[sane-devel] Experimental lexmark backend inclusion in regular CVS

[Fred Odendaal]

Stef,

I've finished testing the experimental code on my X1185 and it works 
well. It is much slower than the current backend - probably due to 
shading calculations during the scanner reads. This may be optimized. I 
know there is a double buffer copy that I put in that can probably be 
removed. So,  I don't see anything that can't be worked around on the 
performance side.

The quality of the scans are much better with the experimental backend 
probably due to the shading, and the gain and offset calculations. Very 
nice!

The code changes look good too. I do have a small concern about name 
space collision with the use of functions named "rts88xx_*", but they 
are only used in a single file and aren't exported so shouldn't be a 
problem. If they are moved into a separate file as indicated in the code 
they may have to be renamed...

So, all in all I'd say the good outweighs the bad and we should move the 
experimental code into the main repository. I'd be interested to hear 
other opinions.

regards,
Fred Odendaal

Stéphane VOLTZ wrote:
> 	Hello,
>
> 	the experimental lexmark backend is now quite in a good shape. It brings 
> other the current backend:
> 	- 2 new USB id support (Dell A920 and Lexmark X12xx models)
> 	- add support for new X11xx models
> 	- add arbitrary scan area selection
> 	- offset and gain calibration
> 	- software shading correction
>
> 	It has been tested on A920, X1200 in USB1.1 and newer X1100 model. Only the 
> orignal model supported by the current backend should need some testing.
>
> 	So I'd like to know what everybody, and especially Fred, think about moving 
> the experimental backend into regular CVS.
>
> Regards,
> 	Stef
>
>