Re: AWGS 1.1 (was: general ? on 2gs)

[Rubywand <rubywand@swbell.net>]

Pim Blokland writes ...
> 
> Rubywand wrote:
> 
> >      Many users have IIgs's which include power-sucking plug-ins. If
> > they have not taken the usual steps to eliminate noise on the power
> > rails, their machine is virtually certain to have noise glitces which
> > can mess up memory, interfer with processing of interrupts, etc.. With a
> > large system like AWGS, the odds of hitting messed up memory or having
> > an interrupt bomb are greater.
> 
> I'm not at all sure. I use AWGS on my real GS and on my emulator on the
> Mac, and on both machines AWGS exhibits the same kind of problems.
> Therefore I don't think hardware problems such as these could cause
> them.

     Your experience is a pretty good indication of some significant
problem with AWGS software.  


> Besides, when such a "noise glitch" occurs, wouldn't the machine
> respond with a "sliding apple" error?

     Not necessarily. 'Random' changes in memory and/or registers can
produce many outcomes from no effect or harmless to display corruption,
hanging, and damaged files. These errors may not detected as fatal to
system integrity.  


> I really do think it's a software-only problem.

     The original question was 'How come so many users have major AWGS
problems while so many others do not?' So, part of the answer is certain
to be hardware glitching on some systems. The reason this needs to be
considered is that, otherwise, you have unexplainable failures vs.
successes on systems with virtually identical software configurations.

     Granted, the above does not eliminate 'software' as the big
problem. We have your own experience and many other error reports. The
indication is that something is seriously wrong with AWGS as it runs
under GS/OS with some software/firmware setups. 

     

> Starting up with no
> Inits/DAs could be a way to find out if the problems persist.
>

     That plus adding Inits/DA's one by one would be an approach which
could help pinpoint problems.

 
> >      Does the presence of a RamFAST increase the probability of an AWGS
> > crash? The answer is probably "Yes". But, then, it would be the same for
> > any hard disk interface. Cards like Second Sight and PC Transporter
> > should be even worse. Any large complex software system which depends
> > heavily upon OS features is going to be more crash-prone as OS-user
> > hardware/firmware/software complexity increases.
> 
> Hmmm... maybe if we all worked together, we could build a list of what
> problems AWGS has on what machines. If we all would write a message with
> what hardware and DAs etc we have, and what goes wrong with which module
> of AWGS, perhaps we can pinpoint some of the bugs and think of ways to
> circumvent them.
>

     Neat idea! 



Rubywand