[12:01] <mjc> I'm playing with serel also
[12:04] <|QuaD-> tseng: cool :)
[12:05] <tseng> DV commited a inotify patch, its supposed to pass the regression tests now
[12:06] <|QuaD-> dv?
[12:13] <bluefoxicy> fabbione:  need any info on bug #9723?
[12:14] <bluefoxicy> It's gonna be a bitch of a problem, because I've had it reported to me by a few users who have gentoo or ubuntu on the same exact model that the stuff that's broke for me actually works
[12:14] <bluefoxicy> :(
[12:16] <fabbione> bluefoxicy: probably.. i didn't look at it yet
[12:17] <bluefoxicy> ah, 'cause it's assigned to you :)
[12:17] <bluefoxicy> heh
[12:17] <bluefoxicy> I should get best buy to support linux
[12:18] <bluefoxicy> "Log onto STAR and see what the problem was"  ". . . customer could not operate USB camera."  "Resolution?"  "No drivers for camera, data gathered, dmesg logs shipped to bugzilla of distro and to LKML"
[12:19] <mjc> anyone have a serel package?
[12:21] <fabbione> bluefoxicy: yes... it's assigned to me, but i have been awake for 20 hours and i am too tired to think straight right now
[12:21] <bluefoxicy> alright, tomorrow then.
[12:21] <fabbione> good night
[12:21] <ogra> night fabbione 
[12:27] <mvo> elmo: can you please sync findutils?
[12:27] <mjc> tseng, pkg-mono thinks I'm building 1.0.5 but I'm building 1.1.6
[12:32] <zyga> re
[12:39] <CarlK> should I bugzilla this?  nfs warning: mount version older than kernel
[12:40] <zyga> CarlK: what are you mounting? I've seen the same message flood my logs
[12:41] <mjc> edited the changelog and removed the patches in the patch dir
[12:41] <mjc> seems to be running currently
[12:42] <CarlK> 3 day old ubuntu box is the client, 4 year old 2.4.9-ac7  NFS server
[12:43] <CarlK> is the current ubuntu still called hoary, or did that go away when it went... um... stable?
[12:43] <zyga> CarlK: similar setup, I've got 2.4.26 slack
[12:43] <ogra> CarlK, the name stays...even if its dead in 18 months ;)
[12:44] <CarlK> groovy
[12:44] <ogra> like wartys name does...
[12:44] <CarlK> oh yeah...
[12:46] <CarlK> ok, hoary boxes on both sides, still get "nfs warning: mount version older than kernel"
[12:47] <CarlK> hmm, one is still from beta
[12:58] <ogra> hmm, amu ? 
[01:00] <trulux> any python hacker around?
[01:01] <ajmitch_> trulux: what's needed?
[01:01] <trulux> ajmitch: you're going to reveal my surprises for Breezy.... a port of redhat-config-* tools to Ubuntu
[01:01] <trulux> of course not those who do ame job as current ones in the gst packages
[01:02] <ajmitch_> trulux: well you're the one who revealed it ;)
[01:03] <trulux> ajmitch: I just mean that you were making me *revealing* it, and no, this is not intended to be sent to bash.org
[01:03] <|QuaD-> i think i am missing something in this conversation, i am not understanding anything
[01:03] <ajmitch_> |QuaD-: don't worry
[01:06] <trulux> ajmitch: yeah, it's not my best day
[01:06] <trulux> just I need to sleep
[01:08] <trulux> CarlK: haha
[01:10] <trulux>   File "/home/lorenzo/proyectos/ubuntu/redhat-logviewer/src/LogBuffer.py", line 117, in insert_into_buffer_at_offset
[01:10] <trulux>     theiconv = iconv_codec.open(self.codeset, 'iso-8859-1')
[01:10] <trulux> AttributeError: 'module' object has no attribute 'open'
[01:14] <ajmitch_> trulux: want me to take a look? :)
[01:14] <trulux> ajmitch: only if you don't have other things to do
[01:15] <trulux> http://fedora.redhat.com/projects/config-tools/redhat-logviewer.html
[01:28] <doko> good morning all
[01:29] <Mithrandir> hi doko
[01:29] <Mithrandir> you're in .au now?
[01:30] <doko> yes, arrived two hours ago
[01:30] <ajmitch_> morning doko
[01:32] <ajmitch_> trulux: seems to have changed in the cvs version
[01:32] <doko> pitti, mvo, seb128: still around?
[01:44] <mjc> hmm, how do I go about turning off gpg checking for a certain repo? the meeby mono repo has no gpg, I don't think
[01:45] <schweeb> mjc: check the apt manpage/documentation
[01:47] <mjc> haven't found anything
[01:48] <mjc> ah here
[01:48] <mjc> /usr/share/doc/apt/examples/configure-index.gz
[01:51] <mjc> hmm still nothing to turn off stuff
[01:56] <ogra> doko, they're all packing their bags ;)
[01:56] <ajmitch> ogra: when do you fly out?
[01:57] <ogra> tomorrow night
[01:57] <ogra> 5 to 12 :)
[01:57] <ajmitch> fun :)
[01:57] <ogra> yeah
[01:57] <schweeb> it's either gonna be really dead in here the next couple weeks, or really really busy :)
[01:58] <ogra> schweeb, at least in the UdU week we'll have net access...
[01:58] <schweeb> :)
[01:58] <ajmitch> thankfully
[01:58] <ogra> so we'll be around from time to time:)
[01:58] <ogra> ajmitch, yes, dont forget that
[02:00] <zul> its going to be dead because everyone is going to be on airline flights
[02:00] <schweeb> I'll be stuck here!
[02:01] <ogra> yeah, you fly two days from here :)
[02:01] <schweeb> but I can't IRC from work anymore :(
[02:01] <zul> how come?
