[cairo] cairo_new_path; cairo_clip [was Re: Explanation/justification for tolerance change?]

[Carl Worth]

On Thu, 06 Dec 2007 01:11:13 -0500, Behdad Esfahbod wrote:
> You said last night that it's inconsistent with
> cairo_new_path;cairo_fill.  I took your word last night but sounds wrong
> to me now too.

Yeah, it's surprising how much a good night's sleep can help. :-)

> cairo_new_path;cairo_fill should fill nothing, and
> cairo_new_path;cairo_clip should clip everything out.  The logic is
> obvious: you start with the empty set and add regions using various path
> construction functions.

My faulty logic was "cairo_fill with no path does nothing so
cairo_clip with no path should do nothing". My subsequent thinking was
imagining the result of both cairo_clip and cairo_fill with a path
that progressively approaches an empty path, (imagine a shrinking
rectangle).

Obviously, either operation should do what those do in the limit.

> So yes, I guess this was a non-issue.  Maybe we need tests for both fill
> and clip if we don't already have.

Chris did add the clip-empty test when he did this work. That's the
important one. I'm not too worried about accidentally introducing a
bug to the NO-OP fill case. :-)

Anyway, my apologies for the noise on this one.

-Carl
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://lists.cairographics.org/archives/cairo/attachments/20071206/a45294f5/attachment.pgp