I've got a
3?" SCSI floppy drive (yes, a floppy drive with a SCSI
interface!) that reports itself as having 256-byte blocks.
Does it have any
provision to reconfigure?
Not that I see. It's possible frobbing the right value in the right
mode page, or passing the right arguments to a FORMAT UNIT operation,
or some such, might change it. It has a bank of ten jumpers, but they
are not labeled very usefully - the only documentation I see on the
thing (the silk-screening on the visible side of the board) labels
them, in order, EJC, J, H, G, F, LEV, HDS, STL, PAR, MON.
It appears to be an interface board in front of a more or less ordinary
floppy drive, though I suspect appearances are deceptive, since I don't
see any power feed to the front-end board except possibly over what
looks like an ordinary floppy interface - but an ordinary floppy
interface does not supply power from the drive to the ribbon cable.
At first I thought there was no model number marking, but on careful
inspection I do find a sticker indicating it's a TEAC FD-235HS.
ftp.rodents-montreal.org:/mouse/misc/scsi-floppy.jpg is a photo I just
now took of it; you can see the ten jumpers at the bottom left of the
front-end board (the easily visible one). Of the three 2xN headers at
the right, the 50-pin one at the far right is the SCSI interface; the
34-pin one, the lower of the left two, is the "ordinary floppy
interface" from above. The 22-pin one above it is just a socket on the
other side; I conjecture it's for SCSI termination resistor packs. At
the far left you can see a sticker (the one with the T?V and CSA and
assorted other logos); the yellow patch on that sticker, just before
FD-235HS, is the TEAC name in silver-on-yellow.
[...]; the OS
support for SCSI I have is broken enough on devices
with non-512-byte sectors that [...]
Prob'ly because the guy who wrote that
code had just read:
>>> "From the time of their first availability in the 1950s until
>>> about 2010, the sector size on disks has been 512 bytes."
Ha. That would entail some interesting causality tangles, since the
SCSI support in question goes back at least a decade.
I don't know why it was written that way. I suspect it was a case of
supporting the common case first, then never getting around to going
back and dealing with the uncommon cases. Then, after a few years of
accretion of code that blindly assumes everything uses the same sector
size, changing it became a major undertaking.
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at
rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
