From: "Charles Sandmann" <sandmann AT clio DOT rice DOT edu>
Newsgroups: comp.os.msdos.djgpp
Subject: Re: Quake contributions [was: Is Allegro too ...]
Date: Tue,  9 Mar 1999 13:03:40
Organization: Aspen Technology, Inc.
Lines: 18
Message-ID: <36e51c2c.sandmann@clio.rice.edu>
References: <199903090358 DOT WAA20971 AT envy DOT delorie DOT com>
NNTP-Posting-Host: dmcap2.aco.aspentech.com
X-NewsEditor: ED-1.5.8
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Reply-To: djgpp AT delorie DOT com

> > Would the __djgpp_nearptr_enable/disable functions have been one of
> > the results of iD's 'footprint in the sand'? What other changes did
> > they ask for?
> 
> That was the biggie.  I think the rest were bug fixes and minor
> features, but I wasn't the one working directly with them.

Actually, Quake was delivered using V2.00 beta 3 - which didn't support
the nearptr as we know it today.  The problem was that in beta 3 that
non-move sbrk didn't exist, so they used the unixy sbrk and a prototype
for the nearptr stuff.  

There were some interesting things written to work around W95 DPMI bugs
(especially DMA and locking memory) - but those routines were not part
of the standard DJGPP libraries.

I know that the DJGPP V2.0 (and later) releases wouldn't work properly with
the source written for beta 3 due to changes in sbrk algorithms, etc.