Re: SVG12: !important in property animation

* Chris Lilley wrote:
>The attributeType is purely to allow disambiguation in the case of a
>name clash between a property and an attribute.

Yes, to identify the target, which has a type with an associated base
type and an associated domain. SMIL 2.1 discusses this in detail and
requires that values are valid for the domain of the target. I agree
that the copied and pasted bits in the draft under discussion are not
clear about a number of details relevant here, but thankfully it is
pointed out that the normative definitions can be found in SMIL 2.1.

>It does not allow in unusual syntax for attribute values, as you seem to
>believe.

I agree that the confused references to attributes in SMIL Animation and
all the specifications that copied text from it is unfortunate. Clearly,
when attributeType="CSS", the target of the animation is a property, not
an attribute. So all the things you say about attributes are irrelevant.
I'm not really sure what's so hard to understand about the fact that for
attributeType="CSS" all processing is defined in terms of DOM APIs and
style sheets.

>Wrong. It would indeed be allowed for a CSS property, if that CSS
>property took scientific notation.

Yes, but since "scientific notation" is specifically prohibited for all
CSS properties defined in the family of SVG and CSS specifications, that
is not really relevant. I can see no evidence in SMIL 2.1 that e.g.

  <svg:... xml:id="t" ...>
  <set attributeType="CSS" attributeName="font-size" to="1e2" ...
  </svg:...>

would be different from

  ...
  document.getOverrideStyle(t, '').setProperty('font-size', '1e2', '');
  ...

which, 1e2 beeing an invalid value, cannot have an effect. You seem to
think the process is rather

  ...
  document.getOverrideStyle(t, '')
    .setProperty('font-size', SomeMagicFunction('1e2'), '');
  ...

You've so far failed to cite where SMIL 2.1 defines SomeMagicFunction.
In the light of this, you might want to argue that it is not clear
whether to="red!important" maps to which of

  d.getOverrideStyle(t, '').setProperty('fill', 'red!important', '');
  d.getOverrideStyle(t, '').setProperty('fill', 'red', '!important');

or what the result would be in case of the former; I agree that it's
not clear, and as a result SVG implementations do not interoperate,
hence this issue.
-- 
Björn Höhrmann · mailto:bjoern@hoehrmann.de · http://bjoern.hoehrmann.de
Weinh. Str. 22 · Telefon: +49(0)621/4309674 · http://www.bjoernsworld.de
68309 Mannheim · PGP Pub. KeyID: 0xA4357E78 · http://www.websitedev.de/

Received on Friday, 10 March 2006 13:54:38 UTC