-----BEGIN PGP SIGNED MESSAGE-----
I know exactly how you feel.. I get so anoyed with all the hardware
reviews on the web that are all about speed, squeezing a few more
bytes/sec out of the hardware. but I never see stability reports.. all
the windows users just shrug off the crashes as windows problem.. when it
all comes down to it.. it's a hardware bug 50% of the time. (for bad
box-killing crashes) some of these problems are so wierd, I almost start
to blame _linux_ for crahses, when it's actualy hardware. (not being able
to see kernel pannic's in GUI mode can be really frustrating)
what I've done over the last few years with PC hardware, was to stop
buying all the really cheap generic PC hardware, and to stick with more
thuroughly tested systems. the only non-fun thing about this, is you
loose some of the nice things like standard motherboards, and
latest/greatest CPU speeds. and video hardware. frankly, most of the
stuff is fast enough for what I do these days.
I love dell optiplex systems for general user use, and if you need
preformance and stability.. check out things like the precision
workstation. it's about all you can do to guarntee stability in any kind
of quantity
Thank You,
Ben Kochie (ben at nerp.net)
"Unix is user friendly, Its just picky about its friends."
On Tue, 15 May 2001, Mike Hicks wrote:
> Ben Kochie <ben at nerp.net> wrote:
> >
> > I agree.. the x86 hardware needs a re-vamp. and hopefully the 64bit
> > platforms will get most of that fixed.
>
> Heh, I think that the Pentium was actually supposed to drop the legacy x86
> support. The same was supposed to be true for their 64-bit chips, though
> I thought I heard that they are going to be x86 compatible as well.
>
>
> Unfortunately, I think that Intel, Microsoft, or someone has been rubbing
> off on other companies who have been putting out consumer hardware. Maybe
> I just wasn't paying attention before, but there seem to be a lot of buggy
> chipsets out there. I think there are some manufacturers who believe that
> `consumer-grade' means that there can be four times as many bugs when it
> gets shipped.
>
> That's a bad mode of thought, in my opinion, as many consumers can really
> push their hardware. Conversely, look into any server room and you're
> probably going to find at least a handful of PCs. This means that some
> really important stuff could be depending on potentially buggy
> `consumer-grade' hardware.
>
> Part of my reason for looking at different architectures is to get away
> from that mode of thought. I'd be willing to pay 2x or 3x the `consumer'
> price in order to get hardware that has been more fully tested, and is
> therefore less likely to completely toast my data. Unfortunately, it
> appears that sellers want me to pay even more (4x-10x or more) to get such
> a level of quality. Who am I? NASA?
>
> --
> _ _ _ _ _ ___ _ _ _ ___ _ _ __ I'll listen to reason
> / \/ \(_)| ' // ._\ / - \(_)/ ./| ' /(__ when it comes out on CD
> \_||_/|_||_|_\\___/ \_-_/|_|\__\|_|_\ __)
> [ Mike Hicks | http://umn.edu/~hick0088/ | mailto:hick0088 at tc.umn.edu ]
> _______________________________________________
> tclug-list mailing list
> tclug-list at mn-linux.org
> https://mailman.mn-linux.org/mailman/listinfo/tclug-list
>
-----BEGIN PGP SIGNATURE-----
Version: 2.6.3ia
Charset: noconv
iQCVAwUBOwHwx8tpDhsSpvgtAQGQmgP+Jd8/BKSV66wwtKX/1RPq3WrDUynd+2gZ
1HAty1iNUa9pEHhvmYHiDvVEI4bw07KqhTskAiOPMt3ZoAmh7hnJgXx00YYFvGNS
fl9rda5gNlKK5LldvGvcT9DExurm9NAKAiYTeDkAbf5/Y4ChR1mE8GwCgH6EyIPL
6Ca+En4s45Q=
=E4jj
-----END PGP SIGNATURE-----