Karl Heinz Kremer wrote:
> I would guess that
> ever other scanner with an IEEE-1394 interface would also work
> as long as SANE supports the SCSI version of the scanner.
>
In my case there is no SCSI equivalent scanner [*]. It is true that the scanner may use a SCSI chipset somewhere, and I'll try to find the spec and all that, but these are newly developed scanners and I don't see why they would use the SCSI control set.
I have not yet had a close look at the SANE back-end, but I take it from the responses so far that
1. If the scanner uses the SCSI command set, then I should be able to hack the sbp2-ieee1394 driver and get the commands going. Worst case I'll have to write a new SCSI-like backend (not nice, but at least plenty of good examples to start from).
2. If the scanner does not use the SCSI command set then I'm, eh, set for more of a challenge? How coupled are the backend architectures to the underlying protocol (SCSI, USB, Parallel)? It seems I'll have to write a new backend with (a) a new command set and (b) a new underlying protocol. [This assumes I can get the specs, of course.] What's the ratio of complexity of (a) to (b)?
[I guess the scanner may use an USB-like interface? It seems to me that FireWire is more like USB (like, serial for one) than SCSI.]
Thanks for all the answers so far. I really appreciate it.
Allan.
[*] Actually, come to think of it, perhaps this is not strictly true. There was an older SCSI model, though with quite different specs. I'll try http://firewire.zawa.com/sbp2/sbp2_1394_012401.tar.gz as a starting point [if I ever get the h/w]: thanks for the pointer. The scanner I'm looking at is the new Nikon film scanner.
-- Source code, list archive, and docs: http://www.mostang.com/sane/ To unsubscribe: echo unsubscribe sane-devel | mail majordomo@mostang.com
This archive was generated by hypermail 2b29 : Sun Jan 28 2001 - 09:42:26 PST