Re: Question about getURL() (SVGT 1.2)

> I read while ago that GPAC [4] also has the method implemented;

Correction: GPAC seems to support XMLHttpRequest. Sorry for the
misinformation. Please search for "XMLHttpRequest" on the link
provided in the Viewer Matrix [1].

> It would be interesting to gather more information and publish the results:
> I'd suggest adding a row within the extensive viewer matrix
> ("Communication", after "Script"?), stating which implementations
> supported communication. I suggested not adding it as "getURL" as there
> are several existing implementations which already support XMLHttpRequest.

Done [1]! :-)

> As we're on the subject, this is something which IMO should/could be
> evaluated in possible test cases which be part of (an optional
> portion) the test suite: optional as it probably had to be live for
> the typical security issues fetching remote data involves.
>> I'm worried that if the behaviour of getURL() is not clear enough people
>> will simply use the XMLHttpRequest() function and the specification of
>> getURL() will just add noise to the SVG specification ...
> IMHO, as XMLHttpRequest [7] is getting more mature (not only at
> specification level but also in implementations), why not propose it
> (at least as a possible alternative to getURL for UAs) within the SVG
> specification?

I'd love to see some feedback on this. I seriously believe that SVG
success will raise from the ability to support Rich Internet
Applications (a.k.a., in some way, AJAX) functionallity. This is only
possible if communication mechanisms are clear within the
specification and tested within an official test suite: for example,
through the proposal already stated.

Best regards,

 Helder Magalh„es


Received on Wednesday, 30 July 2008 16:45:57 UTC