The "mode_sense_1" command on the Microtek scanners is used to set the
channel-by-channel brightness offset. In the current backend, it just
sends zeros because I haven't yet figured out why negative offsets will
not work, and it's sometimes easier on the developer to not release
broken features.
This command *can* be commented out, at least for a Scanmaker E6 (my scanner).
The simple patch attached illustrates this. However, this may not be the
case for the Scanmaker III.
It sounds to me like the Solaris SCSI implementation is trying to be a bit
too clever or too strict --- 0x19 is designated as "vendor specific", so I
don't see why the Solaris kernel assumes that it is being sent to a tape
drive.
Someone should probably harass Sun about this.
I would rather not just remove this command from the Microtek backend, since
it's one particular OS with a flaw that is complaining.
-Matt Marjanovic
--- microtek.c.dist Wed Feb 4 14:44:07 1998
+++ microtek.c Wed Feb 4 14:46:38 1998
@@ -2514,7 +2514,7 @@
if ((status = wait_ready(s)) != SANE_STATUS_GOOD) return status;
if (s->dev->info.does_mode1) {
if ((status = mode_select_1(s)) != SANE_STATUS_GOOD) return status;
- if ((status = mode_sense_1(s)) != SANE_STATUS_GOOD) return status;
+ /* if ((status = mode_sense_1(s)) != SANE_STATUS_GOOD) return status;*/
}
if (s->dev->info.source_options & (MI_SRC_FEED_BT | MI_SRC_HAS_TRANS)) {
if ((status = accessory(s)) != SANE_STATUS_GOOD) return status;
-- Source code, list archive, and docs: http://www.mostang.com/sane/ To unsubscribe: echo unsubscribe sane-devel | mail majordomo@mostang.com