Xref:	news2.mv.net comp.os.msdos.djgpp:1064
From:	Charles Sandmann <sandmann AT clio DOT rice DOT edu>
Newsgroups:	comp.os.msdos.djgpp
Subject:	Re: Infernal compiler error: CWSDPMI
Date:	Mon, 12 Feb 1996 22:36:50 CST
Organization:	Rice University, Houston, Texas
Lines:	25
Message-ID:	<31201562.sandmann@clio.rice.edu>
References:	<271E42B1417 AT fs2 DOT mt DOT umist DOT ac DOT uk>
Reply-To:	sandmann AT clio DOT rice DOT edu
NNTP-Posting-Host:	clio.rice.edu
To:	djgpp AT delorie DOT com
DJ-Gateway:	from newsgroup comp.os.msdos.djgpp

>   How common is this `Cooperative Wholesale Society DPMI' (or whatever the
> letters mean)? I've never had that trouble with the old djgpp.

DPMI is pervasive in protected mode 32-bit environments (you can't get to
32-bit protected mode without it in OS/2, Windows, NT, etc) so you are 
forced to deal with it, since 100% of new computers get shipped with it
as the default environment.

If Microsoft had made it just a tiny bit more pervasive and shipped a DPMI
provider with EMM386 (like all of the other memory manager do) we wouldn't
have to worry about CWSDPMI at all (which is the C.W.Sandmann provided
DPMI package).  And if DPMI hadn't stopped developing before it even 
reached a full 1.0 (yes, MS is still shipping 0.9, 5 years later) it would
be a decent API.  Null pointer protection in V2 is due to a DPMI 1.0
extension built into CWSDPMI - if I had infinite time and infinite $$$
I would code the full 1.0 specification in CWSDPMI and we could have 
dynamic stack expansion, simulated linear frame buffers, and the whole
DJGPP v1.x feature set in V2.x.

Noone is forcing you to migrate to DJGPP V2, V1.12 is still out there,
with full source, so you can continue to use/upgrade/merge_V2_features as
you see fit.  If you or none of your programs ever see any operating
system enviroment than DOS, don't use hardware interrupts, don't need
signals, don't need posix or ansi compatibility, then V1.x is probably
a better choice.

- Raw text -