[02:01] <schweeb> it just got blocked
[02:01] <ogra> even if its only 22h i start here at sat. and arrive on mon.
[02:01] <ogra> weird TZ crossing ;)
[02:01] <zul> bloody long flight
[02:01] <schweeb> I gotta figure out something to do while I work at Chrysler... there are proxies, but I just have to find them
[02:02] <ogra> zul, i'm a strong smoker....
[02:02] <ogra> i'll die i guess
[02:02] <zul> heh...cant smoke on airplanes
[02:02] <schweeb> haha
[02:02] <schweeb> you're screwed
[02:02] <ogra> yep
[02:02] <schweeb> it's a good time to start the quitting process!
[02:03] <zul> or smoke a lot before going on the airplane
[02:03] <schweeb> could just buy some patches
[02:03] <ogra> thats the idea....
[02:03] <zul> that isnt any fun
[02:03] <ogra> nope, but better then going mad
[02:03] <zul> id pay to see that
[02:04] <ogra> zul, sorry, no videocam :)
[02:04] <zul> dang
[02:04] <schweeb> yea, I'd like to see him go off on a stewardess
[02:04] <ogra> i'd tape it for you, really
[02:04] <ogra> lol
[02:04] <zul> i puked on a stewardess once
[02:04] <schweeb> "my beer is warm, bitch!"
[02:04] <zul> when i was a kid...probably too much info
[02:04] <schweeb> hehe
[02:04] <ogra> *g*
[02:07] <zul> eww...why am i listening to dr dre
[02:08] <schweeb> good question.
[02:09] <mjc> http://325i.org/media/bootcharts/
[02:09] <mjc> my bootcharts still stink
[02:13] <zul> nah...iron maiden
[02:18] <mjc> mono-1.1.6 doesn't build on amd64. poo
[02:19] <schweeb> really?  I thought they touted that it actually worked on amd64
[02:19] <schweeb> if I read the release notes right
[02:22] <mjc> using buildpackage
[02:22] <mjc> soemthing's possibly not right with the configuration stuffs
[02:27] <mjc> anyone doing boot process speedup stuff around?
[04:15] <queuetue> Hi, all.  I've got a running ubuntu hoary Athlon system, and I pulled the drives and put them into a new Athlon 64 system... and the disks get errors - buffer I/O on logical block 0...  I can boot the system with DSL cd just fine, and mount, fsck, manipulate the disks just fine, no errors.  But I try to boot ubuntu, and the disk errors come back.  Does anyone have a clue?
[04:16] <Burgundavia> queuetue, #ubuntu for help please
[04:17] <queuetue> Actually, I tried there first, and I thought it might take someone at the developer level  to answer the question...  But if it's not possible to get help here, I'll go elsewhere.
[04:19] <Mithrandir> when the kernel has started loading or at the grub stage?
[04:20] <queuetue> Mithrandir, After loading - Immediately after "starting Ubuntu" or "Welcome to Ubuntu"... Forget exact string.
[04:21] <Mithrandir> ok
[04:21] <Mithrandir> sounds like a kernel bug.
[04:21] <Mithrandir> what kind of motherboard do you have?
[04:22] <queuetue> MSI K8N Neo4 Platinum/SLI
[04:23] <queuetue> (It's Nforce 4 SLI based, if that helps)
[04:23] <Mithrandir> hmm
[04:24] <Mithrandir> others report they have it working "with a little work", which can mean _anything_
[04:25] <queuetue> Mithrandir, WHat direction would that work take?  I don't mind experimenting, but I'm at a loss as to where I should even start.
[04:25] <Mithrandir> queuetue: I'd try booting with acpi=off, noapic, nolapic and any combination of those
[04:26] <Mithrandir> also, does the 32 bit live CD work?
[04:26] <queuetue> Mithrandir, I don't have on on hand, and no cd-burner (excet the one on the MSI) so I can't check atm.
[04:29] <queuetue> All three does not do it...  If all three do not fix it, is there valeu in trying other combos?
[04:33] <Mithrandir> not really :/
[04:34] <Mithrandir> I'm leaving for Australia in a couple of minutes, so I can't really help you more; sorry.
[04:34] <queuetue> Mithrandir, Ok, can you point me towards some topics to research?  (if not, good trip!)
[04:34] <Mithrandir> queuetue: I think there might be some stuff on the wiki -- wiki.ubuntu.com
[04:34] <Mithrandir> sorry about not being more specific.
[04:34] <daniels> in general, nforce4 seems to be fine
[04:35] <daniels> i've never had a problem with my nf4-sli board
[04:36] <queuetue> daniels, Would it matter that this is an install that started on 32-bit and then the disks were migrated?
[04:36] <queuetue> I thought AMD64 was back-compatable...
[04:39] <daniels> it is in theory
[04:39] <daniels> i've never tried it
[04:40] <queuetue> daniels, Any way  you can think of for me to install a "proper" amd64 kernel withotu being able to boot (although can access disks fine from DSL)
[04:41] <queuetue> Could I just chroot an use apt?
[04:41] <daniels> yeah, you could just chroot
[04:42] <queuetue> How do people even *work* on windows machines? :)
[04:42] <queuetue> The lack of flexibility mus be stifling.
[04:43] <zenwhen> queuetue, if you were born in a ten foot room and never left it, you would be satisfied with your situation.
[04:44] <zenwhen> as such windows users who have never tried an open source operating system are also satisfied.
[04:44] <zenwhen> the literally don't know they are constrained.
[04:49] <queuetue> Ok, the chroot has worked - do I need to switch repositories?  I'm not sure how to "do 64" on ubuntu...
[06:25] <erich> Any hints on how to get Java 1.5 working again?
