Re: Proposed charter, please comment

[John Martin <jmartin@netapp.com>]

Mark: I think that we need to add the various multicast groups to the list 
of other related groups that we need to keep an eye on.

Whilst I'm not keen on embracing the complications of inter-domain 
multicast - which has been an issue for many years before CDNs existed - 
I'd prefer us not to do work which specifically excludes it. As far as I 
can tell, we should be able to design at least the request-routing - and 
possibily even the distribution requirements - so that the underlying 
transport may or may not be multicast.

As for streaming, most of it today is unicast. And to further complicate, a 
large proportion is actually delivered over HTTP (usually because of 
firewalls).

The fact is that continuous media - whether multicast, unicast using a 
control protocol, or embedded in HTTP - places different requirements on a 
CDN than traditional HTTP objects. For that reason, we need to explicitly 
include it in our charter.

John

At 02:14 PM 26/01/01 -0800, Maciocco, Christian wrote:
>Streaming kind of implies multicast distribution.
>Will multicast support be required only for streams distribution or for
>other parts of CDNP, e.g. re-direction, accounting?
>Christian
>
> > -----Original Message-----
> > From: Kobus van der Merwe [mailto:kobus@research.att.com]
> > Sent: Friday, January 26, 2001 10:41 AM
> > To: tme@21rst-century.com
> > Cc: John Martin; Abbie Barbir; Mark Day; cdn
> > Subject: Re: Proposed charter, please comment
> >
> >
> > Like John I have always thought of RTSP as not part of "Web-based" but
> > have recently learned that some people consider Web-based to also
> > cover streaming protocols (i.e. RTSP and possibly MMS).
> > For my money that is fine but I would like to see streaming explicitly
> > included in the charter.
> >
> > Kobus
> > Marshall Eubanks wrote:
> > >
> > > John Martin wrote:
> > >
> > > > Noone is saying that we should be psychic about new
> > protocols. I'm talking
> > > > about the apparent *restriction* to "Web-based" protocols.
> > > >
> > > > RTSP is certainly "existing" but, by my definition, it is
> > certainly not
> > > > "Web-based". To me "Web-based" == HTTP.
> > > >
> > > > John
> > >
> > > RTSP has a very html like interface, and is designed to be
> > used from browsers. FWIW, I would consider it
> > > web based.
> > >
> > > >
> > > >
> > > > At 01:02 PM 26/01/01 -0500, Abbie Barbir wrote:
> > > >
> > > > >I do agree with mark here, I think we should try to ride
> > on existing
> > > > >protocols.
> > > > >
> > > > >Regarding the charter, it looks fine to me
> > > > >
> > > > >Abbie Barbir
> > > > >Nortel Networks
> > > > >
> > > > >  -----Original Message-----
> > > > >From:   Mark Day
>[<mailto:markday@cisco.com>mailto:markday@cisco.com]
> > > >Sent:   Thursday, January 25, 2001 11:45 AM
> > > >To:     John Martin; cdn
> > > >Subject:        RE: Proposed charter, please comment
> > > >
> > > > > >A content network is an architecture of Web-based network
> > > > > >elements, arranged
> > > > > >for efficient delivery of digital content.
> > > > >
> > > > > Why Web-based? Does that exclude non-HTTP objects? Actually, if you
>just
> > > > > remove the phrase "Web-based", this reads just fine (to me).
> > > >
> > > >While technically correct, I worry that this moves us in the wrong
>direction
> > > >(increasing generality).  I consider it quite important that the
>working
> > > >group focus on the problems of delivering content within the frameworks
>of
> > > >DNS, HTTP, RTSP, etc. as they exist now.
> > > >
> > > >For example, I think it would be a tremendous waste of time to use this
> > > >group to develop a system for interoperation of content networks that
> > > >required replacing the existing protocol infrastructure of the web.
>That
> > > >could well be a worthy and fun project, but it would be a research
>project
> > > >as opposed to a standards project.
> > > >
> > > >Now, I'm pretty sure that you're not trying to get us to do that, but I
>want
> > > >to be sure that the charter rules it out.  I would be happy with more
> > > >careful phrasing, and perhaps we need to catalog the things that we
>will not
> > > >do instead of relying on the fuzzy phrase "Web-based." But I am wary of
> > > >broadening the scope to any arrangement of network elements that
>delivers
> > > >content efficiently.
> > > >
> > > >--Mark
> > > >
> > > >
> > >
> > > ---------------------------------------------------------------
> > > Network Appliance           Direct / Voicemail: +31 23 567 9615
> > > Kruisweg 799                               Fax: +31 23 567 9699
> > > NL-2132 NG Hoofddorp               Main Office: +31 23 567 9600
> > > ---------------------------------------------------------------
> >
> > --
> >                                  Regards
> >                                  Marshall Eubanks
> >
> > T.M. Eubanks
> > Multicast Technologies, Inc
> > 10301 Democracy Lane, Suite 410
> > Fairfax, Virginia 22030
> > Phone : 703-293-9624
> > Fax     : 703-293-9609
> > e-mail : tme@on-the-i.com     tme@multicasttech.com
> >
> > http://www.on-the-i.com http://www.buzzwaves.com

---------------------------------------------------------------
Network Appliance           Direct / Voicemail: +31 23 567 9615
Kruisweg 799                               Fax: +31 23 567 9699
NL-2132 NG Hoofddorp               Main Office: +31 23 567 9600
---------------------------------------------------------------