Juan Antonio Martinez wrote:
> Our solution was to provide a generic interface to the driver layer, and 
> define a "generic" card: a simply wrapper to talk via socket (or dll, or 
> so)
> with a propietary driver...

I made the decision *not* to allow external drivers of any kind, 
deliberately. I was worried that fprint may become the defacto standard 
for fingerprint scanners, and then vendors would start producing binary 
drivers which would gain lots of adoption and cause us lots of 
headaches. I want everything open and in one place, like how it works 
with drivers in the kernel.

I am still not interested in interacting with closed drivers, but I am 
prepared to add the functionality necessary for some of the "good" 
use-cases that have been described here. e.g. capturing the fingerprint 
image on a very low spec machine where mindtct would take minutes, 
sending the image data to a high-spec machine to do the actual 
processing. That may make it possible for external drivers to interact 
with fprint, an unwanted (but reluctantly accepted) side effect.

Daniel

_______________________________________________
fprint mailing list
[email protected]
http://lists.reactivated.net/mailman/listinfo/fprint

Reply via email to