[06:42] <cc> has anyone here successfully compiled qemu with gcc4? are there required patches, and if so, would love to be pointed to them; thanks
[07:22] <fabbione> morning
[07:27] <robitaille> good evening!  it has been very quiet around here tonight
[08:25] <pitti> *wave*
[08:26] <ajmitch> hi pitti 
[08:26] <pitti> hey ajmitch
[09:34] <zyga> hey everyone :)
[09:35] <robitaille> hi!
[09:35] <zyga> wow, new firefox security edition is out
[09:35] <zyga> 1.0.3
[09:37] <robitaille> a new firefox means a lot of questions in the mailing list:  "where is Firefox 1.0.3 in Hoary?".  Similar to Warty and Firefox 1.0
[09:39] <Burgundavia> already seen a few
[09:39] <Burgundavia> almost makes it enough to just put it there, just to make them be quiet
[09:39] <zyga> is hoary going to get it as a part of hoary-updates?
[09:39] <desrt> 'security edition'?
[09:39] <zyga> desrt: bugfix release really
[09:39] <zyga> http://www.mozilla.org/projects/security/known-vulnerabilities.html
[09:39] <desrt> heh
[09:40] <robitaille> I believe most of the bug fixes in 1.0.3 are already in Hoary's firefox (from a comment from Thom a while back)
[09:40] <zyga> pretty ugly bugs were fixed
[09:40] <desrt> holy crap
[09:40] <desrt> javascript has lambda?
[09:40] <zyga> desrt: err.. no?
[09:40] <Burgundavia> robitaille, then bumb the number, just to make people happy
[09:40] <desrt> it does
[09:40] <desrt> http://www.mozilla.org/security/announce/mfsa2005-33.html
[09:40] <desrt> or at least the mozilla people think it does :)
[09:41] <zyga> desrt: I'm reading that ATM
[09:41] <zyga> desrt: I think they mean something different than what you and I think ;] 
[09:41] <zyga> anonymous function
[09:41] <desrt> i mean the lambda that ...
[09:41] <zyga> hmm okaaay ;] 
[09:41] <desrt> ya. that's the lambda i mean
[09:41] <zyga> :)
[09:42] <desrt> i don't imagine you get partial application with javascript :P
[09:42] <desrt> although you can use lambda as a basis for closures
[09:42] <zyga> desrt: partial application?
[09:42] <desrt> \x -> func arg x
[09:42] <desrt> ya.. say like you have some function
[09:42] <desrt> plus x y = x + y
[09:42] <zyga> ah
[09:43] <desrt> you can partially apply it
[09:43] <zyga> let's use lisp notatnion - it's easier to read
[09:43] <desrt> (ie: give it a single argument
[09:43] <desrt> (plus 2)
[09:43] <desrt> then you have a function that takes 1 argument and adds 2 to it
[09:43] <zyga> desrt: that's a closure, no?
[09:43] <desrt> the act of giving 1 argument to a 2-argument function (and in general, n arguments to an m-argument function n<m) is called partial application
[09:44] <desrt> and it's a way of generating closures
[09:44] <zyga> :-)
[09:44] <zyga> I see, I didn't know that term but I understand the thing it does
[09:44] <desrt> so, the idea is that any language that has lambda supports partial application
[09:44] <desrt> say you have some function f x y
[09:44] <desrt> you might not be able to say (f 5)
[09:44] <desrt> but you can always say \y -> f 5 y
[09:44] <zyga> and call the result that with the remaining argument
[09:45] <desrt> exactly
[09:45] <zyga> :-)
[09:45] <zyga> robitaille: does ubuntu use vanilla firefox?
[09:46] <desrt> hell no
[09:46] <robitaille> zyga,  no
[09:46] <desrt> ubuntu's firefox is so greatly patched that they're not even allowed to use the firefox logo anymore :)
[09:46] <robitaille> zyga,  a few differences...starting by the icon :)
[09:46] <zyga> hmm
[09:47] <zyga> too bad it does not include the patch that fixes the infamous slashdot bug ;-)
[09:47] <robitaille> zyga,  you can always download and use a vanilla firefox if you want; it's pretty easy to install
[09:47] <zyga> robitaille: I like gnome integration :>
[09:48] <robitaille> zyga,  I do as well, that's why for Hoary I use Ubuntu's firefox.  For Warty, I wanted 1.0, so I was using my own then
[09:50] <desrt> *hit
[10:14] <GheRivero> res
[10:37] <pitti> mvo: hi
[10:39] <mvo> hey pitti 
[11:10] <seb128> elmo: around ?
[11:17] <seb128> and who broke my box ? :p
[11:18] <Lathiat> me me me pick me
[11:18] <seb128> gdm doesn't start because there is no mouse (apparently psmouse doesn't get loaded)
[11:18] <Burgundavia> nice
[11:18] <maswan> can I break your box tomorrow? please?
[11:18] <seb128> I'm away for a week so feel free :p
[12:39] <abelli> ciao a tutti
[12:41] <abelli> a friend has done a nice manager for boot services, and i'd like to add "service description"s, where can i look for them?
[12:42] <abelli> - /ill be back in 5 min/
[01:33] <mvo> hey pitti 
[01:36] <bluefoxicy> is there a utility to find all files in /etc that have been changed since install time, or aren't related to an installed package, and save them?
[01:37] <bluefoxicy>  /etc is the missing link in erasing / and reinstalling as long as /home is elsewhere :)
[01:37] <infinity> Woohoo, new firefox and mozilla.
[01:37] <bluefoxicy> (that's my answer to any problems incurred from damaging your system)
[01:50] <zul> heylo
[01:51] <abelli> ciao
[03:30] <pitti> Good bye guys, see you down under!
