[Openicc] colord Printing Plans

[Till Kamppeter]

On 02/24/2011 05:10 PM, Richard Hughes wrote:
> On 24 February 2011 15:51, Till Kamppeter<till.kamppeter at gmail.com>  wrote:
>> What happens if the application from which the job is sent and the printer
>> are not on the same machine or not even on the same local network. CUPS has
>> all capabilities for networked printing, even through the internet. Does
>> this mechanism also work in such a situation?
>
> At the moment, I'm concentrating on the local system use case. The
> idea is that a remote print server will also have an instance of
> colord, and the profiles will be setup on the remote server with the
> icc profiles there. If there is a need to upload the icc file, we can
> look at adding the logic in CUPS to make that possible. That's
> certainly phase 2, so to speak.
>

OK.

>> Is it really needed to have an
>> extra daemon? Is it not possible to do everything with CUPS (perhaps with
>> patches to CUPS)?
>
> I'm sure you could do it all in CUPS, but it's better if you can have
> it a a separate process that hold all the mappings from devices to
> profiles for all devices, e.g. scanners, cameras and most importantly
> displays. This allows you to have display correction without
> installing cups, for instance.
>

OK, for display/scanner/camera/printer color management it really makes 
sense.

>> Do have made the code of pdftoraster and pstoraster available somewhere?
>> pstoraster is perhaps even not needed as due to the nature of Ghostscript
>> accepting both PDF and PostScript as input probably pdftoraster also accepts
>> PostScript and does the right thing with it.
>
> Sure, I've got a private branch here. I've abstracted out a lot of the
> code in pstoraster.c into common.c and am now calling that code from
> both pstoraster.c and pdftoraster.c. If the ghostscript guys want me
> to, I'll happily combine them into one executable and make the pdf
> specific stuff in pdftoraster conditional on the mime type that's been
> sent.
>
> Hopefully today or tomorrow I'll split my mega-commit up into easy to
> review chunks and I'll post it to the gs-devel.
>

Make also sure that the new p*raster filter(s) also work if there is no 
colord running (then at least without color management, as they worked 
before.

    Till