dmb: hey
dmb: the catchum
dmb: i think is maybe broken for the forgot your password link
dmb: there is no letters or numbers on it
dmb: unless i'm blind
dmb: http://www.phoronix.com/forums/login.php?do=lostpw
dmb: anyone else agree?
wrongdevice: os[Linux 2.6.29-rc8-zen1 x86_64] distro[Gentoo Linux "~amd64"] cpu[1 x AMD Sempron(tm) Processor 3000+ (AuthenticAMD) @ 1,80GHz] mem[Physical: 2,0GB, 92,4% free] disk[Total: 74,8GB, 36,8% free] video[nVidia Corporation G72 [GeForce 7300 SE]] sound[Audigy2 - SB Audigy 2 [Unknown]]
wrongdevice: :)
peppot: hi, anyone on Jaunty with ATI (mobile) and latest fglrx xorg drivers?
maligor: the pre-release eh
maligor: what ati mobile?
peppot: hd 3470. latest is in jaunty repos apparently
peppot: getting ~1100 fps with fgl_glxgears
maligor: yeah, it's catalyst 9.4 for xserver 1.6 I think
peppot: right.
peppot: I'm having abysmal performance with compositing in metacity
peppot: but stuff like extremetuxracer works fine
maligor: don't use compositing?
peppot: good performance in 1900x800, everything on which couldn't be software rendering
peppot: maligor, I'd rather use it, actually? trying to get to the root of the problem
maligor: 9.4 should work okeyish with opengl compositors
maligor: but dunno what metacity uses for compositing
peppot: it works slow as molasses here
maligor: resizing is slow tho
peppot: very. toggling maximize state on off takes like 1-2 secs, so does alt-tabing
peppot: it seems when I try compiz, it's using indirect rendering. hm!
peppot: or what does this look like to you? http://rafb.net/p/xWJlB110.html
maligor: indirect just means that the gl commands are passed through X
maligor: it doesn't mean it isn't accelerated
Kano: hi redeeman ,does gentoo has got already kernel 2.6.29?
Kano: i need aufs or unionfs for it
maligor: http://developer.amd.com/gpu/ADLSDK/Pages/default.aspx
peppot: well, is performance expected to be this horrible with the catalyst drivers?
maligor: amd definately releases lots of stuff these days
maligor: compiz is a bit sucky to start with
wrongdevice: Kano, No but you can install zen-sources :)
peppot: I am definitely hoping I've run into a bug
Kano: ah
Kano: zen has it?
peppot: maligor, works extremely snappily on my desktop with nvidia
wrongdevice: Kano, yes
Kano: will take a look
wrongdevice: :)
wrongdevice: Kano, ^
maligor: ADL seems like a API for the stuff aticonfig does
Kano: maligor: did you get that benchmark tool runningß
maligor: what benchmark tool?
maligor: oh.. the pcie
maligor: no, didn't manage
maligor: hmm.. one could write a better control panel for fglrx now
maligor: amdcccle is pretty abysmal
maligor: this seems pretty complete
maligor: it also allow one to query workstation capabilities
Kano: the FireGL string check?
maligor: err... no
maligor: it seems to have a bitfield for STEREO feature's
maligor: can't find any way to configure 3d settings tho
maligor: Kano, someone else has it working on linux tho
Kano: hmm with help of the zen source i got aufs to compile
Kano: lets see if it workds
maligor: aufs?
Kano: yes i need it for live mode
Kano: ubuntu has a git for 2.6.29 but without aufs it is useless
Kano: michaellarabel: if it works you can try it on that x58 board you have got
Ivanovic: ha!
Ivanovic: wesnoth is accepted in the summer of code this year
Ivanovic: michaellarabel: since the list of orgs in general might be a nice news item: http://socghop.appspot.com/program/accepted_orgs/google/gsoc2009
michaellarabel: Wow, did X.Org not get accepted?
michaellarabel: oops
michaellarabel: Found page 2 :)
raevol: hahaha
maligor: http://forums.amd.com/devforum/messageview.cfm?catid=328&threadid=110351&enterthread=y
maligor: has some ubuntu people pasting speeds, dunno why it doesn't work for me
maligor: maybe it only works on hd 4000 series?
redeeman: Kano: what do you mean?`gentoo doesn't offer binary kernels, you may compile any kernel you like
Kano: redeeman: well you need extra modules usually
redeeman: suck as?
