:42] <Keybuk> psusi: then that's the one the kernel kills
[05:43] <ianm_> I can't say for sure, I'm not a kernel hacker, but maybe just preventing one process from allocating more than X ram?
[05:43] <Keybuk> ianm_: then you're talking about this in the wrong channel
[05:43] <Keybuk> this is the channel for people with proposed solutions
[05:44] <grndslm> anyway...could you guys tell me how i could find out if a certain patch was applied in an ubuntu release version, specifically lvm?
[05:44] <grndslm> after apt-get sourcing?
[05:44] <psusi> can someone teach ubotu to quit telling users to use the disks applet to mount partitions since it has been removed in edgy?
[05:44] <ianm_> Keybuk: heh and must someone be 100% sure of their proposal..?
[05:45] <ianm_> Keybuk: "preventing one process from allocating more than X ram"  <--- proposed solution
[05:45] <psusi> ianm_: there is no good way to choose a value of X that will not annoy people who actually want to load a really big app
[05:45] <psusi> ianm_: and you can just have n tasks that use X and still thrash the system
[05:46] <ianm_> how about locking critical stuff in RAM, so that the system stays responsive?
[05:46] <tfheen> ianm_: sure, set ulimits sensibly.
[05:46] <Keybuk> ianm_: what's critical?
[05:46] <Keybuk> the kernel already does that
[05:46] <ianm_> Keybuk: the things that keep the GUI responsive
[05:46] <psusi> have to be careful about that because then it lowers the ram availible to other things when you don't need the "critical stuff"
[05:46] <Keybuk> the primary thing that keeps the gui responsive is also the primary allocator of RAM
[05:48] <sivang> Keybuk: just wanted to ask, I poked at eject for making sure ubuntu changes were not dropped in the merge, but couldn't find the base packag. Do you have an idea where it went ? (I couldn't find it on snaphost/qa.d.o)
[05:49] <Keybuk> sivang: merges.ubuntu.com
[05:50] <ianm_> psusi: I agree re: careful, but aren't there core gnome things that have to stay in RAM for the GUI to be used?
[05:50] <Keybuk> ianm_: no
[05:50] <slomo> Keybuk: hi :) if you have some free time could you take a look at bug #69647? the sync is needed for some other stuff and the only ubuntu change is trivial
[05:50] <Ubugtu> Malone bug 69647 in cli-common "Please sync cli-common 0.4.6 from debian/unstable (main)" [Undecided,Confirmed]  http://launchpad.net/bugs/69647
[05:50] <Keybuk> slomo: no.
[05:50] <Keybuk> syncs are in auto again now
[05:50] <psusi> ianm_: there is a LOT of stuff that would need to be locked... which is the problem
[05:50] <Keybuk> they will take 3-4 days
[05:50] <ianm_> Keybuk: like... the gtk library?  glib?  no??
[05:50] <Keybuk> ianm_: they aren't processes
[05:51] <sivang> Keybuk: you store the base source somewhere on m.u.c ?
[05:51] <Keybuk> sivang: naturally
[05:51] <ianm_> Keybuk: didn't say processes, said "core gnome things"
[05:51] <psusi> Keybuk: yea, but they still could be locked in ram if you wanted to
[05:51] <Keybuk> psusi: it wouldn't help
[05:52] <psusi> Keybuk: well, it would... but it would be costly
[05:52] <sivang> Keybuk: okay, I was just puzzled by not being able to find the base anywhere else then :)
[05:52] <Keybuk> psusi: no, it wouldn't
[05:52] <Keybuk> sivang: please don't try and do merges by hand, use mom
[05:52] <psusi> how would it not help to not have to swap them back in?
[05:52] <ianm_> psusi: costly only in the sense that no process could take up 100% of RAM and swap? 
