Folks

I urge all to read:

draft-itojun-ipv6-flowlabel-api-00.txt
draft-metzler-ipv6-flowlabel-00.txt

I believe these two pieces of work have great synergy and unify 
the ideas I believe we need to pursue to select what the 20bit
flow label should do for IPv6 and get us on the right track.

I am traveling and also want to say I fully agree with Christian's
mail to Brian regarding the role of setting the diffserv bits.

I also do not believe that the orginal thoughts early in SIP and IPv6
regarding the flowlabel time has passed but is now with 3GPP more than
ever needed than before to be used by the client and set by the client.

I also know this works on our implementation of IPv6 and we will be
showing it at ITU Asia I think with our RSVP demo of v4 and v6 to the 
Asian Telco community and others, but I need to verify it made the demo
/ interop code drop.  IF not here definitely at the next show.

I also know the code very well and Itojun's draft would make the
programming of flows even more elegant than what we did.  Also to look
at the IPv4 code for RSVP to hack into the payload to determine the 
5-tuple vs IPv6 just using the flowlabel and srcaddr is no comparison at
all.

Also to Brian:  RSVP does define the use of the flowlabel for IPv6 and
does specify how to use it and Itojun's statements to you in that
context are axiomatic.    

my .02/cents
/jim
--------------------------------------------------------------------
IETF IPng Working Group Mailing List
IPng Home Page:                      http://playground.sun.com/ipng
FTP archive:                      ftp://playground.sun.com/pub/ipng
Direct all administrative requests to [EMAIL PROTECTED]
--------------------------------------------------------------------

Reply via email to