[cairo] line width = 0.0 ?

[Carl Worth]

On Mon, 09 Jan 2006 12:53:47 -0800, Bill Spitzak wrote:
> However underflow is *not* a problem or excuse for this. It would be 
> trivial to remember whether the number passed in was non-zero, no matter 
> what happens to the transform.

Underflow can also happen in the user applications floating-point
code, and not just due to cairo's CTM.

> No this is not what is wanted. The resulting lines should be the same no 
> matter what the current transformation is. If you scaled by .5, these 
> 72/300 lines would draw 1/2 as bright. We need lines that draw exactly 
> the same (and are thus visible) no matter what the transform is.

I've definitely seen applications where hairlines are a desirable
feature, but I have not seen compelling evidence that this is
something that should be handled within cairo. As we've discussed here
before, their is device-pixel snapping needed for hairlines, and the
proper snapping is really a global issue, which cairo cannot address
well with the local information it receives.

Even if cairo did add hairlines, it would still be only a band-aid,
since proper snapping is needed on fills and strokes with widths other
than a single-device pixel in order to get the best-looking effects.

-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.freedesktop.org/archives/cairo/attachments/20060109/958dd498/attachment.pgp