Re: [css3-images] Image Fragments and SVG URIs

On Jul 6, 2012, at 10:17 AM, "Dr. Olaf Hoffmann" <Dr.O.Hoffmann@gmx.de> wrote:

> ~fantasai:
> ...
>> 
>> I'm going with the theory that Media Fragments supercedes the SVG spec
>> in this respect, since Media Fragments does not co-opt any existing SVG
>> fragment syntax and is defined to work for all image formats, of which
>> SVG is one. Therefore a new SVG version is not needed for MF to work on
>> SVG images.
>> 
>> ~fantasai
> 
> As others already mentioned, the Media Fragments PR already excludes
> this, it contains a section, that clarifiies, that especially SVG has already 
> its own syntax. SVG explictly specifies, how such views into documents
> work, this practically excludes other options - if other formats are silent
> about this, I think, for them it does not exclude, that the Media Fragments 
> PR may apply.
> Of course, it is not wrong to add this #xywh=... to a URI of an SVG, but
> it simply has no meaning for the interpretation of the SVG document,
> because it is explictly defined, what kind of strings after the # have
> currently a meaning for the interpretation of an SVG document.
> 
> Therefore such examples using the syntax of Media Fragments PR
> will become meaningful for a new SVG version, if this specifies, that
> Media Fragments PR applies. If this is for example SVG 2.0, the description
> of the CSS example using this syntax has to contain the information, that
> this only applies to documents conforming to SVG 2.0 ff), for 1.0, 1.1, 1.2
> still only the original SVG syntax applies. If the document indicates, that
> a specific version is used, nothing can supersede this (except of course
> the (wrong) interpretation of such documents with viewers ignoring such
> version indication, buth this applies only for the interpretation, not for the
> question, whether something is applicable or correct interpretation - 
> interpretations are always subjective, the audience should not trust in the
> interpretation of a viewer. 
> This is the advantage of versioning, one always knows, what applies 
> and which interpretation is right and the author has the option to
> explictly mention what is intended, what cannot be superseded by
> newer recommendations, whatever they say ;o)
There is no problem if media fragments extends SVG 1.1. The backward compatibility needs to be  considered. And as discussed during the last SVG WG telcon yesterday, the syntax is a bit different, so that URI is not necessarily affected. URI needs XML names, while the introduced fragments on media fragments need equal signs (which can not be used for XML names IIRC). Therefore there is a way to differ between predefined fragments and ids #xywh and #t.

Dirk

> 
> Olaf
> 
> 
>

Received on Friday, 6 July 2012 10:52:38 UTC