[03:30] <Lathiat> cya pitti 
[03:30] <Lathiat> going myself now
[03:31] <Lathiat> cept im already down under, just going accross :)
[03:31] <infinity> pitti : See you there.
[03:31] <ogra> infinity, really ?
[03:31] <abelli> huh ... have a good time ppl
[03:32] <infinity> ogra : Really!
[03:32] <ogra> YAY
[03:32] <infinity> Huzzah, even.
[03:32] <abelli> ogra: im waiting for your sms.
[03:32] <ogra> infinity, looking forward to meet you then...(i'll meet pitti in 4h at the airport)
[03:33] <infinity> I'm looking forward to getting horribly lost in Sydney after I land.  <nod>
[03:33] <abelli> ogra: mind your change.
[03:34] <\sh> grmpf...python2.4-qt3  but not python2.4-kde3
[03:34] <\sh> why are the bindings not build from kdebindings
[03:36] <Riddell> \sh: not sure, been meaning to ask that
[03:37] <\sh> Riddell: well...possible explanation: the standalone packages are also in the repos, but it looks like they're outdated...
[03:39] <\sh> i will play with kdebindings ;)
[03:41] <Riddell> \sh: cool
[03:41] <Riddell> \sh: try asking haggai first, he did kdebindings
[03:41] <\sh> Riddell: http://www.ubuntulinux.org/wiki/UniversePythonTransitionTODO
[03:41] <\sh> "let kde people handling the problem" ;)
[03:43] <\sh> finally they commented them in in the configure.in file
[03:46] <Riddell> ssh is horribly laggy from here
[03:46] <\sh> hehe
[03:50] <\sh> well...i will do some serious stuff now..showering, shopping, cleaning the community places in this house
[03:53] <mvo> hey ogra 
[04:04] <abelli> mvo: dont stress him now ... otherwise he'll forget shirts.. :)
[04:08] <queuetue> Hello, I've done a fresh install of hoary-AMD64 and I'm encountring the same kernel bug as last night:  "Buffer I/O error on device hda, logical block 0" and a kernel panic on boot...  (This was an MSI K8n Neo4 Platinum/SLI  with the nvidia nforce4 chipset)  ... can anyone lend a hand in helping me understand this?
[04:08] <queuetue> (This error occurs immediately after "Starting Ubuntu"
[04:19] <vvl> queuetue: do other operating systems work normally?
[04:19] <abelli> vvl: ?
[04:20] <queuetue> vvl, DSL (32-bit Dman small linux) works fine - I don't *have* any other oses. ;)
[04:20] <vvl> ok :)
[04:21] <vvl> I got a similar error once when I had my jumper settings on my hard drives the wrong way
[04:21] <vvl> but it's not that then
[04:22] <queuetue> "Wrong way" how?
[04:23] <vvl> don't remember exactly how they were configured, but I think one was configured as cable select and one as a master
[04:24] <vvl> and then they were plugged in wrong
[04:24] <vvl> so linux saw the drives but they didn't respond to I/O requests very well :)
[04:25] <vvl> but if you can boot another kernel that isn't the problem
[04:29] <queuetue> vvl, Well, I can boot a livecd, nothing has booted fromt he HD yet...
[04:42] <vvl> queuetue: I'm absolutely no expert, but check all your HD cables and jumpers, and maybe your bios settings
[04:43] <vvl> some drives require LBA to be switched on, others require it to be switched off
[04:46] <queuetue> vvl, Why would that not be an issue during the install?
[04:48] <queuetue> The fact that the install and DSL can use the drive fine is really the strange thing here...
[04:49] <vvl> hmm yes
[04:51] <vvl> but I'd rather not give any more "advice", like I said I'm no expert at these things :)
[05:47] <fgx> where are bugs concerning universe stuff to be submitted?
[05:49] <Burgundavia> fgx
[05:49] <Burgundavia> malone
[05:49] <koke> fgx: https://launchpad.ubuntu.com/distros/ubuntu/+bugs
[05:50] <fgx> is it useful, as in bugzilla, to include my hd db ID?
[05:53] <zul> bbl
[06:04] <ross> mjg59, ping?
[06:10] <rburton> thom, ping?
[06:12] <kent> Didn't firefox/mozilla get a securityupdate in hoary? I thought i read something about that. But I have not yet got it on my Hoary :(  Whats the version of the update? So that i can check if its already installed.. 
[06:22] <Burgundavia> rburton, there are a lot who are leaving for lca right around now
[06:22] <Burgundavia> kent, it is coming, don't worry
[06:23] <kent> Burgundavia, ah, ok. I was just worried that perhaps my list of repositories is wrong, but I dont think so.. it should be ok :)
[06:24] <rburton> Burgundavia: yeah, that was my guess
[06:25] <Burgundavia> kent, you probably won't see a version bump, they will just backport the fixes
[06:34] <schweeb> fabbione: around?
[06:36] <fabbione> schweeb: more or less :)
[06:37] <queuetue> It definitely seems like a kernel problem - the disks work fine in another machine, the system works fine with both the livecd and a windows PE disk... Wht can I do here?
[06:39] <schweeb> fabbione: you have a recommended netboot image for an ultrasparc to get ubuntu on it?
[06:40] <queuetue> An old gentoo HD I scrounged up has the same problem, btw.