redeeman: such*
Kano: and those need patches if they are not supported officially
Kano: aufs for example
redeeman: well, i don't know if they have patched aufs
redeeman: but typically patches do get in before kernel release if required
redeeman: and if it's a package anyone actually gives a damn about
Kano: well i looked at xen sources and patched the versino in the ubuntu git
Kano: you need aufs for live images
punzada: the captcha on the phoronix forums is screwed up entirely, it's not displaying letters to input only graphical images
punzada: i can't register :(
punzada: example: http://dl.getdropbox.com/u/23960/phoronix.png
punzada: every refresh is another image without a single value on it
Kano: disabled scripts?
punzada: nope, everything standard is enabled, even tried with adblock disabled
mattmatteh: just checked too, there are no letters, only image
mattmatteh: does not have any java, javascript, crapy flash, or ad block... just plain simple working html
punzada: just tried in my vm as well
punzada: lol
punzada: stock ibex install
punzada: i just wanted to post on that new virtualbox beta thread, got opengl accel working perfectly in linux guest
punzada: very awesome stuff
mattmatteh: search is broke too, only the image, no letters
mattmatteh: but.... search is such a PITA with that letter crap
mattmatteh: i just use google site:phoronix.com so thats not really all that important
bkero: :( The cacti charts when parrot got slashdotted are blank. :( OOM
Kano: punzada: you got it working perfectly? it worked with u 8.10 but when compiz was on then i did not see a mouse cursor
punzada: my mouse cursor was displaying fine
punzada: using nvidia drivers on my host machine, newest ones 185 series
punzada: only thing that wasn't 100% was rotating the cube with the mouse, it slowed down then, everything else seemed perfect
Kano: Host/Kernel/OS "Kanotix" running Linux 2.6.29-1-generic x86_64 [ Kanotix Excalibur 20090318-22:09 ]
Kano: CPU Info (1) Intel Core2 Duo E8400 @ clocked at [ 1999.000 MHz ]
Kano: (2) Intel Core2 Duo E8400 @ clocked at [ 2999.000 MHz ]
Kano: Videocard ATI Mobility Radeon HD 3450 X.Org 1.4.2 [ 1680x1050 ]
Kano: Processes 131 | Uptime 3min | Memory 223.3/1951.5MB | HDD Size 321GB (0%used) | Client Konversation 1.1 | Infobash v2.67.1
Kano: so first test iso with 2.6.29, running live mode
punzada: o.O
Kano: had to recompress the squashfs with squashfs 4.0
Kano: drm code in 2.6.29 still too old for exa support
maligor: didn't know x86 could adjust core clock separately for each core
punzada: I wish my laptop had a more substancial bios, it has boot order and date/time and that's it, newest update from manufacture doesn't update that any either
punzada: I'd like some underclocking/overclocking options
Kano: i guess fglrx will need 2.6.29 patches
maligor: overclocking on laptops tends to be a bad idea
maligor: the thermal solutions on them are pretty borderline to start with
punzada: indeed
Kano: haha, it is not laptop, the id is wrong *g*
punzada: and I doubt I would do it often, I'd just twant the option to
Kano: redeeman: has gentoo 2.6.29 fglrx patches
redeeman: i don't know
redeeman: i kinda don't give a rats ass about fglrx
Kano: found something
Kano: lets try it
Kano: hmm 9-2 is even wrong packaged
Kano: it has 2 patches dirs
Kano: what a clever guy
punzada: I've weened myself away from distros like gentoo, I'm too much of a tweaker and I would spend days just tweaking things that would either 1) cause me to have little actual productivity or 2) patch things to the point that updates make me have to repeat the process
punzada: it just isn't worth it anymore
redeeman: and instead of not using gentoo, you couldnt fathom simply..... altering that behavior?
redeeman: or are you saying, that you are unable to control yourself from eating the candy available?
redeeman: kindof reminds me of those people that claim KDE is bad because it has too many options, that people don't know how to use....
redeeman: well... easy... DONT USE WHAT YOU DONT KNOW YOU NEED
redeeman: yet some people seem to feel that they simply MUST press buttons they don't know what is
punzada: no kde is bad for poor as shit ui design and trying to suck on the design schemes of windows with every major release
redeeman: lol
punzada: lol :D
redeeman: anyone that thinks KDE takes designs from winblows, haven't ever actually used it
punzada: I used it
punzada: i started using it when i first started using linux with slackware
punzada: and you're going to tell me kde4 doesn't suck on the vista display style?
