Re: XML

[Lars Clausen <lrclause cs uiuc edu>]

On 03 Feb 2003, Lars Clausen wrote:
> On Mon, 3 Feb 2003, John Edstrom wrote:
> > On Mon, Feb 03, 2003 at 09:39:58AM -0600, Lars Clausen wrote:
> > > On Sun, 2 Feb 2003, John Edstrom wrote:
> > > > So far my biggest concern with a dot2dia transformation is that dot
> > > > uses splines and dia doesn't.  Some of dot's edges get pretty baroque
> > > > in complex graphs and arcs and polylines can't do equivalently
> > > > tortuous paths as well.
> > >
> > > Can't you use Dia's beziers?  As I understand it, there are standard
> > > translations between most such things.
> >
> > My background is not in graphics and I didn't know that.  That would
> > simplify things immensely.  Is there a list of such algorithms
> > anywhere?
[...] 
> Note also that Dia's "bezier" isn't really a Bezier line.  It's a string
> of line segments (I believe Bezier) with 4 control points each.  The xfig
> import plugin has some hacky translation from FIG style splines to Dia
> beziers, but it's not exact.

Some more update on this:  While Dia's lines are indeed fragments of cubic
beziers, it's not trivial to translate from B-splines or Calmutt-Rom
splines.  Boehm has a paper that describes the appropriate insertion of
knots to do a full translation, but it's beyond my current time limits.
I'll let the xfig import just use the rough hack I did earlier.

It appears that dot splines are actually beziers, which should make the
conversion easier.  I haven't found any more info other than the fact that
they're called beziers internally.

-Lars

-- 
Lars Clausen (http://shasta.cs.uiuc.edu/~lrclause)| Hårdgrim of Numenor
"I do not agree with a word that you say, but I   |----------------------------
will defend to the death your right to say it."   | Where are we going, and
    --Evelyn Beatrice Hall paraphrasing Voltaire  | what's with the handbasket?