[muglo] Re: VPC (was- Re: OS X Upgrade)

Yipes, you've got a lot of variables in there that make it hard to trouble
shoot. I am still partial to the conclusion that it's your hardware,
especially with 4 RAM chips and a G4 550. Since you have four of 'em what
I'd do is pull one each week and see if it's one of the chips. Actually I'd
pull two or even three at a time!!! You may have a bit of a slow down b/c of
the 256 MB RAM *but* b/c of the one chip you'll be pushing the CPU harder
than normal and be able to see if it's the CPU.

PS Most G4 CPUs can't register temperature worth a damn, and, even if they
do it's only for curiosity sake... unless someone actually went to the
trouble of calibrating the CPU when it was installed the readings you get
are about as reliable as using lips on a person's forehead (actually, the
CPUs are a lot less reliable than lips since a good pair of lips can detect
a temperature to within 0.5 C and CPUs are off by up to 20 C... like when
the CPU reads *below* room temperature (which isn't possible without
evaporative cooling or refrigeration)).

on 28/7/03 12:03 AM, Alex at admeddemda@xxxxxx wrote:

> i dont know. I installed a cpu heatsink fan over the heatsink
> recently. But there's some glitch in the processor (recognized by
> powerlogix) where it cant register temperature, which sucks. I dont
> know if its too hot. Anyway, I am running a G4 550 chip from
> Powerlogix. It's less than a year old.

Age won't matter! If it was a marginal CPU to begin with it'll be a marginal
one now. As for heat, if installing an extra fan didn't alleviate the
problem it's not likely to be the CPU.

> Also, my ram is from OWC, also less than a year old. Lifetime
> warranty on the ram, and I doubt its bad.

RAM is RAM is RAM. Even the most reputable of sources have bad batches.
Kingston has a good warrantee and a good reputation but they still have
returns! Don't count your RAM out simply b/c it came from a good source. I
have a 64 MB Apple SO-DIMM that was a piece of crap and caused no shortage
of kernel panics in a 320 MB config but when I pulled it and went down to
256 all kernel panics disappeared.

I'm still partial to blaming the RAM *unless* your kernel panic logs have
*repeatable* error message. From everything that I gather *random* panics
are more often than not symptomatic of bad RAM.

> yes, that. The instructions on screen and also sometimes freezes
> where the mouse moves but clciking anything does not work.

Mouse moving but not clicking is possibly not a 'real' crash -- i.e. if you
were knowledgeable you could re-launch the GUI from a remote terminal. If
you can still login to your computer with ssh then it's just a GUI crash (as
opposed to a fault in the kernel... i.e. kernel panic).

> I dont worry abt the apps crashing, i just restart them. A bit
> annoying, but ok. But I have crashes every few days. And white, yes a
> problem,. And my monitor is already super dull... the brightness is
> at 0% and the contrast at 30% if I go any duller I wont be seeing
> anything. That's why I cant use an LCD... toooooo bright even at the
> lowest setting. I always liked passive matrix laptop screens... much
> duller!

You'd hate my monitors and LCDs. I have brightness usually maxed out and
absolutely *love* LCDs (they're so easy on the eyes).

Do you compute in complete darkness? If I had brightness set to 0% my
monitors would be black.

> Well, I think it is software... why? bc I have a whole bunch of
> weirdo hacks installed in the OS, and also things like PCIextreme,
> which I know has made things more unstable.

I guess in that case you shouldn't be claiming that OS X is unstable ;). I
have a question for you: "why?" (especially PCIextreme)? Yeah, it'd be neat
to have stuff like WindowShade operating, or fiddle with the command-tab
behaviour but why make your system unstable? I am not joking when I say I
*never* have a crash -- I'm not about to mess with a beautiful formula (so,
instead of applying hacks I push my computer in other ways).

Speaking of pushing in other ways... I had some fun watching an OS X newbie
setting up his PB 867 (*nice* machine... much nicer than the towers...
especially the fact that the 15" LCD is as big as a normal 17" CRT): he kept
working in the old modality. Set something in motion and *wait* for it to be
done before he moved on to something else. After I pointed out to him the
fact that he could actually multi-task he started doing other stuff :) :)
:).

> And, I've always felt upgrade cpu chips were more unstable than Apple ones.

Not necessarily true. Where a lot of upgraded computers get their
instability from is the reality that the upgraded CPU pushes the mobo
components much harder than their design specs rate them for.

> But really, this crashiness happens most frequently when I'm using
> Eudora. Or maybe only when I'm using Eudora.... too bad I love
> Eudora...

Hmm. You may want to check out your kernel panic logs. Do they have a
consistent theme to them (&, yes IIRC Eudora is a buggy app). Perhaps what
you might want to try is to use Eudora for Classic instead and see if that
alleviates your kernel panics? (unless you're hooked on you haxies there's
no need to be destabilising your computer that way).

PS Are you actually running "Haxies" or are these "hacks" and "patches"
other stuff (like ASM?). Haxies are a very specific form of kernel patch.

Eric.


_________________________________________________

For information concerning the MUGLO List just click on

           http://muglo.on.ca/pages/members.html#Joinmuglo

Don't forget to periodically check our web site at:

           http://muglo.on.ca/

Other related posts: