🚀 go-pugleaf

In article <1993Jan6.214710.26976@schunix.uucp> sonix@schunix.uucp (Duane Morin) writes:
>Perhaps someone recognizes this problem.  My BC compile, which used to work
>just dandy, now randomly crashes with "FATAL ERROR:  GP fault" or somesuch
>similar message.  Not always on the same file, and not necessarily occurring

I've encountered the same problem with BC++ 3.1 IDE version.  I was rather
confused since BC++ 3.0 worked just fine with the same program.  

The tech guys at Borland told me that BC++ 3.1 barfs on some compile-time
errors.  The tech I spoke to recommended running prj2mak to convert the .prj
file to a make file and use the commandline version bcc.exe instead.  Turns out
he was right.  The commandline compiler gave me some error messages before
bombing out with a GP Fault.  I fixed the errors in the source, and it works
just peachy.  Odd that 3.0 doesn't do that...

Try running the commandline version and see if that gets any farther during the
compile process.  Worked for me.

-- 
Grant Boggs
boggs@a.cs.okstate.edu
Oklahoma State University
Stillwater, Oklahoma

In article <1993Jan6.214710.26976@schunix.uucp> sonix@schunix.uucp (Duane Morin) writes:
... Stuff about GP errors deleted ...
>2) Got 3.1.  Can't even install the damn thing.  ZIPs are intermittently 
>	corrupted.  Little examination shows that, with cache memory off,
>	I have no problem unzipping, with it on, sometimes they get CRC

This is fair proof that your motherboard is not functioning in a happy way.
486 motherboards are often very sensitive to the 5volt power supply voltage.
If the 5volt supply is running at the low end of tolerance, the cache
controller chips might not be able to function quickly enough.

Check to see that you have 5.00volts -0%, +5% on the motherboard's power
connector (red and black leads), when you have all of your system's
cards/disks installed.  If you aren't getting 5.00volts you should either
adjust your supply (HA HA HA, good luck!)  or get a new supply that does
give you 5.00volts.

You wouldn't have this problem if the motherboard suppliers wouldn't cheat.
What they do is buy all 25MHz 486 motherboards, without CPU or RAM.  They then
plug in the proper speed parts to get either 33MHz or 50MHz.  Well, this would
work, except that the motherboard manufacturer's grade the motherboards for
speed.  The ones that fail 50MHz become 33MHz boards, the ones that fail
33MHz become 25MHz...

Now, you ask, "Why does the board work sometimes on my system?"

Well, It's like this, the boards are designed to run at 5v +/- 5%, and from
0-70degrees C.  Logic chips such as the cache controller run their fastest
when they are cold and at the high voltage margin, and at their slowest when
they are hot, and at their low voltage margin.  Most of the time, the power
supplies will provide 5.0v on the nose, and you all will run your machines
in comfortable rooms, so you can get away with using parts that failed to
run full speed at 4.75v and 70C.  The clone builders are banking on this
happy coincedence.

Chuck Harris - WA3UQV
chuck@eng.umd.edu

|> 
|> In article <1993Jan6.214710.26976@schunix.uucp> sonix@schunix.uucp (Duane Morin) writes:
|> ... Stuff about GP errors deleted ...
|> >2) Got 3.1.  Can't even install the damn thing.  ZIPs are intermittently 
|> >	corrupted.  Little examination shows that, with cache memory off,
|> >	I have no problem unzipping, with it on, sometimes they get CRC
|> 
|> This is fair proof that your motherboard is not functioning in a happy way.
|> 486 motherboards are often very sensitive to the 5volt power supply voltage.
|> If the 5volt supply is running at the low end of tolerance, the cache
|> controller chips might not be able to function quickly enough.
|> 
|> Check to see that you have 5.00volts -0%, +5% on the motherboard's power
|> connector (red and black leads), when you have all of your system's
|> cards/disks installed.  If you aren't getting 5.00volts you should either
|> adjust your supply (HA HA HA, good luck!)  or get a new supply that does
|> give you 5.00volts.
|> 
|> You wouldn't have this problem if the motherboard suppliers wouldn't cheat.
|> What they do is buy all 25MHz 486 motherboards, without CPU or RAM.  They then
|> plug in the proper speed parts to get either 33MHz or 50MHz.  Well, this would
|> work, except that the motherboard manufacturer's grade the motherboards for
|> speed.  The ones that fail 50MHz become 33MHz boards, the ones that fail
|> 33MHz become 25MHz...
|> 
|> Now, you ask, "Why does the board work sometimes on my system?"
|> 
|> Well, It's like this, the boards are designed to run at 5v +/- 5%, and from
|> 0-70degrees C.  Logic chips such as the cache controller run their fastest
|> when they are cold and at the high voltage margin, and at their slowest when
|> they are hot, and at their low voltage margin.  Most of the time, the power
|> supplies will provide 5.0v on the nose, and you all will run your machines
|> in comfortable rooms, so you can get away with using parts that failed to
|> run full speed at 4.75v and 70C.  The clone builders are banking on this
|> happy coincedence.
|> 
|> Chuck Harris - WA3UQV
|> chuck@eng.umd.edu
|> 
|>