maligor: KDE eats babies!
maligor: punzada, well, not really
punzada: it's vista with bits of osx splashed into it, which i won't hate too badly on it's a good idea
maligor: KDE's approach is quite different
redeeman: punzada: actually, in case you didn't know, most of the vista styles were created for compiz which m$ then took
redeeman: oh and btw, those new things which you say are from osx, actually originated in compiz aswell, crapple stole it
mattmatteh: crapple .. haha
Hxal: KDE's approach is "squeeze everything into a single process and implement auto-restart if it crashes" :P
punzada: um, no, compiz borrowed ideas from osx as well, not saying that didn't go both ways
redeeman: Hxal: not really, KDE has ALOT of things separated out, but obviously stuff restarts if it crashes, would you have it not do that?
punzada: i'd rather have stuff not crash :)
Hxal: I'd rather have separation between plasmoids
mattmatteh: agrees with punzada
redeeman: obviously stuff shouldn't crash, and in 4.2 that's an extremely rare thing, haven't happened to me
redeeman: but stuff can also crash for a variety of other reasons
redeeman: such as bad hardware
Hxal: until you start downloading third party plasmoids :P
redeeman: and should a crash happen, what would you have happen? should it not restart it?
redeeman: well then don't download crapply plasmoids
punzada: I actually don't hate kde4 like I did kde3 through, if I was forced on to a system using kde4 I wouldn't hate using it, if someone tries to force me to use 3.x though I'd rather use console and framebuffer
Hxal: ... a single plasmoid shouldn't be taking the whole desktop down
Hxal: granted, it restarts nicely, but the approach still sucks
redeeman: Hxal: plugins run in the context of a process can obviously crash the process
redeeman: and that aproach is done for a reason
Hxal: obviously, so the solution to design is so that they run in a separate process
Hxal: erm
Hxal: the solution is to design it*
redeeman: to quote punzada earlier, "i'd rather have them not crash"
punzada: nods
redeeman: so the solution is: don't install third party BUGGY software
punzada: well, new buggy software is fun :D
maligor: and if you make a system where every atomic process is a separate process, it'd be invincible!
Kano: Host/Kernel/OS "Kanotix" running Linux 2.6.29-1-generic x86_64 [ Kanotix Excalibur 20090318-22:09 ]
Hxal: uh, if it wasn't for third party plugins, I wouldn't be using kde4
Kano: CPU Info (1) Intel Core2 Duo E8400 @ clocked at [ 2999.000 MHz ]
Kano: (2) Intel Core2 Duo E8400 @ clocked at [ 2999.000 MHz ]
Kano: Videocard ATI Mobility Radeon HD 3450 X.Org 1.4.2 [ 1680x1050 ]
Kano: Processes 132 | Uptime 24min | Memory 311.9/1951.5MB | HDD Size 321GB (0%used) | GLX Renderer ATI Radeon HD 3450 | GLX Version 2.1.8494 Release | Client Konversation 1.1 | Infobash v2.67.1
Hxal: community involvement is very important
Kano: found a working one..
redeeman: Hxal: yes, but community BUGGY plugins are not important
maligor: Hxal, there's a first party?
redeeman: Hxal: you want KDE to work around an issue with your(well, maybe someone elses) crappy plugins
redeeman: that's not how stuff rolls
Hxal: oh yes, everyone writes perfect C++ code...
redeeman: obviously bugs can happen, but then complain to the piece of code that causes it
Hxal: of course, it's an extensible platform, it should be designed to handle crappy plugins gracefully
redeeman: besides, you can be assured that KDE has considered carefully whether to run plasmoids in their own process, and there are obviously reasons for what they do
redeeman: and why?
redeeman: are the kernel designed to handle crappy modules gracefully?
redeeman: maybe xchat is designed to handle crappy plugins too?
redeeman: or openoffice?
maligor: *(NULL) = 10;
Hxal: look at the next generation browsers... chrome and ie8
redeeman: Hxal: handling crap is the same as inviting people to create it
maligor: process separation means you have to use a IPC
Kano: that patch is really small, only 1753 lines ;)
Hxal: having such elitist attitudes is not inviting people to create anything
Kano: a tiny bit overkill i would say
punzada: I don't think there's anything elitist about calling out bad code when you see it
redeeman: Hxal: yes, people are welcome to create stuff, they should know be aware that people arent going to clean up when they smear poo outside the toilet
redeeman: Hxal: KDE is not some housecall toilet cleaning service
Hxal: ^ case in point
redeeman: and besides, YOU are the ones complaining
redeeman: KDE has already made a very working solution, restart if crashes
redeeman: and as such, people should fix their plasmoids
maligor: redeeman, I have to say.. YOU incited it
maligor: I could quote you if you like ;)
redeeman: please do
maligor: [01:05] kindof reminds me of those people that claim KDE is bad because it has too many options, that people don't know how to use....
