On Apr 6, 2008, at 2:43 PM, Cyrille Berger wrote:
That model also lacks swatch informations (name, copyright, license,
...)
Yes this is still under consideration. Wether to add some 'simple'
home made
metadata, or use XMP, or use an other RDF schema, or allow anything.
So one opinion on this was that there are a few standard ways this
information is dealt with now (including Dublin Core, Creative
Commons, etc), and thus there is no need to reinvent the wheel for
that. Just use metadata like can be used for SVG.
http://www.w3.org/TR/SVG/metadata.html
and swatch-wide settings ("all rgb values are in Adobe1998 color
space", "all color names start with 'PANTONE' and end with
'C'", ...).
I am not sure about the use of this ? Since the file is not really
intended
for being written by hand by an human, I can only see compression,
but then I
think bzip2 or zip will do a better job. And I don't like that
change, since
it makes the parsing more complicated.
Yes, I strongly agree with this point.
In fact, I've seen things like that in XML being kept verbose
actually improve compression.
Constraints and such like that are easy do to in code. For example,
you can just use a regexp for the name check. Better to leave the
format clear and easy to use._______________________________________________
CREATE mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/create
