Bill Degnan wrote:
Thanks Dave - It has been many years since I genned a
N* disk. I don't
always have success with the port assigning but eventually I get it to work.
Although my Altair was "fully expanded" when I first owned it, I was involved
with it quite some time before that:
At University of New Brunswick I made friend: Gary -staff at computer center
Gary found: Dan -former student -built the Altair -Decided to sell to Gary
-- Dan had assembled a minimal system: Chassis, all Mits: CPU, 1K RAM, Serial
-- Fortunately he had put in 3 of the 4-slot S100 boards (12 slots)
Together Gary and I expanded it to a much more usable system
--- 64K RAM, NorthStar Disks, ADM3A terminal and more
I wrote an 8080 assembler on UNBs IBM 360 mainframe, which gave a good listing
(code values) of our I/O routines. Using front panel: Run at E900 (Boot ROM -
unpersonalized N*DOS), HALT, toggle-in I/O code, restart DOS in memory.
This gave DOS talking to the ADM3A (yes - it took a few trys before we had it
working) - then we loaded clean DOS into memory (DOS self-modifies in memory
as it boots), use N*Monitor to copy our I/O to this copy, initialize
a second disk and save DOS to it - this gave us a working DOS that booted!
If you want to experience doing this, it can all be done on my Virtual Altair!
For the true experience, you should stick to the front panel switches, but if
you want to "cheat" - Virtual Altair has additional capabilities that original
didn't: Full debugger with execution (disassembly) display, memory/register
editors, 'L'oad Intel/Motorola hex images to memory while stopped!
Sometime later, while testing bare code on a second Altair without disks, I
created: MinimalCodeEntrySerialBoot
This let me load/run large programs, having to toggle in only 18 bytes via
the front panel!
Ah, fond memories!
-Dave