I have an old Yamaha cdr100 with a very small buffer. (512k). When
recording at 4x that gives me a bit less than 1.5 seconds of buffer
space. Used to be that a user starting Netscape would cause buffer
underflows on my 200MHZ workstation. The cost of a CD is only about a
buck at the moment, but after the third one, you are ready to tear your
hair out. You've wasted a lot of TIME. This is clearly classified as a
'failure' by the user.
With the advent of the POSIX 'soft' real time scheduling in the 2.0
kernel I can now reliably burn CD's at load factors of 2 with much
larger 'bursts' of intensive activities. Haven't messed one up in
ages.
In the end that is all the user is going to care about. If they can't
get the job done they can't use the tool. If you tell them that if they
want to burn CD's they need to use RTlinux, they will go back to
something else simpler in most cases.
Sure, there are limits to what you can do in the standard kernel, but
since the POSIX stuff is there, I'm glad to see that consideration is
being given to keeping the stock kernel capable of supporting many real
time applications with 'medium' tight timing requirements.
I think Linux needs to concentrate on what it has and continue to improve
where it doesn't negatively impact other system aspects. I've seen a
few posts that imply that ANY real time considerations do not apply in
that developers view.
As I understand it, performance decisions tend to favor the interactive
user in Linux. Nevertheless, we should keep in mind that the
interactive user may want to run his CD-ROM burner/etc..
-- +--------------------------------------------+ Rick Bressler |Mushrooms and other fungi have several | |important roles in nature. They help things| |grow, they are a source of food, they | bressler@mushroom.ca.boeing.com |decompose organic matter and they | |infect, debilitate and kill organisms. | Linux: Because a PC is a +--------------------------------------------+ terrible thing to waste.- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/