All things related to multicast

["Garry Peirce" <>]

While I agree interdomain multicast has not been used as much as envisioned,
I'd rather keep it in the toolbox.
It can require more time to work with, but from my perspective, it's
certainly stabilized over the past few years.

Also as one in the R&E realm, I think it's suffered from being under
supported and promoted.
We could help generate content but there are always higher priorities.
However, I think R&E's should retain such capabilities as a function of
their role in researching/enabling the network as a resource itself.
In the commercial realm, perhaps the economics of multicast are more the
reason for it's absence.

Interested in how v4 interdomain issues are solved by AMT - by converting a
mcast streams to unicast? 
That would seem to be avoiding it rather than improving it.

I also agree mcast use within 802.11 is different,  and although I've not
yet enabled it myself for other reasons, there have been improvements here
too, with vendor C's implementation looking more AMT-like.

Had to laugh at this Nov 2012 article's first paragraph...mentioning mcast
as 'an innovative solution'!
http://www.v-net.tv/multicast-nanocdn-might-redefine-streaming-economics/

Perhaps I'm just a lorax -   "..who speaks for the trees against the greedy
Once-ler"    ;-)


> -----Original Message-----
> From: 
> 
>  [
> ]
>  On Behalf Of Bill Owens
> Sent: Thursday, December 20, 2012 3:49 PM
> To: 
> ;
>  
> 
> Subject: Re: Whither multicast?
> 
> I had written a much longer note about the current state of interdomain
IPv4
> multicast, from our vantage point, but. . . it doesn't really matter.
> 
> The point, I believe, is that the critical period for widespread adoption
of
> interdomain native multicast has passed. I don't think it matters to talk
about
> IPv4 or IPv6; either way there are still considerable issues associated
with
> building, maintaining and troubleshooting interdomain multicast
connectivity.
> The only important difference is that IPv4 multicast seemed to have some
> promise, and by dint of great effort we were able to get some adoption in
> the R&E community. I don't see that happening with IPv6 multicast.
> 
> Even though the train has left the station there are still legitimate uses
of
> IPv4 multicast; we aren't going to turn it off.  And when we start talking
with
> IP equipment manufacturers in a month or two, trying to decide what could
> be the basis for an upgraded NYSERNet R&E network, I'm going to have
> requirements listed for IPv4/IPv6, unicast/multicast. But if a vendor
comes to
> the top of the list with everything else that we want except interdomain
IPv6
> multicast, it won't stop us from buying their gear. . . and even if they
do have
> excellent support, I'm not sure we'll turn it on.
> 
> Bill.