Re-compiling old DOS

From: Richard A. Cini, Jr. <rcini_at_email.msn.com>
Date: Thu Jul 16 18:54:41 1998

    One of my recent time sink-hole projects is de-compiling DOS 1.1. I
figure that I needed a computing challenge :)

    Anyway, this plus a message from one of our own, results in an
interesting question. How, in 1982 (or in 1980-81, when Project Acorn was in
beta), would Microsoft have written and compiled DOS 1?

    I know that DOS as we know it was purchased from Tim Patterson of
Seattle Computing Products (Is this the right name? I don't have a copy of
one of my many Gates books handy.), so the source code in printed, magnetic,
or paper tape form would have existed. Would they have used a PC simulator
running on another computer (the same way they wrote BASIC for the Altair)?
Would they have waited to receive a beta unit to test/develop it live after
months of writing blind code based on IBM's specifications documents?

    Anyway, I digress. I've pretty much finished the boot sector code, and
IBMBIO is done. So, now I'm working on IBMDOS. Does anyone know anything
about how early DOS versions loaded? Looking at the DOS1 code and comparing
it to DOS6 code, it seems like the DOS1 boot sector loads both BIO and DOS
back-to-back in memory (producing a single image, if you will), and then
jumps to BIO. DOS6, in comparison, the boot sector loads IO, which then
loads DOS.

    Any thoughts? I felt this qualifies as "classic" since the file dates
are May 7, 1982 <g>

Rich Cini/WUGNET <nospam_rcini_at_msn.com>
  - ClubWin! Charter Member
  - MCP Windows 95/Windows Networking
  - Preserver of "classic" computers
<<<< ========== reply separator ========== >>>>>
Received on Thu Jul 16 1998 - 18:54:41 BST