Arrow questions

[Lars Clausen <lrclause cs uiuc edu>]

Cyrille (and others who have worked on the arrows), I have a couple
questions before I start doing serious dama^H^H^H^Hchanges to the arrows
code.

Firstly, I notice explicit differences between hollow_ellipse and
filled_ellipse in whether to account for the line width.  Same for the dots
and boxes.  Shouldn't all of the account for the line width, so as to line
up nicely with the connection point?

Second, I notice you're using beziers rather than ellipses to draw the
round shapes.  Any particular reason for that?  Beziers are notoriously
difficult to translate into other formats, as they aren't really standard.

Thirdly, I notice that the various objects have arrow information as Arrow
start_arrow, rather than Arrow *start_arrow.  Not only does that waste some
space, it also means that we can't add onto the Arrow structure without
causing binary incompatibilities.  Any reason to not change that?

Lastly, a general question.  With the upcoming arrow adjustment system, it
will be possible to have transparent arrowheads where the line doesn't show
through.  Is that desirable?  Better than the current white-filled version?
Should we have both?

I'm planning to start out with a non-caching transformation handled by the
objects themselves.  It seems that by keeping the inherited points, things
like loading and saving will still work.  Renderers that know about arrows
will have to untransform the lines, but that's ok.

-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?