[05:52] <slomo> Keybuk: this one involves dropping ubuntu changes though
[05:52] <psusi> costly in the sense that there would be a lot less ram that could be used by other things that might need it more than the gui at times
[05:52] <Keybuk> slomo: then it'll be done in roughly two weeks
[05:53] <slomo> ok
[05:54] <psusi> now what might be really cool is if there were a magic sysreq key or something you could hit that would stop the kernel from swapping out from a list of important tasks for a while... so you could hit that to regain control of a thrashing gui
[05:54] <Keybuk> psusi: there is
[05:54] <psusi> then switch it back after
[05:54] <psusi> there is?
[05:55] <Keybuk> well, there's a key to stop swapping, and others to kill processes
[05:55] <Keybuk> if swapping is actually causing a system problem, you've got bigger problems than just the wrong things in the wrong place
[05:55] <Keybuk> you really need to kill things
[05:56] <ianm_> Keybuk: are we back to that?  how does a normal user kill processes?
[05:57] <psusi> yea, but you don't just want to have the kernel pick something to kill, and you may in fact, want to let the process keep running.... just not at the cost of gui responsiveness
[05:57] <sivang> Keybuk: okay, I'll try to get less confused by it :)
[05:59] <psusi> and when you say there is a key to stop swapping, what exactly does that mean?  just swapoff -a?
[05:59] <psusi> cause that's not what you want either
[05:59] <ianm_> if we assume that 1) a normal user will never use the command line, 2) the GUI gets unresponsive in these conditions, 3) when the system is unresponsive, a normal users assumes it needs a reboot, then aren't we *effectively* back in windows 3.1 days where any app can hard-lock the system? (not literally/technically, but effectively!)
[06:00] <Keybuk> ianm_: then propose a solution
[06:00] <Keybuk> this is NOT a bitching channel, or a bug reporting channel, or even a support channel
[06:00] <Keybuk> this is "I have a concrete idea to fix this problem, here's my idea"
[06:00] <ianm_> I came here to pick your brains on this topic
[06:00] <psusi> practically though the only time I can see this happening is if someone opens a 3 gig image in gimp... and they can just click the close button and it will be killed
[06:00] <Keybuk> you haven't given a single idea yet
[06:00] <Keybuk> ianm_: this is not a brain picking channel either
[06:00] <Keybuk> start a thread on the ubuntu-users mailing list
[06:00] <Keybuk> when you have a concrete idea, start a thread on ubuntu-devel proposing it
[06:01] <tfheen> Keybuk: we could conceivably make c-a-d bring up the task manager or something like that.  Or some thingabob similar to the windows c-a-d dialog
[06:03] <ianm_> tfheen: would whatever components the task manager needs have to be locked in memory for that to work?
[06:04] <ianm_> Keybuk: if you haven't seen a single idea yet, read more carefully
[06:04] <tfheen> ianm_: you seem to be very attached to the "lock in memory" idea which has bad implications for other reasons.
[06:05] <sladen> when swapping gets bad the system responsiveness gets aweful.  Best thing to do is Ctrl-Alt-F1 and come back 20minutes later
[06:05] <ianm_> I'm not attached to any idea, just asking
[06:05] <sladen> at which point the system will be normal again
[06:05] <mjg59> Is mlocked memory still shared?
[06:05] <tfheen> mjg59: unsure, but mlock requires root privs, iirc?
[06:05] <cjwatson_> not any more
[06:06] <tfheen> oh, ok
[06:06] <cjwatson_> that was changed in the Ubuntu kernel (at least) a while back
[06:06] <mjg59> There's a configurable amount that can be done per user
[06:06] <psusi> I like the idea of rather than locking the task manager all the time, setting it up so that ctrl-alt-del will bring it up, and set the task such that the kernel will not prune pages from its working set for a while
[06:06] <mjg59> Right, what you /actually/ want is for the kernel to just stop caring about anything other than that application
[06:06] <psusi> that way it will get itself faulted in and running in a timely manner and then be fully responsive
[06:06] <mjg59> Oh. Other than X, of course.
[06:06] <mjg59> Which makes things trickier
[06:07] <sladen> yup/win 346
[06:07] <tfheen> mjg59: so really just renice a bunch of processes -5 or so?
