Printer Autodetection

[John (J5) Palmieri]

I don't know if the fdi files database would be slow (the foomatics one
surely is).  I was just questioning if it would be a resource hog.  I'm
not sure.  Actually your version of HAL should give the make and model
just fine (assuming it is USB).  HAL callout is what we use in Fedora.
We also use a daemon for printers that can't be detected (i.e. they are
not in the database or their ieee info is not complete so they need user
input).  I am going to change that to a HAL addon so we don't have a
daemon running.  The backends are all distro specific bits but once you
have all that info in HAL the backend doesn't need to do much. I am most
likely getting to this stuff within the next two weeks and it should
filter back to HAL.  I will test the FDI file merging to see if it would
be a bottleneck.  I am most certain it will be faster than foomatics but
the question is will it cause slowdowns for HAL in general since it will
have to troll through huge FDI files?  Someone who understands the way
FDI files work better might be able to answer this. 

On Sun, 2005-11-13 at 17:25 -0500, brendan powers wrote:
> Hi,
> Thanks for the quick response. I did not know that hal did this as am
> using a version of hal that is quite out of date (0.4.7). As for the
> foomatic databse, it has autodetect information that does mach the fdi
> files for some printers, i'm not sure how many printers have that
> information though. If the FDI file database is slow, would it be
> possible to use some sort of collout to detect the printer? I think it
> would be very usefull for hal to provide this information if possible.
> This way more distrobutions could make use of the code. As for
> parralel printers, in my software i checked for them once at bootup,
> this is a fairly simple way to allow support for printers without
> polling.
-- 
John (J5) Palmieri <johnp at redhat.com>