On Thu, 12 Feb 2004, Ian Truelsen wrote:Amen brother!!!
Excuse me for interrupting, but does it make a lot of sense to be
benchmarking what is essentially a development kernel? I would think
that when they are in the single digits in releases, they are simply
trying to make sure that everything works. Once the 2.6 tree has been
around for a while they will work on speed tweaks.
I beg to differ here. The 2.5 series was for making sure everything worked. Once they renumbered to 2.6, it became production-ready. I don't understand how commercial software gets sneering comments about waiting for the first couple of patchsets before using it but open source can be apologized for when you have to do the exact same thing.
Sorry for the rant but sometimes what appears to be hypocrisy just drives me nuts.
Ric
-- [EMAIL PROTECTED] mailing list
.. Anyhow.. I have been following this thread like a crazed madman.. (this and 2 others about pretty much the same thing, including 1 that I started).
What I am seeing now.. is that yes, in fact the 2.6.X kernel has been giving almost everyone the same issues.
Now, my question is.. what are these developers doing to make "their" systems to fast? Could it be that they had slow systems from the start?
I have been in #linuxfriends on irc.freenode.net in a little dispute about this topic also.. it seems that my friend there... who previously used a 2.4.X kernel had such poor performance that he could not even start a game.. say ut2003 for instance. I on the other hand have been playing games on my Linux computers all the way back as far as SuSE 6.0 and Redhat 6.2 and have had no problems at all.
So it could be that our friends are not setting up their kernels correctly in the first place, and somehow installing gentoo and a 2.6.X (lets say a "hybrid") kernel have magically fixed their systems.
I dunno.. Most people whom I have spoken with have nearly the same hardware.. I.E. a newer AMD proc.. Athlon XP etc..etc. and newer motherboards.. I.E. nforce based mobos and the like. So it could also be that their newer hardware also contributed to the fact that there is better support in their newer kernels.
Anyhow I really don't know.. This whole situation has been quite boggling to me so far and that is why I enjoy and encourage as much input about this as possible.
So we can see so far that maybe 1 out of 5 people who prefer a 2.6.X kernel have actually not seen any gains, but still like the 2.6.X kernel enough to stick with it.
This is very strange... and not an anomaly... as this problem actually exists.. and is not in the minds of the people at large.. and not in the configuration of said kernels.
Like I said in earlier posts (and other threads) I currently have at my disposal the following kernels, all configured with the same options, except the newest performance "enhancing" hacks provided by newer 2.6.X kernels.
linux-2.4.23 linux-2.4.23-pre9 linux-2.4.24 linux-2.6.2 linux-2.6.2_rc3-love1 linux-2.6.3-rc1-mm1 linux-2.6.3-rc2
And I use pretty much the same hardware on all of my systems.. (sans mobo and maybe a different model proc) I.E.
All systems have:
Soundblaster Live, or SB16 sound cards Linksys LNE100TX 10/100 Wal-Mart specials NVIDIA based graphics cards AMD processors DDR400 Ram or pc133 ram in my older boxes 3 systems with an nforce chipset maybe 3? systems w/ and VIA chipset
Anyhow on other boxes.. I.E. one of my game servers.. I may even have an old 2.6-test.X kernel running somewhere also.. and I know I have a base stock/vanilla 2.6.0 kernel running somewhere too.
Well, I hope this helps the overall movement.
Thank you
Sincerely,
-- TriKster Abacus irc.freenode.net #cllug #gentoo #linuxfriends irc.cotse.com #linux http://www.cllug.org http://www.trikster.homelinux.org http://www.trikster.homelinux.org/contact.html
-- [EMAIL PROTECTED] mailing list
