info-cpm (12/01/82)
>From mwm@Okc-Unix Wed Dec 1 00:34:33 1982
To: info-cpm@Okc-Unix
Cc: mwm@Okc-Unix
Via: Okc-Unix; 30 Nov 82 23:25-EST
Via: Brl; 30 Nov 82 23:38-EST
Via: Brl-Bmd; 30 Nov 82 23:46-EST
About 6 months ago, I ran into a (rather nasty) MPM 2.0 bug. While discussing
DRI's software support with someone, this came up again. Somebody suggested
posting it here, both as a warning, and on the off chance that sombody
knows of a fix.
The problem is that MPM does NOT reinitialize page zero in a when a process
exits, unless page zero of the partition is also page zero for a bank.
This is a problem if you have a program that twists things in page zero. Any
of the `CCP replacements' (microshell, etc) or a menu-driven system that
uses program chainging (Magic Wand II) will trip over this. The sympton is
that, after running such a program, you get the thing back the very next time
something exits, IF you took a hard (^C or whatever) exit.
When I talked to DRI about it, they said `send us a disk with software that
displays the problem, and we may look at it.' Just one of the many reasons
that I hate DRI.
<mike