Hi, On Wed, Nov 28, 2001 at 12:20:34AM +0100, Oliver Schwartz wrote: > I don't know if it fixes the original problem, but I found a potential > problem in the snapscan backend. In sane_control_option, it didn't guard > against TL_X < BR_X and TL_Y < BR_Y, in particular, when TL_Y = BR_Y the > number of scanlines is zero and bad things may happen (particularly if trying > to divide by the scanlines). > > A quick survey showed that other backends (mustek, epson) may show the same > problem. You have to use the sliders for TL_X etc. to reproduce the error.
Thanks. I will fix the Mustek backend and add a TODO entry for the others. Bye, Henning
