Hi John,
sorry, I couln't respond because I was on holidays. I can speak for the
Phantom 330 / 336 CX scanner, which I am currently working on (with some
success in the meanwhile). At least the 330CX really needs the READ IMAGE
STATUS command, otherwise it shows unpredictable behaviour. In the
documentation is mentioned that the scanner goes into calibration after
this command. And my backend has to query the scanner a few times before
it is ready. So I think we cannot remove the command.
by the way, which SCSI-command-documentation do you have? The newest I
could find on the Microtek ftp site was Rev. 2.2.4 from August 1997. Do
you have a newer one?
Greetings
Karsten
John Fremlin wrote:
> SANE committers please pay attention!
>
> As far as I can see no one has replied to my original mail. I am
> therefore compelled to do so myself ;-)
>
> John Fremlin <chief@bandits.org> writes:
>
> > At the moment scsi_read_image_status sends a READ IMAGE STATUS with
> > a zero transfer length. This wedges my Microtek X6 USB scanner so
> > completely it won't even respond to USB resets.
> >
> > One workaround would be to just not do READ IMAGE STATUS. This is bad
> > because we wouldn't be able to tell if the scanner is ready to scan.
>
> Background information: The READ IMAGE STATUS command is used to see
> if the scanner is ready to scan. In the Microtek SCSI command set
> documentation it says that it should always be used before performing
> a scan.
>
> However, it wedges my scanner unless a rather hideous workaround is
> put in place. Nobody else has tested this hideous workaround on their
> scanner, as far as I can tell so the hideous workaround could quite
> possibly fail to work on other peoples' scanners.
>
> Because SANE opens the SCSI device with O_EXCL it is impossible for
> two scans to proceed at once. Even when running scanimage twice
> without pause, when READ IMAGE STATUS is called the scanner is always
> already ready to scan, as far as I could see from the logs.
>
> I propose that the READ IMAGE STATUS call be removed from microtek2.c
> because it is untested functionality that causes problems for some
> people.
>
> [...]
>
> --
>
> http://www.penguinpowered.com/~vii
-- 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 : Mon Mar 19 2001 - 12:31:28 PST