[06:07] <mjg59> Since by X we actually mean "the X server that this application is going to connect to"
[06:07] <mjg59> tfheen: I'm not sure how well the linux scheduler works when stuff is deep in swap
[06:07] <psusi> true... X needs to not be thrashed as well
[06:07] <tfheen> probably not great.
[06:07] <mjg59> Which is probably what the actual problem is
[06:08] <psusi> nice itself won't help since it isn't an issue of cpu scheduling
[06:08] <iwj> sladen: /var/lib/dpkg/updates is so that dpkg doesn't have to rewrite /var/lib/dpkg/status each time.
[06:08] <mjg59> psusi: If the thrashing tasks don't get scheduled, they won't trigger page faults
[06:08] <psusi> unless the swapper respecs nice?  which I don't think it does
[06:08] <ianm_> is there a time when a user *wants* X to thrash?  
[06:08] <psusi> mjg59: they do get scheduled... then page fault and block again... that's the problem
[06:08] <mjg59> psusi: Right. So to some extent it /is/ a cpu scheduling thing
[06:09] <psusi> mjg59: no... it isn't... it got scheduled... but can not run because it page faulted
[06:09] <mjg59> psusi: In the specific case of us wanting the system to remain responsive enough for people to kill tasks, we don't want the applications that are causing the memory starvation to get scheduled at all
[06:10] <Keybuk> mjg59: and probably dbus and HAL for the task manager to communicate with the other random bits
[06:10] <Keybuk> and the window manager
[06:10] <Keybuk> so it can get painted
[06:10] <psusi> ohh, yes... if you stop the bloated tasks then they won't be faulting in pages anymore
[06:10] <Keybuk> oh, and the composite manager, so it can be drawn
[06:10] <mjg59> But to be honest, my gut feeling is that this is something that would make a nice research paper for someone
[06:10] <psusi> but nice isn't going to do that for you
[06:10] <mjg59> Right, it would need to be rather more invasive
[06:11] <psusi> aye
[06:13] <psusi> I have a question about Suggests: and Recommends:  do these fields exist for any reason other than to inform the inquisitive power user about other things they might want to install?
[06:13] <mjg59> Yes
[06:13] <psusi> if they are supposed to list packages that are not strictly depended on, but things that you really should install to use this package properly, shouldn't the add applications applet and maybe even synaptic at least prompt the user to install them?
[06:14] <mjg59> psusi: http://www.debian.org/doc/debian-policy/ch-relationships.html
[06:16] <psusi> mjg59: that doesn't say what tools actually DO in response to the field... and from my experience, it looks like they do nothing
[06:17] <psusi> i.e. apt-get and synaptic do not prompt you to install the Recommended or Suggested packages
[06:17] <mjg59> psusi: What the tools do is entirely up to the individual tools
[06:18] <psusi> from a user perspective, shouldn't the tool at least prompt to install the Recommended packages?  it seems none of them do
[06:18] <psusi> should I file a feature request for the add applications applet to do this?
[06:19] <tfheen> we need to vet the list of recommends first, iirc
[06:19] <ianm_> psusi: also it seems it would have to explain the choice, which it isn't prepared to do
[06:19] <psusi> tfheen: vet it?
[06:20] <ianm_> psusi: saying "xftgbd is recommended, do you want to install it?" probably isn't appropriate
[06:20] <tfheen> make sure it's sane
[06:20] <psusi> ianm_: I think it would be sufficient to put up a dialog that says "It is strongly recommended that you also install the following to get the full use of out this package:" and have the other packages checked to install unless they uncheck them then hit ok
[06:21] <ianm_> I agree that'd be fine for synaptic, as that's not what normal users will be using anyway
[06:22] <psusi> followed by "The following packages are Suggested to be installed as they may add in some way to the functionality of this package, install?" and list them but have them not to be installed unless the user checks them off before hitting ok... maybe even make the package names clickable so they can read the description
[06:23] <psusi> I raise this issue because I used the simple add applications applet to install Grip to rip some cds, and it does not work out of the box because it needs vorbis-tools to encode... it is listed under Recommends: but since that field is ignored,
[06:23] <psusi> we have a package that does not work out of the box.
