619971 - Timing of string animation changes causing filters-image-02-b to fail

Status: RESOLVED INVALID

(Core :: SVG, defect)

Description

[Cameron McCormack (:heycam)]

filters-image-02-b from the SVG 1.1 test suite is failing.  It looks to me like the test is expecting different timing behaviour for the animation of the xlink:href="" on <feImage> and <image>.  Firefox seems to be switching to the to="" value at the beginning of the 2s duration rather than at 1s or 2s -- I can't remember which is correct for types like strings that don't interpolate, when calcMode="linear", but I think the test is expecting 2s.  Brian/Daniel, could you confirm whether the test is valid?

Comment 1

[Daniel Holbert [:dholbert]]

The test effectively has calcMode="discrete" (since you can't linearly interpolate strings), and given that, this behavior is correct per the SMIL spec.  We actually "fixed" our implementation to have the current behavior in bug 544855.

As noted in that bug's opening comment, this isn't the first test in the SVG testsuite to fall prey to this misconception -- http://lists.w3.org/Archives/Public/www-svg/2010Feb/0015.html is a www-svg post about the same issue in a different test (animate-elem-227-t).

Comment 2

[Daniel Holbert [:dholbert]]

Relevant bits of SMIL spec:
> if the attribute does not support linear interpolation (e.g. for strings),
> the calcMode attribute is ignored and discrete interpolation is used.
...
> For the shorthand form to animation, there is only 1 value; a
> discrete to animation will simply set the "to" value for the
> simple duration.
http://www.w3.org/TR/smil-animation/#AnimFuncCalcMode

Comment 3

[Daniel Holbert [:dholbert]]

(In reply to comment #1)
> this behavior is correct per the SMIL spec

(if it wasn't clear: "this behavior" = "mozilla-central's current behavior")

Comment 4

[Cameron McCormack (:heycam)]

The test has been fixed (and we now pass it).
http://www.w3.org/mid/[e-mail]