[sane-devel] discussion: Future of SANE-project

[Henning Meier-Geinitz]

Hi,


On Tue, Jul 06, 2004 at 10:00:46AM -0400, David N. Paules wrote:
> I subscribed to this developer list to monitor and hopefully
> contribute to getting my lame UMAX Astra 2100U working natively on
> Mac OS X. Apparently that will never happen. I then noticed lots of
> messages about Epsons, and other 'big name' scanner vendors. This got
> me to wonder. While the SANE project is a fascinating and honorable
> goal, is anyone selling the benefits of SANE to the scanner
> manufacturers?

There is a page pointed to manufactureres on our website:
http://www.sane-project.org/manufacturers.html

If you have any additions/corrections or comments, feel free to tell us.

> My point is that this project will constantly have to be supported
> from people like myself who are irritated that their scanner doesn't
> work on different hardware/ OSes than what the maker originally
> intended (or cares about).

The problem with scanners in comparison to other hardware is that
there are so many different chipsets and there is no standard for
communicating with scanners. E.g. there is not even a way to identify
a scanner on the USB. So you really have to start from scratch for
every new chipset.

> However, if the SANE architecture is so
> much better than existing dominant TWAIN architecture, then someone
> should be pitching this story to the scanner manufacturers. If Epson,
> HP, UMAX, Dell, Primax, etc. agree to support this open source
> initiative (i.e, provide SANE backends),

Epson does, Brother does. Mustek did for one scanner. Others provide
information and support to SANE developers.

> the hope is that their scanners will be better supported on more
> hardware, reaching more customers and improving satisfaction with
> EVERY customer.

That's what I tell manufacturers when I ask for information/source
code etc.

> Otherwise, the developers on the SANE project will always be writing
> backends for 'old' scanners. In effect, they will ALWAYS be playing
> catch-up with what already works on at least one OS and driver
> config.

Unfortunately some manufacturers seem to think that ~90% of the market
is enough for them. So there is no reason to support non-Windows
systems. Maybe this changes if MS Windows market share is only 80, 70
or 50%.

> The other option is to get the SANE architecture to be a standard on
> *nix boxes which means getting developers of the various *nix OSes to
> bundle pre-compiled SANE with their OS.

I think that's pretty much the case. On the other hand, is there any
commercial *nix (but MacOS X) that really has a bigger market share?
Servers usually don't have scanners...

Bye,
  Henning