Phoronix IRC Log: 2009-04-23
aanti: somebody knows how safe it is to increase the pci-e bus ? if so, whats a sane number ?
peppo: anyone have any clue why I'd get horrible 5-6 fps performance with the same video files when running mplayer from gnome-terminal and good performance when running the same video with mplayer but launched from nautilus?
peppo: started happening with the latest fglrx, 9-4
nXs: Hi there, got an issue whith result-file-to-pdf, there are no images... running under sled11, any idea ?
nXs: maybe a way to have a debug mod ?
nXs: anyone ?
nXs: ok got it, php5-gd was missing
nXs: nice piece of code btw
whilo: michaellarabel: ping
michaellarabel: hi whilo
whilo: michaellarabel: have you ever thought about benchmarking memory usage?
whilo: i think this is badly missing at least for linux
whilo: it is not easy to do it right, it is mostly done wrong, and it is still very relevant in many cases
whilo: especially now that foss also comes to smaller devices
whilo: michaellarabel: personally i have had issues with kubuntu and a very high ram usage (400-600mb) for a usual desktop session
whilo: while on gnome and kde3 on lenny i have 200-300 mb
whilo: then on fedora it was 200 mb -300 mb or so for kde4, on opensuse 11.1 it was 140-200 mb for kde 4.1
whilo: this is all "free" (without caches and buffers)
whilo: kde devs propose exmap to be the most accurate solution, but the kernel module for example does not build on kubnut 9.04
whilo: michaellarabel: so this is my particular issue, but i think that in general, especially if you use linux on the desktop, this is an important issue and a proper benchmarking would be really helpful
whilo: michaellarabel: what do you think?
michaellarabel: whilo: Out of it right now, quite hectic, but will look at it within a few hours
whilo: michaellarabel: no prob, ok
Ivanovic: hi michaellarabel
Ivanovic: welcome back
Ivanovic: PTS is now part of "mainline" gentoo (okay, not my ebuild and only the amd64 version, but better than nothing, right? ;) )
redeeman: i wonder what hardware michael used in the recent virtualization benchmark in KVM
evocallaghan: Xen ftw !
redeeman: michaellarabel: in the nehalem virtualization benchmark, what disk hardware did you choose in KVM?
michaellarabel: I'll need to let you know when I get back to the states.
redeeman: well if you did "-hda blablaba.img" then it's the ide hardware
redeeman: and furthermore, it also depends on what harddrive image format you use
redeeman: also, do you know about placement on the actual harddrive?
evocallaghan: Xen is f'ing good
redeeman: evocallaghan: abomination :)
redeeman: xen :P
evocallaghan: The Xen HV
evocallaghan: I don't know what your thinking of..
Azerthoth: xen causes HIV ??? I knew it
redeeman: The exact same thing :P
evocallaghan: Tell me a better HV then Xen
evocallaghan: exactly :p
evocallaghan: *know* what your talking about !
evocallaghan: redeeman: Do some research and or work with virtualization in a enterprise class environment before going around pushing propaganda. No offence.
evocallaghan: OK catch you later tonight lads
michaellarabel: Anyone know offhand if there is any known bugs with regard to NetworkManager dropping a WiFi connection after a while on Ubuntu 9.04?
Milyardo: michaellarabel: I think there may had been a problem with a realtek wireless card driver
Milyardo: but not netowork manager itself