primax: Thread: Re: pxscan & primax_scan on the Colaado Direct 9600

On Wed, May 14, 2003 at 06:52:59PM +0100, robert w hall wrote:
> Why are there two similar code bases?
because Gustaf Kroling forked his pxscan from an quite early alpha
stage of primax_scan and made the modifications needed for the D9600.

> pxscan-0.41 works on my Colarado Direct 9600, but primax_scan-.93
> doesn't.
well ... the D9600 and D300 (or simply Colorado Direct) have different
ASICs and so completely different commands and registers ... that's
why each scanner requires an extra driver. merging them together
again would need either loads of #ifdef or better a complete rewrite.

> However, the changes to make it test correctly didn't at first glance
> look large.
it simply starts at the wake-up sequence which is quite different ...
 
> Is there a later version of primax_scan which will run the 9600 already?
no. and I doubt there will be one in the near future ... 
primax_scan is currently moving towards a more modularized design (backend
scanning and hardware library and a SANE backend.) ... maybe it would be
possible to change the hardware abstraction lib (or simply write one for
the D9600 ... Andre Herms who did this thing for the D300 also only
wrapped this stuff around the old primax_scan stuff ... so changing this
library would be nearly the same as changing the original sources)
  
> Is there's some reason why some kind person (not me for a few weeks, but
> possibly there after) can't combine them ?:-)
read above .-)

PS: I'll also CC this message to the mailinglist ... maybe there're
    some more interested and even have time to do it.

-- 
Christian Ordig
Germany