Re: Software for MS-DOS 1.25

Den Fri, 13 Oct 2017 13:43:01 +0200 skrev Michał Pleban
<lists@michau.name>:
> No, it is actually up to 8088 to do the translation. The 6509 function
> expects to receive the exact track and sector to be read. So the 8088
> must translate PC sectors into 8250 sectors. It is a bit tricky to do
> the translation (there are fewer tracks but more sectors per track,
> also the number of sectors per track varies) but I don't see anything
> here that couldn't be done.

Well, the 6509 code can be changed, but with tight space for the code
it might be best to let the 8088 do that job.

> Another thing is that on the PC sectors have 512 bytes, so each sector
> read must be translated to two reads from the 8250, possibly with some
> interleave to do it faster.

If the 8250 and 8050 can do read from disk to internal buffers and from
buffers over IEEE as separate commands (like 1541 can), then it might
be best to tell the drive to read two 256 byte sectors at a time and
read them back to the computer with memory read commands?

> At this point, MS-DOS 1.25 thinks that the 8250 is a 160 kB floppy ;-)

That must be so it would work on a 2031, 3040, 4040 or similar drive.

I would try FDREAD and FDFORMAT and see if you can use larger disk
space :)

This is if FDREAD and FDFORMAT even works on DOS 1.25

On a real PC, DOS 1.x uses 8 sectors per track, and DOS >= 2.0 use 9
sectors per track.

Anyway as the BIOS would be loaded from disk, the correct data for the
disk drive would anyway be loaded from the disk drive. If this project
leads to downloadable images they could be ready configured for the
correct drive depending on which image they are (.d64 , .d80 or .d82).


-- 
(\_/) Copy the bunny to your mails to help
(O.o) him achieve world domination.
(> <) Come join the dark side.
/_|_\ We have cookies.

       Message was sent through the cbm-hackers mailing list