[Mops] FW: Multicast for the Web (and WebTransport in particular)

["Holland, Jake" <[e-mail]>]

Hi Mops,

Ali just mentioned that the mops group also might be
interested in Web multicast, since video is such an important
use case for it.

It's very late notice, sorry, but anyone who can make it is
invited to attend the Web Multicast bar bof starting in 2
minutes.

Best regards,
Jake

On 07-22, 10:44 AM, "Holland, Jake" <[e-mail]> wrote:

Hi WebTransport,

I'd like to let you know about a new W3C Community group
dedicated to incubating Multicast support for web traffic:
https://www.w3.org/community/multicast/
https://w3c.github.io/multicast-cg/multicast-cg-charter.html

For the first phase, we're working toward an integration with
WebTransport, as the charter describes in more detail.

I've set up a bar bof side meeting at 11am Pacific Time on
Thursday July 29th to discuss further.  Details are attached,
the meeting link is at the bottom, and it's registered here:
https://trac.ietf.org/trac/ietf/meeting/wiki/111sidemeetings

(I'd ask for a slot in webtrans, but hilariously, it's at the
same time as mboned, where some other parts of this work are
under way.)


A little more background:

Akamai is wrapping up some trials and discussions with several
ISPs and content owners evaluating the feasibility of standards-
based multicast delivery for our popular production traffic on a
3-5 year timeline.  The trials were checking whether our
prototypes and specs look like they'll be feasible to deploy
after productizing them, and whether the achievable traffic
offload would make investment worthwhile.

The moderately positive results of these trials (as well as some
interest generated during a W3C Interest group presentation) led
to the creation of the new community group as part of our next
steps.

We'll be relying on progress on some of the IETF docs to clarify
and address the networking requirements for web traffic as it
relates to multicast IP delivery.

The first one to consider from the webtrans point of view is
probably our recently-posted security considerations draft:
https://datatracker.ietf.org/doc/html/draft-krose-multicast-security

The aim is for this doc ultimately to perform a similar function
to what RFC 8826 did for WebRTC.

There are other relevant docs that taken together we think can fill
the security needs we know about so far, with a little refinement:
https://datatracker.ietf.org/doc/html/draft-ietf-mboned-ambi
https://datatracker.ietf.org/doc/html/draft-pardue-quic-http-mcast

I understand this to be outside webtransport's current charter,
but we'd like to develop this work between now and IETF 112, and in
November to bring a coherent and well-supported proposal for next
steps either to a venue like secdispatch or to a BoF for an
appropriate new group, depending on the advice we get along the way.

One option that might be worth considering would be rechartering
webtransport to allow for a scope that includes this work.

But even if that turns out not to be the right thing, we expect
that members of webtransport will have crucial expertise and opinions,
and we would value your feedback and request some level of liaison as
we try to move this forward.

I hope to see some of you next Thursday.

Best regards,
Jake

[1] Webex meeting for Thursday at 11am Pacific (1 hour before the
first sessions of the day):
https://akamai.webex.com/akamai/j.php?MTID=mbe0822775e2ba553d4669b441833ec38