[06:40] <schweeb> sittin in front of the sunblade 2002 right now
[06:40] <fabbione> schweeb: the one that is in the archive @ sparc.u.c
[06:40] <schweeb> kk
[06:40] <fabbione> no idea if it will ever work :)
[06:40] <fabbione> you might have to play with it
[06:40] <schweeb> yea
[06:41] <schweeb> we'll see
[06:41] <schweeb> first experience installing /anything/ on a sparc :)
[06:42] <\sh> schweeb: oh..:)
[06:42] <\sh> i'd installed debian on a sparc u5 and on a r420
[06:43] <queuetue> As a test, I've agreed to install XP on the system, and see how it acts with the hardware ... I find it very disturbing that I can't op open another console during the install. :)
[06:44] <fabbione> schweeb: debian should install pretty fine.. if you can install sarge, you can upgrade to ubuntu.. i am 99.9% sure about it
[06:45] <kent> Burgundavia, well, im not interested in getting new versions etc. I was mainly interested in if my sources.list was correct. But i will know it later, i guess :)
[06:51] <Burgundavia> kent, there is no reason for a stable hoary system to change its repos
[06:53] <kent> Burgundavia, well, that is as correct as can be,  but haha, during Warty i changed the repos to install some 3rd party-stuff.. etc, and changed to a mirror etc. So my sources.list was a bit full of stuff and # comments etc. But i think i  have taken out the stuff that I dont use any more. So now it should be a clean list of repos.
[06:53] <Burgundavia> kent, that would not be classified as a stable install in my books
[06:55] <kent> Burgundavia, becaus of the thirdparty programs (i think i tried to/and did install java, flash or something like that during warty), or becaus i fiddled with the repos manually?
[06:56] <Burgundavia> kent, fiddled with repos
[07:30] <zyga> hello
[07:31] <zyga> how does gnome-cd gets launched upon inserting audio-cd?
[07:31] <zyga> (I've disabled 'auto-play' from the menu)
[07:31] <Treenaks> menu?
[07:32] <zyga> Treenaks: system->prefs->sth (cannot remember english name)
[07:32] <zyga> Treenaks: I can make a screenshot - the dialog is very familiar even in foreign language
[07:33] <zyga> anyway, the real problem is that it keeps starting, any insight before I scoop the source?
[07:33] <Treenaks> zyga: I know which one you mean
[07:33] <Treenaks> did you re-login?
[07:34] <zyga> Treenaks: no, do I have to/
[07:34] <Treenaks> (should not be necessary, but you never know)
[07:34] <Treenaks> if that works, file a bug anyway :)
[07:34] <zyga> Treenaks: It's not important - it was turned off long time ago
[07:34] <zyga> Treenaks: (if it's something simple I'd prefer to patch it ;-)
[07:37] <zyga> Treenaks: hmm there is a gconf key somewhere
[07:38] <Treenaks> zyga: I don't doubt iy
[07:38] <Treenaks> it
[07:38] <zyga> Treenaks: but since gnome-cd doesn't launch on its own ...
[07:38] <zyga> Treenaks: what can be starting it?
[07:38] <Treenaks> nautilus, or gnome-volume-manager
[07:39] <zyga> the latter is not running apparently
[07:40] <zyga> ahh my mistake - it is running
[07:45] <schweeb> fabbione: you know the key combo to get to OpenBoot w/o using the shutdown -i 0 -g 0 stuff?
[07:46] <zyga> hmm how to use debian/rules to extract .tar.gz & apply patches?
[07:46] <fabbione> send a break on the console
[07:46] <fabbione> schweeb: do you have a keyboard or are you on ser console?
[07:46] <schweeb> keyboard
[07:46] <fabbione> stop + a
[07:47] <schweeb> that have to be done on boot, or does it automatically work from a booted system?
[07:47] <fabbione> it should always work
[07:47] <schweeb> k
[07:50] <akk> Hi -- any installer devs around?  I had some problems installing to a laptop, and am wondering what bugs I should file.
[07:51] <fabbione> akk: just use bugzilla.ubuntu.com, we will take care of addressing the bug to the right component.
[07:51] <fabbione> akk: please be very specific with the error if you can
[07:51] <fabbione> and check if the bug hasn't been reported yet
[07:51] <akk> Thanks.  I'll do that.
[07:51] <akk> Oh, is anyone interested in bugs related to expert mode?
[07:52] <fabbione> akk: yes.. all bugs are important
[07:52] <fabbione> and needs to be addressed
[07:52] <fabbione> better one more than one less
[07:52] <akk> Okay, I'll report everything I saw.  Thanks.
[07:52] <fabbione> akk: but please do a search on bugzilla first to see if it is a known problem
[07:52] <fabbione> managing duplicates take a lot of time
[07:52] <akk> Understood.  Will do.
[07:53] <fabbione> thanks
[07:57] <dizzy> anyone expert of hotplug here?
[07:59] <schweeb> o_O
[07:59] <schweeb> no rarp in the Ubuntu kernel?
[08:00] <schweeb> hrm... guess I'll have to use bootp
[08:03] <fabbione> schweeb: you don't need rarp in the kernel
[08:04] <fabbione> schweeb: what you need is rarpd and tftpd on another machine
[08:05] <fabbione> schweeb: OBP will take care of using rarpd to get the ip and download from tftp
[08:05] <schweeb> yea, I had that... rarpd doesn't need rarp in the kernel?
[08:05] <fabbione> nope
[08:06] <trulux> ajmitch: ping
[08:06] <trulux> ajmitch_: ping
[08:06] <schweeb> this network's pretty goofy... they're already bootin some sun's via network
[08:06] <schweeb> so I may mess around with it later
[08:06] <fabbione> schweeb: OBP doesn't support dhcp for booting.. only rarpd 
[08:06] <fabbione> schweeb: the booting is binded to the mac address
[08:07] <fabbione> see /etc/ethers
[08:07] <schweeb> when I had rarpd running, it was only getting rarp stuff for one box, which wasn't the one that I'm trying to netboot
[08:07] <schweeb> yea, I had ethers set up
[08:07] <schweeb> on my laptop
[08:07] <fabbione> so if your rarpd answers only for that mac address it would be ok
[08:08] <fabbione> schweeb: well, if you have phisical access, just use a cross cable to netboot
[08:08] <schweeb> yea, that's what I'm lookin for right now :)
[08:08] <fabbione> and plug the machine back to the real net before dhcp :)
[08:08] <schweeb> that'd do
[08:09] <fabbione> yeps
[08:10] <Amaranth> If I had pbuilder setup for hoary and I'm on breezy now using pbuilder will still create packages usable in hoary, right?
