Comments on draft-ietf-mboned-ieee802-mcast-problems

Hi mboned,

After the hackathon yesterday (in which I worked on AMT service
discovery via DNS-SD for a home network) David Lamparter came
to introduce himself and credibly inform me that the multicast
WiFi problems we see are largely an implementation problem with
the Linux bridge code, rather than an inherent problem with WiFi.

He further let me know that several proprietary implementations
do indeed have an approach that does NOT suffer the same
performance problems with multicast over WiFi.

Also that there was a Linux patch a few years ago attempting to
fix it, which was declined because it made the normal bridge path
slightly slower, but if I understood correctly, he believes it's
fixable, and that it would be a more useful approach than tunneling
unicast.

This seems worth following up on, and it might  dramatically
change the contents of draft-ietf-mboned-ieee802-mcast-problems,
after appropriate consideration.

I wanted to let you all know about this development, and to
introduce David and say thanks for raising this.  I look forward
to some further discussion on this point.

Cheers,
Jake

_______________________________________________
MBONED mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/mboned