[Foomatic] Using usb IDs for autodetection

Kurt Pfeifle kpfeifle at danka.de
Thu Apr 8 15:04:51 PDT 2004


Grant Taylor wrote:

> "Kurt Pfeifle" <kpfeifle at danka.de> writes:
> 
> 
>>Does using the GCC force the compile results becoming GPL? No.
>>
>>To be sure, I asked Mike (Sweet) the day it was released the same
>>question and he confirmed that the PPD being compiled with the PPD
>>compiler will not be forced to be GPL'd.
> 
> 
> Oh, no, that's not the concern.  It's a question of the libraries to
> be used by the actual driver code; if these are GPL, then the driver
> code must be, too.
> 

The whole idea of the CUPS filtering architecture is to have CUPS
provide a feature-complete generic CUPS raster (with an IANA-
registered MIME type of application/vnd.cups-raster), that is *meant*
and *designed* for printing (containing page meta data in the headers
of each raster-page), unlike the other raster formats (like bmp,
ppm, pnm etc.) used by the 2-stage <ghostscript-raster> --> <external
filter formats> processes, which were used by the <external filter>
developers just out of convenience, because nothing better was there
and because it did fit their purpose *somehow*...

The "cupsraster-to-somethingvendorspecific" raster driver doesn't
need to be linked at all against any CUPS lib, FWIW. Also, parts of
CUPS are LPGL -- mainly the API library, IIRC.

The "cupsraster-to-somethingvendorspecific" approach is already
slowly being adopted by vendors. I saw a Sharp leaflet at CeBIT which
announced something to this effect, and when I asked they said that
their raster driver will be "Open Source", maybe even GPL....

> This would of course be a good thing from a customer viewpoint, but
> vendors don't operate exclusively for the benefit of their
> customers...
> 

Cheers,
Kurt





More information about the Printing-foomatic mailing list