I think this is a good starting point. I don't know how eager
hardware manufacturers are to write new drivers for different OS, but
a lot of historical experience shows that they cling to their islands
as long as possible. Current example: writeable DVD.
Especially in the WinWorld there had been no real revolutions. There
are still many people running MS-DOS programs in a Win98 DOS-box.
Thus if something should be changed it has to run parallel to
traditional solutions. If SANE or a derivative is offered as a
second and compatible standard, which requires the TWAIN
administration to be involved, both standards will melt together
automatically. And I am convinced that as soon as hardware developers
notice that a SANE driver doesn't make the scanner eat up pages but
produces an interface for most platforms in one run, they will turn
to SANE just by calculating costs and profits.
What I want to say: If we create a new driver concept from scratch,
which is neither SANE nor TWAIN, it is pretty unlikely to be used at
all. Manufacturers will keep to what they know, anything else is
expensive and a high risk. If SANE becomes an official component,
customers will learn its advantages and use it. Applications grow
with possibilities. By now TWAIN is the scanner driver like Win98 is
built into Pentiums.
If the aim is a standard cross-(all)-platform solution, introduce
SANE softly and stable. There is an exploding market with NT-networks
using Linux servers, which will readily accept this. There even a Win
only TWAIN frontend would win.
Just some economical comments,
Lars. (hanke@janeway.physik.uni-dortmund.de)
Tel.: +49-(231)-755-3536
FAX: +49-(231)-755-3674
-- Source code, list archive, and docs: http://www.mostang.com/sane/ To unsubscribe: echo unsubscribe sane-devel | mail majordomo@mostang.com