[08:10] <schweeb> yes
[08:10] <fabbione> Amaranth: please stop this backporting madness.. really
[08:10] <schweeb> heh
[08:10] <fabbione> most of the stuff that will build and work on breezy will not on hoary
[08:11] <fabbione> because of the toolchain changes
[08:11] <Amaranth> fabbione: Heh, two packages my menu editor needs have fixes upstream.
[08:11] <Amaranth> the gnome-menus 2.10 branch and PyXDG
[08:11] <Amaranth> nothing that won't work in hoary
[08:11] <fabbione> sooner or later that won't work.. just be aware of that
[08:11] <Amaranth> why won't it?
[08:11] <Amaranth> pbuilder sets up a chroot, doesn't it?
[08:12] <fabbione> because if we apply a patch to fix compilation with gcc-4, it might not work with gcc-3.3
[08:12] <fabbione> also.. new libc6 will generate dependencies that are not easy to satisfy in hoary
[08:12] <Amaranth> oh, you mean for new packages
[08:12] <fabbione> no.. also for old ones
[08:12] <Amaranth> yeah, i'm not going to be working on the backports project
[08:12] <Amaranth> well, i mean for packages that come later on in breezy
[08:13] <Amaranth> I just need these two because they got fixed upstream less than a week after hoary came out and my menu editor needs them.
[08:13] <Burgundavia> Amaranth, did you do the original backports?
[08:13] <Amaranth> nope
[08:15] <schweeb> ugh, this crossover doesn't seem to be providing link :-/
[08:18] <Amaranth> hrm, the gnome-menus source package doesn't have all the source in it
[08:36] <bluefoxicy> 25698 bluefox   15   0  801m 184m  18m S  1.7 24.7  16:26.80 AbiWord-2.2
[08:36] <bluefoxicy> 801M, 184M
[08:37] <bluefoxicy> yeah I'm gonna call that a problem with the allocator
[08:38] <bluefoxicy>  3772 bluefox   15   0  183m  79m  19m S  0.0 10.7   0:04.75 abiword
[08:38] <bluefoxicy> closed abiword, then reopenned with the same file
[08:39] <zyga> bluefoxicy: could you compile abiword with debug and run it via valgrind?
[08:39] <bluefoxicy> zyga:  no, I have work in 20 minutes.
[08:39] <zyga> bluefoxicy: ok
[08:40] <bluefoxicy> zyga:  but I believe that a lot of memory issues stem from the fact that I'm practically the only person I know of that actually understands that the design of the heap is a really bad idea
[08:40] <zyga> bluefoxicy: ?
[08:40] <zyga> bluefoxicy: why
[08:40] <bluefoxicy> and red hat (who controlls glibc and thus mallo()) doesn't care
[08:40] <bluefoxicy> zyga:  because if you allocate a ton of memory, then free most of it, except one byte at the end, you hold a ton of memory, as the heap is defined by start and length only
[08:41] <bluefoxicy> the heap can't be fragmented in allocation; it's one big block.
[08:41] <bluefoxicy> and thus, those unused areas will eventually get swapped out.
[08:41] <zyga> bluefoxicy: hmm not so sure, are you confident that how it works?
[08:41] <bluefoxicy> then when you allocate again, you just swap them back in because thankfully glibc's allocator doesn't just infinitely allocate more
[08:41] <bluefoxicy> zyga:  I've tested it
[08:41] <bluefoxicy> lemme find a test program.
[08:42] <zyga> bluefoxicy: It's easier to check the source
[08:42] <zyga> bluefoxicy: don't we have a slab allocator?
[08:43] <bluefoxicy> zyga:  not that allocator, the userspace allocator.
[08:43] <bluefoxicy> the kernel allocator handles physical->virtual mappings
[08:43] <bluefoxicy> the userspace allocator handles getting memory from the kernel and using it in such a way that it doesn't just eat up 40 gigs of RAM
[08:44] <zyga> err but it's rather hard to believe that the us allocator is so dumb ;] 
[08:44] <zyga> (not that I really know)
[08:44] <bluefoxicy> zyga:  it uses brk(), except if an allocation is >128KiB it makes an mmap() segment
[08:44] <zyga> mmap is okay6
[08:44] <bluefoxicy> it would be better to use say 4-20 meg mmap() segments as mini heaps so that when they empty out you can just free them back to the system
[08:45] <bluefoxicy> you can't just use mmap() for everything because it always maps full pages
[08:45] <bluefoxicy> i.e. mmap()ing 5 bytes maps in 4096 bytes (think of fat, and small files, and you'll get it)
[08:45] <zyga> bluefoxicy: well personally I don't assume I'm working with a perfect allocator and keep small stuff in their own allocation blocks
[08:46] <zyga> bluefoxicy: I know how that works
[08:46] <bluefoxicy> I know
[08:46] <bluefoxicy> unfortunately using a huge block has problems, i.e. firefox starts at 40M and grows to 100, and stays there no matter what
[08:46] <bluefoxicy> or if you delete 6000 files in nautilus, it allocates 330M of memory to store information about what it's about to do
[08:47] <bluefoxicy> then it frees down to 300 instead of back to the usual 20M that it uses
[08:47] <zyga> well simple likned list of blocks that feeds a custom allocator is simple
[08:47] <bluefoxicy> but if it needs another 200 or so megs it's already allocated; most of that 300M is just empty space in the heap
[08:47] <bluefoxicy> problem being it's empty allocated space that should be going back to the system, not to swap after 10 minutes
[08:49] <zyga> bluefoxicy: I know little about how kernel works in this regard (overallocation)
[08:50] <zyga> bluefoxicy: but you do have a point
[08:51] <zyga> bluefoxicy: but ...
