On 4/21/23 18:15, Brian K. White wrote:
## unk23 is what the old dlplus called "ts-dos mystery command"
## it's unknown what it *really* means, but it's a reliable way to
tell the difference between TPDD1 and TPDD2 without locking up the
drive by giving an invalid command to the wrong drive.
## when you issue a certain request, the real drive gives back this
response, and it's always the same response for all TPDD1 drives.
## so if you issue the request and get the expected response, the
drive is a TPDD1, otherwise it's a TPDD2, and in either case it
doesn't lock up the drive.
another correction, right idea just backwards tpdd1 vs tpdd2
# TS-DOS mystery TPDD2 detection - some versions of TS-DOS send this
# TPDD2 gives this response. TPDD1 does not respond.
# request: 5A 5A 23 00 DC
# return : 14 0F 41 10 01 00 50 05 00 02 00 28 00 E1 00 00 00 2A
There is also a similar 0x11 command where like 0x23, TPDD2 responds
with something and TPDD1 does not respond (but also does not get freaked
out and lock up)
Client sends 0x11 (0x5A 0x5A 0x11 0x00 0x##) ## = checksum
TPDD2 responds: 3A 06 80 13 05 00 10 E1 36
TPDD1 does not respond
and 0x33 is apparently a synonym for 0x11
if the client sends an 0x33 request, TPDD2 responds the same way as for 0x11
No version of TS-DOS uses 0x11 or 0x33 that I have found, I found it by
just feeding all possible commands to a drive to see what it did.
Not all versions of TS-DOS use 0x23 either.
I don't know if all tpdd2 drives respond with the same value like they
do for 0x23
dlplus emulates all of these, which tell TS-DOS it can use some
TPDD2-only commands
or can be told specifically *not* to, to more accurately emulate a
TPDD1, which allows some TPDD1-only client software to work that would
otherwise get screwed up by unexpected responses because the client
software was from before TPDD2 or desklink existed.
--
bkw
