Re: DTC436e SCSI & ScanExpress
Rogier Wolff (R.E.Wolff@BitWizard.nl)
Wed, 15 Apr 1998 15:22:17 +0200 (MET DST)
Dr. Andreas Bolsch wrote:
>
> This card seems to work with the g_NCR5380 driver (but only with
> the mustek patch supplied in the same directory as SANE itself.
> You have to supply some additional parameters, for using as a module
> 'modprobe g_NCR5380 ncr_addr=0x280 ncr_irq=255 dtc_3181e=1' works
> for me.
>
> At first I thought there is some trouble with the driver since my
> scanner (ScanExpress 12000SP) doesn't work with the mustek
> backend. But yours should work.
>
> But now I've made some progress, it seems that the scanner's
> SCSI implementation is *very* crippled, even after totally weird
> commands the scanner never sends any error code at all.
> Only INQUIRY worked first, but according to David's hint I am
> trying SCSI-II commands, and --- what a surprise --- something
> happens. After SET_WINDOW and SCAN the slide moves slightly
> forward, and back to the home position after closing the device.
>
> And READ produces some noise like moving the slide back and forth
> a few scanlines, perhaps continuous backtracking because the
> scanner can't get rid of the data. Unfortunately, the SCSI-II
> standard isn't very precise concerning how to specify the
> amount of data to be transferred. I don't know what the 'blocksize'
> actually is.
Just try to read 28k blocks or so. That should, once you've got things
setup correctly give you a "working scan", but it might be horribly
slow. There is a possibility that the scanner will enforce you to ask
for a multiple of the line-length. You could try to behave correctly
by taking a number that will fit any possible line-length that you
expect.
The Tamarack scanner also speaks a SCSI-II scaner like language. This
does not mean that any SCSI-II scanner will work with any driver using
a SCSI-II commands: There are enough places where there are lots of
differences possible, and since the spec isn't specific enough,
different vendors have done different things.
Ideally, all standard-conforming scanner backends should be merged,
with a small "scanner spec" file defining the specifics for each
scanner.
Roger.
--
If it's there and you can see it, it's REAL |___R.E.Wolff@BitWizard.nl |
If it's there and you can't see it, it's TRANSPARENT | Tel: +31-15-2137555 |
If it's not there and you can see it, it's VIRTUAL |__FAX:_+31-15-2138217 |
If it's not there and you can't see it, it's GONE! -- Roy Wilks, 1983 |_____|
--
Source code, list archive, and docs: http://www.mostang.com/sane/
To unsubscribe: echo unsubscribe sane-devel | mail majordomo@mostang.com