Re: [pdf-devel] Expanation, please?

[Jose E. Marchesi]

    > I reckon the FSF communicate may be confusing.  It is obvious that the
    > project is not "done" at all and that not all the issues motivating it
    > have been solved.  
    
    What was solved? What wasn't?

As I said, the licensing of both ghostscript and the xpdf codebase is
now much better than in 2007: they are both distributed under GPLv3 (in
the case of xpdf it is under a dual license supporting both v2 and v3 in
its version 3.03).  The GNU PDF software is GPLv3+, meaning version 3 or
any later version as published by the FSF.  That is the preferred
licensing schema for GNU packages, and is very important for us.  So the
licensing problem was fixed, but only partially.

The support for interactive features (namely forms and annotations) is
much better now in free software engines.  A big concern was that the
citizens of some countries were being forced to use proprietary software
in order to interact with their governments (like in taxes declarations)
or in some other way.  The situation is not that bad now thanks to the
good work of the libpoppler hackers.

On the architectural side the library we are writing mimics in a big
extent the architecture of the Acrobat SDK (unlike xpdf/poppler and
ghostscript) by providing a layered set of well documented APIs.  Those
layers will provide access to the PDF documents in several levels for
both reading and writing.  The purpose of the library is to not just to
serve as a rendering engine for PDF viewers, but to provide the
foundations to write many different applications manipulating PDF
documents.

We are also working in verifiable support for the several standards
(PDF/A, PDF/X, etc).  This will allow to certify the library in order to
support free software applications in some areas such as long-term
preservation of documents.

Another key aspect of GNU PDF which is missing from the existing free
implementations is extensibility: we want the library to be extensible
to allow people to add the functionality they need.  We wanted to add
support for plugins written both in C and in some interpreted language.
The development of a PDF-specific extensible graphical application in
the same spirit than Acrobat is also included in our plan.

This is not an exhaustive list but it exemplifies why, IMO, we are not
done yet.
    
    So, we're still on the go but don't know clearly what for. How can
    we be sure we're not duplicating efforts?

There will always be some degree of duplication among the PDF engines,
but they are different and they have different goals.  You may think
that some other project provides a better foundation where to implement
the features above... well, hacking time is an extremely valuable
resource and each hacker must decide where her time is best invested :)

-- 
Jose E. Marchesi    address@hidden
GNU Project         http://www.gnu.org