Hi,

On 2020-03-07 6:06 p.m., Olaf Meeuwissen wrote:

Thanks that makes sense to me.
Perhaps the SANE spec could be updated to add some clarity in this regard.
Issue and/or MR welcome at https://gitlab.com/sane-project/standard :-)

That said, I also think SANE_CAP_SOFT_SELECT is the better choice from
the user point of view.  However, you might consider adding a check for
the current hardware side setting to sane_control_option() so it can
signal the frontend in case changes were made hardware side.

# I considered adding a polling loop but that doesn't make much sense.
# You can only signal the frontend via sane_control_option() that it
# needs to SANE_INFO_RELOAD_OPTIONS, so checking there is good enough.


Ah, I have implemented and checked in the change to make those sensors work but as you say, xsane currently does not take account of changes made on the hardware which is a shame.
I will look into the SANE_INFO_RELOAD_OPTIONS aspect.

I think most proprietary scan software implementations that I have observed use a polling loop of some sort, some utilizing USB INTERRUPT, some not.

When I have looked into that, perhaps you would be kind enough to review the changes? :D

Cheers,
Ralph

Reply via email to