So you expect software written for both SVG-specific and
XML-generic to understand both SVG and XML fragment identifier rules? It
seems more
likely that SVG applications would only understand SVG FI
rules and XML applications would understand XPointer - fortunately, a
superset of SVG.
I think the question here is one of intent. If people believe that
generic XML processors are going to be manipulating SVG as part of
common practise, then it makes little sense to have a media-types
specific fragment identifier. If SVG is *not* commonly going to be
manipulated by generic XML processors, then a media-specific fragment
id makes perfect sense. It's event better if they line up.
I remember getting into this debate years ago when I proposed having
a generic tree-structured fragment identifier for *all* media types
(non-herarchical data could be handled with just single-level
qualification). You can see where the debate ended up...