[Gnome-print] [Fwd: Gnome printing issues]

[Chema Celorio <chema celorio com>]

---

• From: Lauris Kaplinski <lauris helixcode com>
• To: Raph Levien <raph acm org>
• Cc: gnome-hackers nuclecu unam mx
• Subject: Re: Gnome printing issues
• Date: 12 Sep 2000 04:40:29 +0500

> Ok, thanks for raising this issue, and also John Trowbridge for raising
> the spectre of patents.
> 
> It is absolutely true that Adobe has change control over the PDF spec.
> In an ideal world, change control for things like file formats would be
> held by free organizations. I think the issues are how good a job the
> organization does in managing the spec, and how much freedom is
> restricted. Historically, Adobe has done well on both counts, although
> there is always the possibility for this to change.

I would not trust Adobe. Just remember, which horrible crap it made of
PostScript (starting from version 2.0) - which was incredibly beautiful
thing at beginning. So horrible crap, it had to develop new file format
(PDF) to get things right.

> So this issue raises several sub-questions:
> 
> 1. Putting aside the question of _which_ PDL to use, should we be using
> a PDL as an intermediate step for printing, or do direct rasterization?
> 
> 2. What are the technical advantages of using PDF compared with growing
> our own?
> 
> 3. What can we do to minimize our vulnerability in case Adobe becomes
> evil in their enforcement of patents?
>
> I believe the answer to (1) is clearly yes. In the local case, it does
> not matter much. In the networked case, using a PDL gives you
> considerably greater flexibility and avoids potentially serious
> performance problems.

Agree.

> While I am definitely sympathetic to the freedom issues, I am reluctant
> to grow our own PDL when there's one of extremely high quality already
> available. Frankly, it smells like NIH to me. I question whether we have
> the resources and commitment to build a PDL that comes close to the
> quality of PDF. It's not trivial, guys! Lastly, we give up
> interoperability with the rest of the world. I realize that people may
> have differing senses of the importance of this interoperability -
> purists will say "freedom is much more important", while others will be

It relly is :)

> out there developing interesting software like Samba, Wine, wv,
> gnapster, x86 compiler back-ends for gcc, etc.

OK. I do not know little about PDF. But definitely we want gnome print
PDL
to be done the right way. So there certainly are lots of other issues
aside
color management etc.
- how easily is it parsed
- how small subset we should use (we certainly do not want to use the
full
spec for print jobs - like hyperlinking etc.)
- how easily we can add our own extensions to it
etc.

> Now, what do we do if Adobe enforces their patent? My favorite idea is
> to define a patent-free profile of PDF 1.4. My feeling is that PDF 1.3
> plus a subset of the PDF 1.4 transparency features (the ca and CA, and
> SMask entries in the extended graphics state, plus SMask images, most
> likely), would meet these needs. I believe this is a good feature set
> for Gnome Print, and it allows a totally free implementation. Much
> useful interoperability is preserved. Particularly, it should be
> possible to produce a document from Gnome, and have it spooled, viewed,
> and managed on other systems. The reverse direction seems to me more of
> a Ghostscript issue than one for Gnome Print.

Well, I have nothing against gnome-print ability to generate PDF files.
But
I am still not sure, PDF is the best internal PDL. For external spooling
systems we can use whatever PDL best suited for them.
Anyway, if we go with that then I suppose we should:
- rename it (calling it PDF has probably some issues with Adobe?)
- define reasonable subset we will use (and only that)
-  will continue developing and extending it ourselves
I.e. we still go with our own PDL, but instead of starting it from
scratch, we'll start from existing Adobe specification.

Lauris

---