No .bfs is faster kernel?

[Rudge]

FYI. Joble. Downloading the phoenix .iso now. I will have a live CD in a few. 

[Joble]

hehe, live-cd.... that's so yesterday.   

[Crow]

"I've got an atom and the .a64 seems to be spunkier on it, though I haven't done any benchmarks."

a64 isn't designed to improve performance for amd 64 bit cops?

am i missing some important info here?

Seems to work well for the Atom too T6.  Thanks to Hootiegibbon for pointing it out to me.

I have an atom, must try this weekend 

[Rudge]

hehe, live-cd.... that's so yesterday.   

Hay,, I told you I try it. I am trying it. LOL

[Crow]

hehe, live-cd.... that's so yesterday.   

Hay,, I told you I try it. I am trying it. LOL

       

[Rudge]

I done had me a learning curve. It seems that xfce is to Gnome, what lxde is to KDE. If you like Gnome but want a light weight version, use Phoenix,, if you like KDE and want a light weight, use LXDE.

Just my observations. I relight my pipe and humble myself.

[Crow]

I done had me a learning curve. It seems that xfce is to Gnome, what lxde is to KDE. If you like Gnome but want a light weight version, use Phoenix,, if you like KDE and want a light weight, use LXDE.

Just my observations. I relight my pipe and humble myself.

And if I use e17 and want a lighter DE, what should I do?   

[Bald Brick]

And if I use e17 and want a lighter DE, what should I do?   

Try MS-DOS. 

[Crow]

And if I use e17 and want a lighter DE, what should I do?   

Try MS-DOS. 

          

[TheChairman]

I am running 2010 LXDE with the 2.6.32.11-pclos2 kernel on a p4 @ 1.6GHz. (1g ram)

It seems faster than with the 2.6.32.11-pclos2.bfs kernel.

Is it just me or is this a reasonable scenario?  Why would this be so?

Thanks 

Personally, I found the BFS (Badly F*ck*d System) kernel to be an annoying problem on my hardware... particularly when attempting to use the Synergy keyboard/mouse sharing client (synergyc). 

Machine #1: Dell Precision M70 laptop 2.26 GHz - dual boot *WinXPP and PCLOS 2009.2 (Gnome)
Machine #2: Dell Precision 450 2.8 GHz Dual Xeon - dual boot *WinXPP and PCLOS 2010 (Gnome)
Machine #3: ASUS P2B-S (440 BX) 600 MHz P3 - triple boot WinServer2000, *PCLOS 2010 (LXDE & Zen)

* = primary OS   |   PC #3 is my LAMP platform for staging/testing websites locally before upload.

With the Synergy server running on #2 (WinXPP), it works flawlessly between all MS Windows systems.

With BFS kernel on machine #3 (LXDE or Zen installation) synergyc would get killed off and fail to restart even with the --restart switch explicitly given.  Other odd behavior showed up on machine #2, so I decided not to 'upgrade' #1 to PCLOS 2010.   FWIW, the PCLOS 2009 versions ran beautifully on all three machines.  And yes, I did a full Synaptic package system update before digging into the problem any further.

After some research on BFS and reading kolivas' site, I decided the BFS kernel might indeed be the culprit...

Thus I removed the BFS kernel and 'updated' to the standard 2.6.32.15-pclos1 kernel (NON-BFS).  Lo and behold, problem solved on all PCLOS 2010 installations! The standard kernel also fixed a few other odd quirks which showed up on #2 and #3 (e.g. jittery mouse) with BFS.

Now I can use Synergy (and the clipboard) between all three machines, regardless of which OS is running, without the frustration of BFS killing the process.

Ironically, BFS was supposed to benefit my type of 'legacy' hardware.

Frankly, BFS does not yet appear mature enough to be a 'default' kernel at this time.

IMO they should remove BFS and use the standard kernel in future releases to appeal to the greatest common denominator of hardware.  Put a section on the website explaining how (and why) to install a different kernel such as BFS (it's not intuitive or obvious if you're a newbie drifting over from a Microsoft OS).

[gseaman]

The bfs scheduler is designed to be more responsive to user input, not to be faster overall. If it is faster in some benchmarks, that's more of an indication that benchmark is measuring processes that have been getting bogged down waiting for interaction of some type.

If a problem happens because of the bfs scheduler, that is not an intended trade off in the designed, a bug report should be made to help improve the scheduler even further. For many of us, the bfs kernel has improved the user experience. If it were not the default, I would install it.

If you installed synergy from the repos and there is a problem with that program and the default kernel, a detailed report of what hardware you have, what you installed since installing the os, etc., might allow others to benefit from an improved out of the box experience. If you've done all of this and still no luck, then switching kernels is a satisfactory workaround.

But, telling Tex and everyone else, what PCLinuxOS ought to do, is a bit presumptuous, don't you think? If your experience was the norm, Tex would not have defaulted to the bfs kernels.

Galen