[Coco] Any resolution to patching a Multipack for Coco 3?

KnudsenMJ at aol.com KnudsenMJ at aol.com
Fri May 7 12:58:37 EDT 2004


In a message dated 5/7/04 1:24:14 AM Eastern Daylight Time, 
johnguin at hotmail.com writes:

> The group that said it needed to be done suggested that slow degeneration of
>  the computer (the GIME chip?) was possible if you used an unpatched
>  multipack.
>  
>  Now that it's been 10+ years, has anyone experienced any faults?  Are the
>  parts still available to patch the 26-3124s?  Does this need to be done?

Since I have always used only an original gray-case MPI, with the "patch" 
(for that model, a replacement PAL chip, no soldering required), I can't really 
say for sure whether failure to upgrade it would damage the Coco3 or not.

However, I do remember the discussions.  The problem is that the unpatched 
MPI will respond to some addresses that are now used only by the GIME chip.  
(This is an objective fact that should be verified, once and for all).

After many years of professional hardware design, I can tell you that having 
more than one device trying to drive onto a bus at the same time is Not A Good 
Thing, and even if it doesn't cause run-time errors (because the good guy 
always wins), it will prematurely wear down both devices due to the current 
spikes.

As for "can I get a witness", I've never heard anyone say that he really did 
break his Coco or peripherals by continuing to run with an unpatched MPI.  
That may just be because everyone who goes so far as to belong to these Lists is 
also the kind of person who upgrades his gear as needed, heeding the warnings.

I'm trying not to take sides here, but I guess it's clear which side I'm on 
:-)

With all the connector contacts, marginal timings, and questionable firmware 
in a Coco system, I believe in not tempting the fates with any more sources of 
glitches.
--Mike K.



More information about the Coco mailing list