RE: xml-include published early for April 1st??

Liam R. E. Quin wrote:

> Is this serious??
> 
> The excuses about not having used entities because the DTD is too
> complicated and not using XLink because it's more fun to invent a new
> standard than use an existing one seem to me to show that people have
> too much time on their hands.
> 
> If XLink doesn't do what the groups using it need, FIX IT.

An XLink-based solution would likely not be significantly different, other
than potential minor syntax changes.  XInclude is the result of exploration
of the necessity and functionality of inclusion in the XLink WG.  Since
linking and inclusion are substantially different, modularization was deemed
desireable, leading to separate specifications.  If you have more ideas on
how XLink could be "fixed" and the costs of not doing so, we'd love to hear
them.

> Don't invent new standardsjust for the fun of it.

Believe me, standards are not created for the fun of it :-).  If you have
feedback on specific requirements or find the whole idea unnecessary, your
rationales would be valuable for us to consider before we get much further
into it.

> I really hope this "standard" goes away.
>
> Will an XML processor in 2 years' time have a list of hundreds
> of "magic' URLs that define namespace-specific behaviour that
> ccould not be deduced by inspection, are cryptic, and interact
> in subtle ways?

The intent is not to introduce a new namespace URI for inclusion, but to use
the xml namespace.  But it is hard to tell whether you are objecting to
XInclude, or more generally to evolution of XML.

> What happened to XML being simple?

I can't quite distinguish whether this is a rant or if there are specific
architectural issues for us to consider.

- Jonathan Marsh

Received on Monday, 27 March 2000 13:56:17 UTC