[cairo] Re: Using ADD operator to prevent seams

[Owen Taylor]

On Wed, 2005-03-23 at 09:35 +1200, Robert O'Callahan wrote:
> Thanks for this thread. It was well written, very educating and
> directly addresses my questions and needs.
> 
> Gecko's in the same boat as Inkscape. Web authors also write documents
> that induce the adjoining edge problem, e.g., something as simple as a
> sequence of line boxes with a background color. (But lots of more
> complicated scenarios exist.) Currently this causes us problems
> because we simply don't round correctly (e.g.
> https://bugzilla.mozilla.org/show_bug.cgi?id=173051) but even if we
> fix that, we're still going to have problems on antialiased devices.
> 
> Like Inkscape, figuring out adjoining edges in general is going to be
> prohibitively complex and probably too slow also ... speed is
> incredibly important to us. That also rules out compositing with lots
> of temporary buffers. The dilation approach will cause artifacts with
> transparent fills. So my guess is we'll want to go with the Flash-like
> supersampling approach. I think we could automatically tune the system
> to use supersampling only when rendering is keeping up with the
> document's demands, given the available resources.

When doing antialiased rendering, it's indispensable to position 
horizontal/vertical edges to integer pixel boundaries. Full-screen
supersampling doesn't help: you'll still have fuzz at the edges of the
boxes, it will just be "correct" fuzz.

Do adjoining diagonal (or curved) edges really come up enough in web
pages to be worth worrying about? (Obviously, you can get whatever
problems you have with SVG by embedding an SVG, but that doesn't seem
to be the normal case...)

Regards,
						Owen

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
Url : http://lists.freedesktop.org/archives/cairo/attachments/20050322/c6468d70/attachment.pgp