The IP over ATM Mailing List Archive by date

Cell Relay Retreat>List Archive>month:1995-May> msg00065



[Date Prev][Date Next][Thread Prev][Thread Next]  
  [Date Index][Thread Index][Author Index][Subject Index]

If broadcast is a special case of multicast, then....Note from SMTP2 at IINUS1

  • From: "Timothy J. Smith" <tjsmith@vnet.ibm.com>
  • Date: Mon, 8 May 95 09:03:59 EDT
  • CC: ip-atm@matmos.hpl.hp.com

Curtis,

>>I definitely do not want to see the draft
>>stating that every host automatically joins the broadcast group and
>>gets the benefits of the packets sprayed on the LIS by badly written
>>applications whether it is going to use them or not.

Your point is a good one, but we have folks pulling hard on
the other side as well.  Others have contacted me off-line
stating that broadcast is more important to their development
than multicast.  I believe their interest is to change as
little as possible to add ATM.  And the
point was made recently that they don't want to have to register
at all to be a part of the bcast, that it should be automatic.

There really is no way to stop a determined bcast application.
One could simply ARP for every address in the LIS and build his
bcast tree.  So, I think it is going to happen.  I would prefer
to get an architecture that people think is reasonable rather
than leave it up to random implementations.

Do others on the mailing list have an opinion?

Cheers,
TJS