--0kRkyLZR5zsR9u2P
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Sun, May 21, 2006 at 07:55:23PM -0700, Jason Thorpe wrote:
>=20
> On May 20, 2006, at 4:19 AM, Daniel Carosone wrote:
>=20
> >Like, say, XDR?
>=20
> IIRC, XDR is not self-describing; you have to know what is coming.

This is true, but not entirely.  XDR doesn't represent protocol, just
data forms - and as you say the data forms are not annotated with
in-band type descriptors. =20

However, it's possible (and even suggested in the RFC) that protocols
built using XDR can be: define an enumeration XDR type for all the
other types you might encode, and sequence them together, each XDR
data element described by the preceeding XDR next-element descriptor.
You can of course build trees and any number of other structures in a
similar fashion, up to and including encoding valid proplib xml
tags/nodes and data as your "compact binary encoding".

Or you could pick one of the several other "compact binary xml
encodings" currently vying for buzzword and standards attention,
including at least one that uses ASN.1 for extra shock value.

But all of that is just letting facts get in the way of a good tease :)

--
Dan.

--0kRkyLZR5zsR9u2P
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.3 (NetBSD)

iD8DBQFEcW4YEAVxvV4N66cRAltCAKDcmAHL/aqgrO3Qx8PQTttZKye78gCaAytn
C5cK6zoccjptJxPnlZfdTuE=
=Oa1o
-----END PGP SIGNATURE-----

--0kRkyLZR5zsR9u2P--