plustek: CanoScan N1220U with Testversion 0.45-TEST5 gives I/O error
Subject:
Re: CanoScan N1220U with Testversion 0.45-TEST5 gives I/O error
From:
Gerhard Jaeger ####@####.####
Date:
7 Jan 2003 08:02:38 -0000
Message-Id: <200301070856.29082.gerhard@gjaeger.de>
Hi,
the changes you notice should be somewhat transparent
to any backend that uses them. The handle fd or later on
dn is simply a value that should reference the correct device.
This value can or better will be obtained by a
sanei_usb_open()
call. That's all...
On Montag, 6. Januar 2003 23:06, Marc Lavergne wrote:
> Just as a side note, I noticed that between 1.0.8 and
> 1.0.9 of Sane, the sanei_usb_write/read_bulk function
> appeared to change in a fairly significant way:
>
> 1.0.8:
> ------
> sanei_usb_write_bulk (SANE_Int fd, ...
> {
> ...
> write_size = write (fd, buffer, *size);
> ...
> }
>
> 1.0.9:
> ------
> sanei_usb_write_bulk (SANE_Int dn, ...
> {
> ...
> write_size = write (devices[dn].fd, buffer, *size);
> ...
> }
>
> When I look at sanei_lm983x.c in PLustek 0.45-TEST5, I
> see the following:
> ...
> sanei_usb_write_bulk( fd, command_buffer, &size );
> ...
>
>
> I do get further into the debug stack when using 1.0.8
> with 0.45-TEST5 than with 1.0.9. Am I correct in
> assuming that 0.45-TEST5 works with 1.0.8 but not
> 1.0.9?
No, at least if should work with both versions. To be honest, I've
not testet 1.0.8, but I'm working with 1.0.9...
One hint, beofre hunting any ghosts here, make sure not to mix
binary installed libraries and self-compiled stuff - be sure to clean
all sane-stuff. Then recompile and install again, this should cure
the strange behaviour you have. Sometimes it happens,
that you need to replug your scanner after the whole box is up.
This behaviour seems to be relatet to the USB subsystem...
Hope this helps a bit
Gerhard
>
> Thanks,
>
> Marc