[08:51] <zyga> bluefoxicy: isn't is then possible to have a custom allocator that knows about 128KiB stiupiditity and always maintains chunks in mmaped pages
[08:51] <bluefoxicy> http://rafb.net/paste/results/1pgZrc68.html
[08:52] <zyga> bluefoxicy: then spitting them out as a page becomes empty?
[08:52] <zyga> (not that I care about performace for a moment)
[08:52] <zyga> (alhorightm performace that is)
[08:53] <zyga> bluefoxicy: dropping heap holder...?
[08:53] <bluefoxicy> zyga:  problem being that if you simply mmap() a segment for every allocation, you allocate a 4096 byte page when you need a 60 .  You have to mmap() a small-ish area of a fair size, and use that as a heap, then scrap it when it empties.
[08:53] <zyga> what does that mean?
[08:53] <bluefoxicy> zyga:  one byte at the end of the heap.
[08:53] <zyga> bluefoxicy: I dont want to mmap everything
[08:53] <zyga> bluefoxicy: mmap -> list of blocks per allocated maps 
[08:53] <bluefoxicy> zyga:  I malloc() like http://rafb.net/paste/results/B1ug1b73.html
[08:54] <bluefoxicy> 25600 segments, then 100 bytes at the end
[08:54] <zyga> bluefoxicy: smart
[08:54] <bluefoxicy> then I free the 25600 segments out of the heap, which now has abotu 100M of unused memory in it
[08:54] <zyga> bluefoxicy: and shows the point
[08:54] <bluefoxicy> but the heap can't brk() down
[08:54] <zyga> bluefoxicy: so let's change that ok?
[08:54] <bluefoxicy> UU----------------------------U
[08:54] <zyga> bluefoxicy: we'll use a custom allocator on top of mmap
[08:54] <bluefoxicy> zyga: http://bugzilla.gnome.org/show_bug.cgi?id=171547
[08:54] <bluefoxicy> zyga:  exactly :)
[08:55] <zyga> bluefoxicy: just say one thing to me - is there any limit to number of mapped pages? (per user)
[08:55] <zyga> bluefoxicy: I once did a test
[08:55] <bluefoxicy> zyga:  Depending on security setup and system limits there is
[08:55] <zyga> bluefoxicy: I've mapped one gig file, each page on it's own
[08:55] <zyga> bluefoxicy: I could not get more than 10 pages 
[08:55] <bluefoxicy> wli reports that the kernel can handle millions of mappings, more than a 32 bit CPU can actually account for managable ram
[08:55] <bluefoxicy> without slowing down at all
[08:55] <zyga> bluefoxicy: mapping the whole file with one call worked
[08:55] <bluefoxicy> heh
[08:56] <zyga> bluefoxicy: I would gladly repeat that test :)
[08:56] <zyga> and re-run your test with a small custom allocator (with malloc interface of course)
[08:57] <bluefoxicy> zyga:  :) perhaps I should make a quick Ubuntu bug
[08:57] <bluefoxicy> that points at that bug
[08:58] <zyga> bluefoxicy: I'm still reading your bugreport
[08:58] <zyga> bluefoxicy: but ... isn't this a generic bug?
[08:58] <zyga> bluefoxicy: why gnome?
[08:58] <zyga> (why ubuntu for that matter)
[08:58] <zyga> create a small library with malloc free calloc and realloc 
[08:59] <zyga> and LD_PRELOAD it or something
[08:59] <bluefoxicy> because gnome devs get hard-ons over using less memory
[08:59] <zyga> bluefoxicy: I've seen that memory wiki - it's a great idea 
[08:59] <lsuactiafner> ok ppl, anyone here running ubuntu-5.04-install-amd64.iso? the kernel in 5.04 doesnt run my hardware setup, and i need someone to compile me a kernel, i got a .config for my system already, but am running 32bits and i dont want to cross-compile
[09:00] <zyga> lsuactiafner: hmm I guess you should ask in #ubuntu
[09:00] <lsuactiafner> i did, they seem useless.
[09:00] <zyga> bluefoxicy: BTW: how does new get it's memory? via malloc?
[09:01] <dhonn> lsuactiafner, you can compile your kernel under any linux distro
[09:01] <lsuactiafner> and btw the lilo package installation in ubuntu-5.04-install-amd64.iso doesnt work
[09:01] <lsuactiafner> dhonn : you cant compile a 64 bit kernel while running a 32 bit system with no cross compiler
[09:01] <zyga> lsuactiafner: grub works okay (/me hugs big amd box)
[09:02] <dhonn> you can
[09:02] <dhonn> get the compiler
[09:02] <lsuactiafner> dhonn : with a cross-compiler... and i do not want to mess with gcc 
[09:02] <zyga> bluefoxicy: is there any convenient .tar.gz with your tests?
[09:02] <bluefoxicy> work time
[09:02] <bluefoxicy> zyga:  download the link I showed with the source as text.
[09:02] <dhonn> what kernel works for you and what linux distro?
[09:02] <lsuactiafner> zyga : lilo is better than grub
[09:03] <zyga> bluefoxicy: msg me when you finish work please
[09:03] <lsuactiafner> i use slackware 2.6.11, some options in the ubuntu kernel conflicts with my hardware.