redeeman: yes
redeeman: which is very true
punzada: that was to my comment that i avoid gentoo now
redeeman: and the next line is golden too
punzada: totally different subject
punzada: lol
maligor: quote so, but it's also practically asking for kde flames ;)
redeeman: punzada: actually it was kindof the same subjet
redeeman: punzada: i don't think you understood my point :P
redeeman: you said you switched from gentoo because you would spend too much time tweaking because you could
redeeman: and i attempted to tell you that nobody forces you to do so on gentoo
redeeman: and that you could simply refrain from doing so
redeeman: and that the argument for ditching gentoo is the same as the argument people use to claim KDE is bad
redeeman: because it supports something one doesn't need
punzada: well i should of elaborated, i meant i spent too much time tweaking, which when update time happened would lead to breakage, which would mean i would have to either have an untweaked system i'm unhappy with, or an un-updated system
maligor: gentoo is just plain bad ;)
punzada: neither of which are good options
redeeman: punzada: but tweaks would also break updates on other distributions?
punzada: the same tweaks possibly, but binary based distributions have me instantly avoid those tweaks
maligor: gentoo has the potential of creating unreproducible bugs
redeeman: but then you are also unhappy with it?
maligor: which is why I don't really like any source distro
redeeman: maligor: yes, but that's not a problem since the user should know what he is doing
redeeman: if some user turns on something very different, or modifies something himself
redeeman: he should know what it entails
redeeman: and atleast provide that information when reporting bugs
maligor: do they ever
maligor: that's why they're called users
redeeman: no, those are called LOOSERS
maligor: the technical term is lusers
redeeman: that is again what im talking about, people shouldn't do stuff they don't know what is
redeeman: like for instance
redeeman: an engineer shouldn't agree to construct a bridge if he cant calculate the required thickness of materials used
maligor: like prod around bootloader source code and make black voodoo changes
maligor: coughs
redeeman: well sure, one can do such things, but one better know that if stuff doesn't go right, one shouldn't smash bugs open that makes no sense
maligor: oh.. my black voodoo change fixed a bug
punzada: redeeman: the problem comes up when new suspension strings for the bridge are installed not matched to the original specs, and all further repairs to the bridge are based off these new strings, meanwhile the old base is still there, hence the many conflicts you run into when dealing with too much source
redeeman: i have also sometimes backed up some files, and made changes i wasnt sure of, but i still knew what i was doing(aka a TEST to see what happens)
maligor: I just don't know why the bug was there in the first place since it's regarding hardware initialization
punzada: compiling from source and optimizing for hardware is great, but it doesn't play well hand in hand with to the minute new software, it's simply not practical
redeeman: yes it does
redeeman: in fact it's far better for bleeding edge stuff
redeeman: as you can easily incorporate new patches, new code etc
punzada: it's far better until there are major updates to required packages
punzada: or dependecies
redeeman: that is not a problem
maligor: redeeman, are you a developer?
redeeman: yes
maligor: most users don't have proficiency
redeeman: and those obviously shouldn't then do some things
redeeman: there are limits to everyones abilities, including myself
redeeman: and it shouldn't be hard to know what one can do
maligor: personally I don't really even want to compile other people's code unless I have a really good reason
redeeman: so some people simply have to think: "hmm, do i know how to get this git checkout, which requires me to also patch this depdendency?"
redeeman: and the answer for most "users" might well be 'no'
redeeman: so while they may have the possibility to attempt, that doesn't mean they should do it
punzada: i'm simply not articulating my point well enough and could be assed to keep attempting to, lol.
maligor: as for kde... with qt 4.5 you can now run kde apps in gnome with native looks ;)
Azerthoth: something I have never been able to work up much of a give a rip about .. program works ..great ... the window decoration gives me warm fuzzy happiness ... you have to be kidding me
superdump: gives the test suite a go in os x
ognirc: help
ognirc: sorry, didn't know what amsg was
ognirc: night all