[sane-devel] Arcard and mustek

[Henning Meier-Geinitz]

Hi,

On Tue, Jan 14, 2003 at 01:54:41PM -0800, alto.iole at earthlink.net wrote:
> At the present I cannot check the sane
> version number because I am not at home, anyway
> it sound to me 1.0.9 ( I downloaded at the begin
> of December.)

That should be ok. There was a bug in old versions of SANE that may
have cause such errors.

> I agree that the problem should be in the data transfer of initialization of
> the transfer. 
> 
> I think that I already comment out the double
> buffering option in the mustek.conf.

You can also try to decrease the size of the buffers to 128 kbytes (or
even lower). Do that also in mustek.conf.

There is also the following iformation that I once had on my website
but someone told me it's no longer necessary:

      Some people told me that their scanner wasn't detected using Acard /
      Advance SCSI adapters. If you look at the debugging output (level 5) you
      can see, that the inquiry information is cut at byte 36. After that byte
      only garbage is returned. As this information is essential for the
      detection of the scanner, the backend can't detect it. Acard has an <a
      href="download/atp870u.c">updated driver</a> which works concerning
      inquiry. But there seems to exist another bug that braeks scanning for
      some scanners. The following should work: remove the following lines
      from atp870u.c:

      if (dev->ata_cdbu[0] == READ_CAPACITY) {
        if (workrequ->request_bufflen > 8) {
                workrequ->request_bufflen = 0x08;
        }
      }


      if (dev->ata_cdbu[0] == INQUIRY) {
        if (workrequ->request_bufflen > 0x24) {
                workrequ->request_bufflen = 0x24;
                dev->ata_cdbu[4] = 0x24;
        }
      }

Otherwise I have no more ideas. Maybe you can find more information in
the archive of the Linux kernel mailing list.

Bye,
  Henning