[06:23] <psusi> this leads to poor user experience
[06:23] <ianm_> that sounds more like a data bug, no?
[06:24] <ianm_> as in vorbis-tools should be required
[06:24] <psusi> no, because it is not required
[06:24] <psusi> you can choose not to encode to ogg
[06:24] <psusi> you can just keep it in wav, or go with mp3
[06:24] <psusi> but for it to just work out of the box... it needs vorbis, which is why it is strongly recommended
[06:25] <psusi> unfortunately, the installer ignores this fact
[06:25] <ianm_> oh I see.  the app could possibly be wise to the situation as well
[06:26] <psusi> yea, I was thinking maybe the debconf script for the package could check to see if any of the encoder packages are installed, and prompt the user to install them?  not sure if you can do that with a debconf script, I'll have to hack around with it
[06:26] <ianm_> there has been talk about media apps making it easy to install codecs.  this seems to fall into that category
[06:31] <pygi> dholbach: thanks, was already approved as part of -qa team :)
[06:32] <zOrK> hello
[06:37] <cbx33> hey jono 
[06:37] <jono> hey cbx33
[06:37] <cbx33> guys, who would I talk to about using the arrows for the undo/redo in the Human theme in a logo for a new ubuntu product
[06:37] <jono> how did you get on with gstreamer?
[06:37] <cbx33> jono I jhave a working prototype
[06:37] <pygi> jono: as always, are you willing to be boethered? :)
[06:37] <cbx33> ;)
[06:37] <pygi> bothered even :)
[06:38] <cbx33> bothered....? with me?
[06:38] <jono> cbx33: awesome :)
[06:38] <pygi> cbx33: no, I wanna bother him :P
[06:38] <cbx33> oh heh
[06:38] <jono> pygi: huh?
[06:38] <cbx33> jono I can now convert ogg mp3 and wav to wav ;)
[06:38] <jono> cbx33: woo!
[06:38] <cbx33> i was hoping for amr
[06:38] <cbx33> but it's not in the repo
[06:38] <pygi> jono: do you have the power to setup a mailing lists at lists.ubuntu.com?
[06:38] <jono> cbx33: I saw some of the gstreamer trick modes yesterday - funky shit
[06:39] <jono> pygi: nope
[06:39] <pygi> jono: you know who does? :)
[06:39] <jono> pygi: yep, lemme check
[06:40] <jono> pygi: mail mailman@lists.ubuntu.com
[06:40] <cbx33> trick modes?
[06:40] <pygi> jono: ok, I'll write there
[06:40] <pygi> jono: thanks
[06:40] <jono> pygi: :)
[06:47] <cbx33> jono tell me more about these trick modes
[06:50] <jono> cbx33: you can stuff like slowing and speeding up audio in realtime, its pretty cool
[06:50] <jono> doing reverse playback and stuff
[06:50] <cbx33> ooooh
[06:52] <sladen> jono: how the heck to do you reverse playback realtime?
[06:52] <Treenaks> sladen: gstreamer ;)
[06:53] <cbx33> heheh
[06:53] <sladen> oooh, gstreamer has gained so much.  I'd missed the Time Machine going in
[06:54] <cbx33> sladen, hiya...had chance to sign my key yet? :p
[06:54] <sladen> cbx33: you control panel got a mention by ogra in the introductary talk
[06:54] <cbx33> oh really?
[06:54] <cbx33> good or bad?
[06:54] <cbx33> :p
[06:54] <sladen> cbx33: sounded good to me
[06:55] <cbx33> w00t
[07:24] <zOrK> Is anyone working on easy-ndsiwrapper?
