Steve Underwood <email@example.com> writes:
> I updated the backend to merge the colours properly in May. Since then I have had
> little feedback. Lawrence is the first person to fully work through anything from
> "I have a problem" to "I have a cure" since then. People need some serious coaxing
This time I think "I have a cure" ;)
I found that ps->pss->expected_read_bytes calculated in SCSISource_get
was slowly decreasing down to 0. At this point
(ps->pss->expected_read_bytes == 0) the scanner would block.
To solve this, I can put it to an arbitrary high value.
// ps->pss->expected_read_bytes =
// ((int) (msecs/ps->pss->ms_per_line))*ps->pss->bytes_per_line;
ps->pss->expected_read_bytes = 50000;
then this line later
ps->pss->expected_read_bytes = MIN(ps->pss->expected_read_bytes,
put it to absolute_max.
In this case, my scanner works fine and does not block anymore. Of
course, it's a dirty trick.
May be the ms_per_line reported by the scanner is not right or
-- Sébastien Sablé firstname.lastname@example.org
-- Source code, list archive, and docs: http://www.mostang.com/sane/ To unsubscribe: echo unsubscribe sane-devel | mail email@example.com
This archive was generated by hypermail 2b29 : Fri Oct 20 2000 - 17:45:21 PDT