> I created a clip-path element: A filled circle "masked" (more
> precisely clip-pathed) by a filled rectangle. On the circle element of
> the svg document, i put a onmouseover and a onmuseout attributes, wich
> The problem is that, when I try this document in batik, the color of
> the shape changes even when the cursor is over the non-apparent part
> of the circle: The cursor is outside the rectangle, but would be on
> the circle if it wasn't "clip-pathed", and the part of the circle
> inside the rectangle changes its color.
I guess the specification isn't clear enough regarding this. See bug
46289 , which contains a patch by Thomas to "fix" it and also links
to relevant discussions in the www-svg mailing list. :-)
> With, firefox, i don't have this problem (the color changes only if my
> mouse is on the part of the circle inside the rectangle).
Unclear or underspecified stuff usually leads to this sort of
unexpected incoherences among implementations... :-|
> So I would like to know if it's a bug of batik? Will it be fixed?
I'm not sure if the patch will be integrated. I'd vote for so, but I
remember the discussion somehow hit a dead end: there was no clear
consensus at the time... :-( Maybe this is addressed in the second
edition of the SVG 1.1 specification. :-)
> Can I try to "fix" it? I tried to search the Javadoc API for the class
> responsibles for this interpretation, but i've no idea! If there are
> some batik developpers on this mailing, can they shox me the
> "incriminated classes"
As stated, in order to "fix" it you may locally apply the patch
already referred. ;-)
As stated within the www-svg debate, another way of working around
this (in an inter-operable way?) would be adjusting the
"pointer-events" property  to "visible".