Re: img issue: should we restrict the URI

On 1/25/08, Anne van Kesteren <annevk@opera.com> wrote:
> On Fri, 25 Jan 2008 18:48:18 +0100, Jeff Schiller <codedread@gmail.com>
> wrote:
> > Thanks for the update from Opera (haven't had time to test/read up on
> > it lately).  I'm all for restricting scriptability on images.
>
> "SVG as image" is basically SVG with scripting disabled and no event
> interaction. Ideally "declarative" animations still run as long as they
> don't depend on events. I believe in Opera they currently run for <img>,
> but not for 'background-image'.
>

Great - this behavior is exactly what I'd like to get defined in the
spec for <img> so that it is non-ambiguous what an "image" is and how
it behaves.

> > What about animated GIFs - what if some user agent allows you to
> > restart/freeze/loop the animated GIF?  That is technically interaction
> > and by my simple definition it would not be allowed on an image...
> > what about panning?  That's interactivity too...
> >
> > So perhaps "non-interactive" is too restrictive a term?
>
> Non-interactive in the context of HTML 5 is as far as the Web page is
> concerned. <blockquote> is also non-interactive yet the user agent could
> offer a way to get to the citation URI.

Fine.  I just want to clarify that things like mouse events (and other
interactive events like focus, scroll, keypress) do not pass down into
the image content.  How about:

"An image is defined as a visual representation that must not receive
interactive events nor execute any script.  If the image content
generates a DOM (as in SVG), the image's DOM is completely detached
from the DOM of the referencing document."

Received on Friday, 25 January 2008 18:22:21 UTC