[
Lists Home |
Date Index |
Thread Index
]
On Fri, Feb 01, 2002 at 01:14:17PM -0500, Simon St.Laurent wrote:
> On Fri, 2002-02-01 at 11:59, David G. Durand wrote:
> > If you want Xpointer killed, you can of course relax, because it
> > seems to be dying quickly.
>
> No, don't relax. (Unless it's RELAX NG.)
>
> XPointer is a huge beast - a good candidate for a 2.0 spec, but enormous
> for a 1.0 spec. I'd be much much happier to see something smaller for a
> first round.
Took me one week to implement the near complete support in libxml
on top of the existing XPath implementation. People from the University
of Bologna wrote 2 implementations in a few weeks, one in JavaScript,
the other one in Java by extending Xalan to support templates to
match XPointer expressions.
http://www.cs.unibo.it/~fabio/XPointer/
And XPointer is small compared to other specifications, quite smaller
than XSLT which reuses XPath too, not that much longuer than for example
RFC 2396, but sure it builds on top of more existing layers. But in itself
it is relatively small.
Some people would like to subset it, right. But the 1.0 to 2.0
evolution is only possible if you keep the parts allowing to version
it (namely the scheme mechanism). Otherwise assuming a non versionnable
pointing spec gets agreed as the fragment identifier for XML (which is
the target of this spec since day one) then you simply cannot make a
version 2.0 later on. Independantly of what to subset, the main problem
is HOW to allow interoperability !
Daniel
--
Daniel Veillard | Red Hat Network https://rhn.redhat.com/
veillard@redhat.com | libxml Gnome XML XSLT toolkit http://xmlsoft.org/
http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/
|
