[ale] kernel numbering

[Barry Hawkins]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Jonathan Rickman wrote:
| On Fri, 11 Mar 2005 22:35:00 -0500, Barry Hawkins <barry at alltc.com> wrote:
|>and/or functionality for your hardware.  Plus, with the robustness of
|>today's bootloaders it's not like you couldn't boot into your previous
|>"sure thing" kernel if you have problems.
|
| This seems to be the core of the problem. Apparently for you, Linux is
| just a fun hobby. For others, it is a critical component of their
| business...and a fun hobby. Your tolerance level for bugs is higher
| because the consequences of a bad build are not as serious.
[...]
Wrong guess.  All my development work, business data, and personal data
are on Linux.  The only reason I run later kernels is that when I have
tested them, they have worked, and I have not had to revert.  2.6.9 was
an exception to that, however.  My two development servers that are x86
are on 2.6.8 kernels, and they have been running fine.  I stay with the
stock Debian kernels there, because they serve my needs without
requiring effort on my part.

~ Linux
| started out as a research project that became a stable and trustworthy
| solution. Now it appears to be turning into a research project again
| at the expense of stability and trustworthiness.
My experience has been that most commercial vendors allow their paying
customer base to license research projects and pay for their support.
They usually carry version numbers like x.0, get fixed on the vendor's
time schedule, and customers don't have a great deal of input into the
process.

[...]
| The stability problems with
| the newer 2.6.x kernels are very real.
Most of the complaints I have heard about are issues most critical with
desktop-centric elements, like video cards and sound.  Have you been
experiencing issue with things like SCSI controllers, RAID arrays, fiber
NICs, etc.?  I have heard no specific complaints in this thread other
than being unhappy with the pace at which the kernel development is
moving and the maturity and stability of releases.

| Companies like Red Hat, Novell,
| IBM, etc have significant investments in Linux and will not hesitate
| to fork the project if customers start jumping ship and they feel that
| investment is in jeopardy.
That's certainly their prerogative.  I would say that it would then
immediately begin to become stale and very biased toward their hardware
offerings and/or alliances, unless they staff a fairly large and
benevolent kernel development team.  Thank God it's GPLed or they
probably would have forked and stagnated long ago.

[...]
| Much as I love Linux, specifically Slackware, lately I find
| myself looking at my Solaris 10 machine and trying to come up with one
| single reason to continue to use Linux. If Slackware goes under for
| some reason, that'll probably be it for me.
It's not a "right or wrong" thing; please don't take my previous post
that way.  The freedom of open source comes at a price, part of which is
a degree of risk and a rate of change that can be uncomfortable.
Depending upon needs and available hardware, Linux can simply be an
unworkable option.  It has been for me at times in the past.

Regards,
- --
Barry Hawkins
All Things Computed
site: www.alltc.com
weblog: www.yepthatsme.com

Registered Linux User #368650

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (GNU/Linux)

iD8DBQFCM0Fw7bZ6kUftWZwRAjJ4AKCdjSF3GUasvFJZPyd4GUMD2JSy/QCghX6H
z+z3hwynjF/p9jnFta6ZsoY=
=jb7y
-----END PGP SIGNATURE-----