Re: application/xml in ietf-xml-guidelines

Keith Moore wrote:
> 
> > Why are HTTP URIs unstable over the long term? Why shouldn't the URI
> > "http://www.ietf.org/rfc/rfc0821.txt" last for 100 years?
> 
> That particular URI might last.  The vast majority of URIs will not.

But if we agree that it is possible to construct URIs that are both
likely to survive for a long, long time, AND be dereferencable using
widely deployed software then what are we arguing about? Why not change
the RFC to say:

The IETF hereby commits to maintaining a namespace rooted in
"http://www.ietf/org/xmlns". This should be used as the root for
IETF-defined namespaces.

> So in general it's not a good idea to trust URIs to be a stable
> reference over a long period of time.

In general no. But just as the "IETF URN" subset of the URI space is
going to be stable because the IETF says so, couldn't the "IETF URL"
subset of the URI space be stable for the same reason?

> ... But you still won't be able to assume good stability for http
> URIs in general.

But who cares about HTTP URIs in general? We're talking about the tiny
subset defined in IETF specifications.

 Paul Prescod