> And you're completely WRONG. > > What drivers are you going to get with your scanner for Linux ? > Binary-only i386 crappy proprietary drivers written in C++ ? > > 1. this won't help non-i386 users > 2. there are no TWAIN applications for Linux (yet) > 3. C++ means the drivers *WILL* be broken at some point by an ABI > change (because, yes, that's going to happen AGAIN). > 4. as the drivers will basically go unmaintained after a year, and the > source won't be released anyway, you'll be stuck with an unusable > scanner. > > Looks like an improvement to me, indeed. Did you read the code Mustek > sent to Henning ? No ? You should. That's what you'll get with your > scanner. > > JB.
I think it is not unlikely that in 10 years most scanners will be sold with a Linux driver in the box. So the question becomes what is the best way to help manufactures supply a good driver. The recent Mustek code shows the problems manufactures have in supporting the windows world and open source. I think Mustek should be congratulated for making an effort - more than most companies to date. The code falls far short of the expected standard, but making excessive demands will only discourage companies from contributing. We should be looking for a path that allows companies a easy transition from closed x86 drivers to open source. And it looks like that transition will be via TWAIN 2.0. David
