To: djgpp AT delorie DOT com
Subject: Re: speeding up command.com
Message-ID: <19970203.183343.4575.5.chambersb@juno.com>
References: <s2f5dd7c DOT 003 AT bielbit DOT bielsko DOT pl>
From: chambersb AT juno DOT com (Benjamin D Chambers)
Date: Mon, 03 Feb 1997 21:32:46 EST


On Mon, 03 Feb 1997 12:42:47 +0100 Bartosz Polednia
<bartosz AT bielbit DOT bielsko DOT pl> writes:
>> Question:
>> OpenDOS is for x86 processors, right?
>> THEN WHO THE HELL NEEDS IT TO BE PORTABLE?????????
>> You're not going to break things by inserting a little ASM code - >
>> it's not like it'll run on the Mac anyways :) :) :)
>> And who's going to be running the command.com from Linux???  Go
>> ahead and make things fast!
>
>Hi,
>
>#1.
>  What kind video card do you want to implement direct writes for ?
>     If it is VGA, what about all Hercules, EGA, CGA users ?

All depends.  See below.

>
>#2.
>  Maybe we can use some 386, 486, 586 or MMX specyfic opcodes inside
>Open DOS. Executables will be smaller and faster but all machines
>with less than (386...MMX) could be dropped into trash.

Fair enough, that's a very good question.  Now:
Why must the code be included in command.com?
Why must the code not be included in run-time loadable objects?

I don't know if it would work, but I believe the DLM project has
something _similar_ (though it would probably need to be modified to work
in command.com)
I unfortunately can't find the URL right now (sigh :( ),
but someone on this list should have it.

...Chambers