To: opendos AT delorie DOT com, crough45 AT amc DOT de
References: <97Sep11.091906gmt+0100 DOT 11657 AT internet01 DOT amc DOT de>
Message-Id: <ADes56qqfH@belous.munic.msk.su>
From: "Arkady V.Belousov" <ark AT belous DOT munic DOT msk DOT su>
Date: Fri, 12 Sep 1997 01:07:52 +0400 (MSD)
Organization: Locus
Reply-To: ark AT mos DOT ru
Subject: Re: ClosedDOS???
Lines: 46
MIME-Version: 1.0
Content-Type: text/plain; charset=koi8-r
Content-Transfer-Encoding: 8bit
Precedence: bulk

X-Comment-To: Christopher Croughton

Hi!

11-σΕΞ-97 08:24 crough45 AT amc DOT de (Christopher Croughton) wrote to ark AT belous DOT munic DOT msk DOT su, opendos AT delorie DOT com:
 > Arkady V.Belousov wrote:

 > >  > And what's wrong with DOS extender if it is free?
 > >      You think - this is wise? DOS utilities and kernel itself with external
 > > DOS extender? Only for 386+? How many memory additionaly this requires? How
 > > many time to initialize this requires? How many stability and compatability
 > > this add?
 > DJGPP does not have an external DOS extender, necessarily, it is
 > normally bound into the utilities.

     I, may be, not correctly express my mean. "External" there I understand
as additional _pretentious_ module, even they built-in.

 > All they need is a DPMI server (one that works!).

     Ha!

 > Yes, they are 386+, which means that one of the
 > main target areas for OpenDOS (low-end 8086 and 286 machines) is
 > not possible.

     And what I win by this for small and not pretentious DOS utilities
(except, may be, some like "sort")? Speed? Size? Compatability?

 > >  > BTW i always wonder *why* there is no GNU compiler for 16-bit DOS..
 > >      Some think, this because no one (!), capable to port gcc, interesting
 > > by this. :(
 > Plenty of people are capable, no-one is interested in putting in the work.
 > If someone wanted to pay me for doing it I'd be willing, but it would
 > have to be at my full commercial rate because I don't have time to do
[...]
 > The sources are available, if anyone who really wants a 16-bit version
 > wants to port it they can do it.  If they either don't want to make the
 > effort themselves, or if they can't do it and aren't willing to reimburse
 > someone to do it for them, they should stop complaining.  (Or in other
 > words, "put up or shut up".)

     I, may be, can - but this requires from me (as and for you) additional
work to generate machine-definition, but _I_ prefer to test and enchance
someone else's software, and up to now my requirements be satisfacted by
existing compilers like BCC.