Re: Mac OS X driver.

[Ethan M. <merritt@u.washington.edu>]

On Wednesday 31 March 2004 04:22 am, Hans-Bernhard Broeker wrote:
> On Tue, 30 Mar 2004, Ethan Merritt wrote:
> > On Tuesday 30 March 2004 01:17 pm, Per Persson wrote:
> > > I was a little curious about the reason for the overlap between
> > > filled_polygon() and boxfill(), any particular reason?
> >
> > Two parallel projects that got incompletely merged.
>
> That's not the entire truth, though.  There's also the problem that
> doing cross-hatching pattern-fill of arbitrary polygons is at least an
> order of magnitude harder than either cross-hatching of axis-aligned
> rectangles or solid-colour filling of arbitrary polygons.

I worried about that originally, but it turns out not to be a problem
in practice.  Other people have already done the hard work, and
included the necessary support in the relevant libraries or languages.

The patch on SourceForge shows that pattern-fill of polygons
works with no special effort for pdf, png, svg and x11.

There is an issue with PostScript, but it's more a question 
of policy than of difficulty.  PostScript pattern-fill requires 
PostScript Level 2 language support. This is nearly universal
by now, but not quite.  The question is, what should we fall
back to if it is requested but not present?  It seems to me that
at worst we fall back to doing nothing at all, which is what
happens in the current version anyway. 

That leaves Windows as the largest gap in implementation.
Is pattern-fill so horribly difficult in Windows?  The current 
code in src/win/wgraph.c does this:
	idx = <expression that evaluates to pattern number>;
	SelectObject(hdc, pattern_brush[idx]);
That suggests to me that the pattern-fill support can be
applied to any sort of object, but I could be wrong.


-- 
Ethan A Merritt       merritt@u.washington.edu
Biomolecular Structure Center
Mailstop 357742
University of Washington, Seattle, WA 98195