Re: Comments on SVG Compositing

[just clipping away the emails, since they're long and mostly don't
need direct replies]

Re: clip-to-self
----------------

Even though the example clearly shows that clip-to-self runs off of
geometry rather than alpha, in my head I kept thinking about it in
alpha terms.  I see now why it's not equivalent to porter-duff
operations, even for something simple like 'clear' (using
comp-op:clear and clip-to-self:object would clear out the object's
shape even if the object was fully transparent).

I *much* prefer the suggested name of comp-op-region, with the current
'canvas' and 'object' values.  That makes good, intuitive sense, and
the 'object' value is decent at suggesting that the property cares
about geometry, not alpha.

The spec says that "For... filters, the bounds of the object are
converted to a clipping region".  Are the bounds of a filter
well-defined?  (They may be; I haven't completely absorbed a lot of
parts of SVG yet.)


Re: enable-background
---------------------

Darn, too bad about the name already being used in Filters.  In that
case, yeah, keep it as it is.  The name isn't *horrible*, it's just
less than ideal imo.

Replacing the example would do wonders, I think.  It's too difficult,
imo, to understand what is going on in the current example, where the
structure of the markup (namely, that the first two items are in a <g>
or similar, which is the thing that has the enable-background property
set on it) is unclear, which is vital to understanding that the
property only affects the construction of a compositing group.

~TJ

Received on Thursday, 14 April 2011 17:26:42 UTC