Re: Dia & Schemas/DTDs

["Durk Strooisma" <durk kern nl>]

Hi again,

First of all, thanks for your replies!

We have concluded that Dia is not the way to go for us. The main reasons for
this are:

- Dia is not very extensible. For example, adding extra attributes to shapes
to put extra meta data in is not possible within XML boundaries.
- Validation of the XML source files is not possible. The problem with this
is that you can not safely write Dia XML in other editors, like a
XML/text-editor. Additionally, in my opinion, all XML documents should be
validatable.

Like James K. Lowden said earlier, it wouldn't be too easy for us to
contribute fixes to cover our critism. Maybe there are internal
architectural changes needed, well that's not easy for an outsider and would
possibly take too much time.

To draw our network diagrams, we looking at RDF/GraphViz/IsaViz as
alternative.Information about RDF:
http://www.xml.com/pub/a/2001/01/24/rdf.html
http://www.w3.org/RDF/

Describing our network in RDF is possible because a descriptional view of a
network consists all of meta data.

Example:
http://www.w3.org/2001/Talks/0506-semwai/sirpac-image.html


Durk

> On 2002.11.06 22:40 James K. Lowden wrote:
> > On 06 Nov 2002 12:26:43 -0600, Lars Clausen <lrclause cs uiuc edu>
> > wrote:
> > > On Wed, 6 Nov 2002, Durk Strooisma wrote:
> > > > We noticed that Dia doens't make use of DTDs or XML Schemas (for
> > > > example to validate Dia XML documents). Even correct DTDs or XML
> > > > Schemas aren't provided, which would be handy to validate Dia XML
> > > > documents within XML authoring tools.
> > >
> > > I don't know enough about DTDs and XML Schemas to judge whether this
> > > is a useful thing.  We do have DTDs, but they're not used by Dia.
> >
> > Durk makes a good point.  It would be useful, and it wouldn't be easy
> > for an outsider to contribute.  Forgive me if I'm belaboring the
> > obvious.
> >
> > If Dia relied on its DTD to validate a file prior to/while loading it,
> > then the DTD would be assured of being a good measure of a good Dia
> > file.
>
> Not that I know much about stylesheet validation either, but isn't it
> fairly computationally expensive, and thus typically only done as a
> "sanity check" apart from the application?
>
> > That would give non-Dia programs a benchmark for creating Dia files.
> > And it would assist non-Dia programs in interpreting Dia files.  (BTW,
> > do I
>
> I agree that formalizing a stylesheet validation would probably be a
> good thing.  I have a hacky Perl script right now that writes a Dia
> file from a data dictionary, and stylesheet validation would help
> pinpoint errors in the Dia file caused by errors in my Perl code.
>
> Andy
> _______________________________________________
> Dia-list mailing list
> Dia-list gnome org
> http://mail.gnome.org/mailman/listinfo/dia-list
> FAQ at http://www.lysator.liu.se/~alla/dia/faq.html
> Main page at http://www.lysator.liu.se/~alla/dia