🚀 go-pugleaf

RetroBBS NetNews Server

Inspired by RockSolid Light RIP Retro Guy

Thread View: comp.os.msdos.programmer
3 messages
3 total messages Started by Ralf Brown Tue, 17 Sep 1996 00:00
Re: Divide Overflow ??????
#8119
Author: Ralf Brown
Date: Tue, 17 Sep 1996 00:00
38 lines
2858 bytes
}ric@radmic.com asked:
}:       I have an old game that I got when I had a 386SX/25 w/ dos 6.  It ran 
}: good and was a good game.  I haven't played it for years and thought that I'd 
}: give it a whirl.
}:       I now have a Pentium 90 w/ dos 6.22.  When trying to run it, and I 
}: don't even get to the start screen.  Fact is, I don't get to any screen.  What I get 
}: is a message that says "Divide overflow" and then the dos prompt.  

In article <51hbnv$qqo@csugrad.cs.vt.edu>,
dcraigg@csugrad.cs.vt.edu (Danny Craigg) replied:
}This is pure conjecture, and thus may be completely off base.  That said, is
}it possible you have one of the buggy Pentium CPU's?  I'm probably way off
}base, so someone correct me if I'm wrong.

In article <51hvja$l8p@goanna.cs.rmit.edu.au>,
pfg@yallara.cs.rmit.EDU.AU (Paul Francis Gilbert) replied:
}This a very old problem with DOS itself. Some games were written in the
}pre UMB days, when everything, including drivers was loaded in the main
}640Kb. Now, with the new managers for HMA, it isn't always that way, so
}the old programs can spit the dummy... giving an erraneous "Divide" error.

You're both wrong.  It's neither a buggy CPU nor the EXEPACK "Packed file
is corrupt" bug.  The Pentium-90 is simply too fast for the program!  Since
games need to run at the same speed no matter how fast the CPU, they include
a speed-detection routine at startup.  In the present case, the program is
clearly counting the number of timer ticks it takes to run a fixed number
of iterations of some action.  On the P5-90, it is finishing in less than
one timer tick, and thus trying to divide by zero.  If it didn't bomb out,
it would be unplayably fast....

The solution: try putting the Pentium into "slow" or "deturbo" mode.

-- 
mailto:ralf@pobox.com -=- http://www.pobox.com/~ralf -=- #include <disclaimer>
"The Internet is like a freight train roaring along while people are laying
 tracks in front of it.  It's not just gaining on those laying tracks; it's
 gaining on the steel mills."  -- Matt Mathis
Re: Divide Overflow ??????
#8120
Author: Charles Dye
Date: Sat, 21 Sep 1996 00:00
22 lines
1841 bytes
In article <323d5d83@ralf> Ralf Brown <ralf@telerama.lm.com> writes:

>You're both wrong.  It's neither a buggy CPU nor the EXEPACK "Packed file
>is corrupt" bug.  The Pentium-90 is simply too fast for the program!  Since
>games need to run at the same speed no matter how fast the CPU, they include
>a speed-detection routine at startup.  In the present case, the program is
>clearly counting the number of timer ticks it takes to run a fixed number
>of iterations of some action.  On the P5-90, it is finishing in less than
>one timer tick, and thus trying to divide by zero.  If it didn't bomb out,
>it would be unplayably fast....

>The solution: try putting the Pentium into "slow" or "deturbo" mode.

If your Pentium, like many, doesn't have a "turbo" button, you can fake
it by going into the CMOS setup program and disabling any cache.  Don't
forget to re-enable it when you're done with the game, though!

There are also freeware programs out there that enable/disable the cache
from the command line.

raster@indirect.com
Re: Divide Overflow ??????
#8121
Author: Charles Dye
Date: Sat, 21 Sep 1996 00:00
24 lines
1892 bytes
Reposting article removed by rogue canceller.

In article <323d5d83@ralf> Ralf Brown <ralf@telerama.lm.com> writes:

>You're both wrong.  It's neither a buggy CPU nor the EXEPACK "Packed file
>is corrupt" bug.  The Pentium-90 is simply too fast for the program!  Since
>games need to run at the same speed no matter how fast the CPU, they include
>a speed-detection routine at startup.  In the present case, the program is
>clearly counting the number of timer ticks it takes to run a fixed number
>of iterations of some action.  On the P5-90, it is finishing in less than
>one timer tick, and thus trying to divide by zero.  If it didn't bomb out,
>it would be unplayably fast....

>The solution: try putting the Pentium into "slow" or "deturbo" mode.

If your Pentium, like many, doesn't have a "turbo" button, you can fake
it by going into the CMOS setup program and disabling any cache.  Don't
forget to re-enable it when you're done with the game, though!

There are also freeware programs out there that enable/disable the cache
from the command line.

raster@indirect.com
Thread Navigation

This is a paginated view of messages in the thread with full content displayed inline.

Messages are displayed in chronological order, with the original post highlighted in green.

Use pagination controls to navigate through all messages in large threads.

Back to All Threads