[07:24] <zOrK> https://features.launchpad.net/distros/ubuntu/+spec/easy-ndiswrapper
[07:58] <DShepherd> I don't know if anyone else noticed but the firefox icon in the main menu differs from the one that shows on the application itself... was this intended? 
[07:58] <robertj> already filed
[07:59] <DShepherd> robertj:ok 
[07:59] <jdong> DShepherd: hehe, the DFSG spirit always finds ways to cling onto our packages :D
[08:00] <DShepherd> jdong: ;-)
[08:25] <sladen> jono: re: Rosetta, stub, carlos and danilo are all marked down to be in the session running concurrently
[08:28] <jono> sladen: ahhh cool
[08:44] <zOrK> is there any patch for the broadcom wireless card?, I am thinking about make one
[09:20] <volvoguy> hey folks. rebuke me if you must, but i can't find any documentation anywhere that mentions whether the server kernel is SMP capable. can i get a quick yes or no?
[09:21] <infinity> volvoguy: Yes.
[09:21] <volvoguy> infinity, thank you. :) bye bye now. 
[09:26] <geser> infinity: hello. are you still working on bug 65266 ?
[09:26] <Ubugtu> Malone bug 65266 in php4 "[UVF Exception]  Sync php4 4.4.4 from Debian unstable" [Medium,Confirmed]  http://launchpad.net/bugs/65266
[09:33] <claviola> where can I find this year's signing keys for the archives?
[11:12] <tepsipakki> Seveaz: hi, do you know of any efforts on packaging nx-2.1.0?
[11:13] <Seveaz> tepsipakki, no, focus in the free software world is on the 2x.com packages and not nomachines mess
[11:13] <tepsipakki> oh
[11:14] <tepsipakki> didn't know such exists
[11:14] <ajmitch> Seveaz: where are these packaging efforts?
[11:14] <ajmitch> I haven't seen 2x.com as much of an improvement on the nx mess
[11:14] <Seveaz> kanotix
[11:14] <tepsipakki> in the meantime, I have the 2.1.0 mess compiling atm :)
[11:14] <Seveaz> it's an improvement, all parts are free
[11:14] <tepsipakki> thats good
[11:15] <Seveaz> it's still messy though
[11:17] <tepsipakki> I took the nx-1.4..+1.5.0-11ubuntu1 and made some changes to the patches so at least it is compiling right now.. we'll see if it get's through
[11:20] <tfheen> have they stopped shipping a complete copy of an X source tree yet?
[11:21] <tepsipakki> nope :)
[11:21] <tfheen> won't make the archive, then
[11:22] <pygi> tfheen: why would someone do that? :-/
[11:22] <tepsipakki> oh that's the reason.. wondered why there were no packages
[11:22] <tfheen> pygi: because somebody had a very bad idea one morning, or something?
[11:23] <pygi> tfheen: right :-/
[11:32] <doko> gicmo: slacker!
[11:32] <gicmo> hey!
[11:34] <sivang> hey gicmo 
[11:34] <dholbach> doko, gicmo: having fun? :)
[11:35] <gicmo> dholbach: sure enough!
[11:35] <doko> dholbach: stay at your session, no more place on the couch ...
[11:35] <seb128> dholbach: aren't you suposed to draft telepathy? :p
[11:35] <bhale> seb128: will we get galago in main?
[11:35] <dholbach> doko, gicmo: don't worry - I'll leave you guys the couch - enjoy your time there together ;-)
[11:36] <bhale> seb128: with telepathy?
[11:36] <pygi> seb128: we did that already I think? :P
[11:36] <dholbach> seb128: i'm working on it
[11:36] <seb128> bhale: we didn't really speak about galago 
[11:36] <seb128> dholbach: suuuuuure :)
[11:36] <gicmo> dholbach: yeah, the couch! ;-D
[11:36] <seb128> couch rulez!
[11:36] <seb128> heh
[11:36] <bhale> vibe the fuck out?
[11:37] <seb128> we are working like house elves there!
[11:37] <tfheen> cuddly house elves you are
[11:51] <sivang> couch ?