RE: DAV-Enabled field (was RE: A case for GETSRC)

The bottom line here for me is that the source is very different
from the generated output from that source.  Users care deeply
about this difference.  This means that you have to make it easy to
pass around, from application to application, which one you want.

Now we could require every application that passed around a reference
to be upgraded to understand a set of syntactic extensions that would
tell each client along the way what headers, or what alternative
methods, should be used to get the contents or otherwise manipulate
the intended meaning of that resource, *OR* we could require that
a server come up with a different URL, so that all of the existing
URL passing machinery (e.g. pasting the URL in an email message)
continues to work.  Given the relative cost of getting consensus on
these syntactic extensions (and getting all applications/users to
adhere to them) vs. the cost of generating a source URL, the answer
to me is clear.

There are some points that are worth addressing, such as "how can a
server tell a client that a corresponding tree of authoring resources
is available".  I (and I'm sure Julian) would agree that this is the
kind of thing we need to fix about the DAV:source mechanism.  In this
case, a straightforward approach is to define that the DAV:source of
a collection is another collection containing all the sources for
members of that collection.

Cheers,
Geoff

Received on Wednesday, 6 March 2002 09:53:30 UTC