On Sun, May 31, 2009 at 9:14 PM, Olaf Meeuwissen <olaf.meeuwissen at avasys.jp> wrote: > "m. allan noah" <kitno455 at gmail.com> writes: > >> On Sun, May 31, 2009 at 7:23 PM, Olaf Meeuwissen >> <olaf.meeuwissen at avasys.jp> wrote: >>> "m. allan noah" <kitno455 at gmail.com> writes: >>> >>>> [snip] >>>> I also would like to add a generic 'SANE_STATUS_RETRY' for those cases >>>> where the backend has to block during sane_start because it must >>>> ingest the entire image before the size is known. >>> >>> The standard allows backends to return an unknown height, so what kind >>> of cases are there when the width can't be known before slurping in the >>> whole image? >> >> most medium/large fujitsu machines will auto crop or auto rotate, but >> the command to request the size returns the data for the previous >> page, so you have to read until eof from the machine. > > I see. > >>>> I'd rather not block... >>> >>> I'd rather not hog memory ... ;-) >> >> do you support any machines which interlace front-side and back-side >> duplex data? then you already do... > > Yes, but in that case I have no alternative.
nor do i. allan -- "The truth is an offense, but not a sin"