[09:03] <dhonn> just copy the kernel over to ubuntu
[09:03] <lsuactiafner> slackware 10.1 but kernel 2.6.11-4
[09:03] <zyga> lsuactiafner: I see
[09:04] <dhonn> and copy the /lib/modules/*
[09:04] <lsuactiafner> dhonn : a 32bit kernel WILL NOT run binaries compiled for 64bit
[09:04] <dhonn> get the 64bit kernel from slack
[09:05] <dhonn> lsuactiafner, you say the slack 64 kernel works for you?
[09:05] <lsuactiafner> i'm running slackware in 32bit mode.
[09:05] <dhonn> damn
[09:05] <lsuactiafner> slackware 10.1 is 32bit, patrick didnt like some nasty bugs with 64bit systems
[09:05] <dhonn> is there a 64bit kernel disto that works for you
[09:06] <lsuactiafner> no. thats why i need someone currently running a 64bit system to compile me a kernel that will work for me, i got a .config file that will work
[09:07] <schweeb> fabbione: still there?
[09:08] <lsuactiafner> dhonn : you aint a developer are you?
[09:08] <dhonn> im a different kind of developer
[09:10] <lsuactiafner> graphics?
[09:10] <dhonn> yep
[09:11] <dhonn> i use to compile kernels a lot though
[09:12] <dhonn> you can compile a 64bit kernel in a 32bit/os partition, then copy it to your 64bit/os partition
[09:12] <lsuactiafner> you can not.
[09:12] <dhonn> yes
[09:13] <lsuactiafner> dhonn : --march=athlon64 
[09:13] <dhonn> yes
[09:13] <lsuactiafner> try compiling a program with that if you got a 32bit compiler
[09:13] <dhonn> you can compile any arch
[09:13] <dhonn> under any os
[09:13] <lsuactiafner> trust me.. it does not work.
[09:13] <lsuactiafner> ok cd /usr/src/linux 
[09:14] <lsuactiafner> edit the Makefile
[09:14] <lsuactiafner> ill get a search quick
[09:14] <|QuaD-> is there a reason that all of unstable hasn't been imported to breezy? looking at gnome-blog, sid has v0.8-2, we are on 0.7-4
[09:15] <schweeb> fabbione: I keep getting a "Fast Data Access MMU Miss" on a tftp download from the Sun client... using tftp manually from an x86 box works just peachy keen
[09:15] <lsuactiafner> change the CFLAGS to export CFLAGS="-march=athlon64 -O3 -pipe -funroll-loops"
[09:15] <schweeb> and on my x86 tftp box, the only error I get is "read: Connection refused"
[09:15] <lsuactiafner> see if it does compile...
[09:15] <lsuactiafner> HOSTCFLAGS ...
[09:16] <dhonn> something isnt installed
[09:16] <dhonn> dhonn@laptop:~/temp$ gcc -march=athlon64 blah.c
[09:16] <dhonn> i get errors too
[09:17] <lsuactiafner> exactly.
[09:17] <dhonn> works fine march=pentium4
[09:17] <lsuactiafner> well you know what, an AMD64 3200+ Hammer aint the same thing as a Pentium 4
[09:18] <lsuactiafner> i686 is compatible. 
[09:18] <dhonn> you have to install something else, you can compile programs for other archs i know it
[09:18] <lsuactiafner> a cross-compiler...
[09:18] <lsuactiafner> and you do not want to fsck with gcc on any pc.
[09:18] <lsuactiafner> ok now i'm wasting time here.
[09:18] <lsuactiafner> dhonn : dont ever touch anything todo with ubuntu please.
[09:20] <dhonn> gcc is a cross compiler
[09:20] <dhonn> lol
[09:24] <fabbione> schweeb: oh... that error!
[09:24] <srbaker> jdub, are you on tv right now?
[09:24] <fabbione> schweeb: even with debian?
[09:25] <fabbione> schweeb: because iirc the problem. it means that your cpu isn't supported
[09:26] <schweeb> :-/
[09:26] <schweeb> I'll try debian
[09:26] <schweeb> google says it's a kernel problem
[09:27] <schweeb> it is definitely an ultrasparc though
[09:27] <fabbione> schweeb: it can also be silo
[09:27] <fabbione> that's why i am asking you to try debian
[09:27] <schweeb> yea
[09:27] <fabbione> it compiles silo with another gcc
[09:27] <schweeb> looking for the image now
[09:27] <schweeb> it's an UltraSparc-III+
[09:28] <fabbione> not sure it's supported by the kernel
[09:28] <schweeb> o_O... the banner says Sun-Blade-1000... it's supposed to be a 2000
[09:30] <schweeb> debian's boot.img does the same thing too
[09:30] <schweeb> trying the 2.6 one...
[09:35] <zyga> confirm this please - it's not possible to get 4MB pages on x86?
[09:38] <jbailey> zyga: Superpages I think are available from pentium up.
[09:39] <zyga> jbailey: any good docs on the subject? (other than google)
[09:39] <jbailey> Intel's archtecture books.
[09:40] <zyga> jbailey: thanks
[09:40] <jbailey> Dunno if they're online or not.
[09:40] <zyga> jbailey: I've got some of them at my university library - that will do
[09:40] <jbailey> Cool.
[09:40] <schweeb> fabbione: okay, now I'm getting it to say "138200" then is says to check the tftp config.  frickin SPARCs
[10:00] <fabbione> schweeb: so it starts the download and than it stops?
[10:00] <schweeb> yep
[10:01] <schweeb> I'll do some more research and mess with it later
[10:01] <schweeb> gotta go to a bbq with whiprush in a little while, and this is pissing me off too much
[10:11] <zyga> does LD_PRELOAD take symbol versions into account?
[10:15] <schweeb> headin out... later