In your previous mail you wrote:
> => I agree, the whole stuff about MF re-classification on
> the 5/6 tuple seems silly to me.
People do much worse things already, by looking even deeper into
the packet.
=> I know... I'd like to see how this will work at 10Gbits/s.
On the principle I don't want to have ISPs that I don't trust
to look at the transport header or the payload of my packet.
Of course, if the answer is to use ESP we have only shown
the silliness of the whole thing!
In practice, I agree that port numbers aren't too much use,
and probably very often only the addresses will be used. The whole
idea Alex and I had was to give IPv6 an advantage by using the flow label
to support more effective MF classification.
=> this is what I have understood so even you proposed a PHB ID in fact
alternatives in appendix A are still options...
My concern is that if the MF re-classification on the 5/6 tuple or
worse is given up then your idea will be given up too: should we
discuss about the provision of effective MF classification or
discuss about the MF classification itself?
Regards
[EMAIL PROTECTED]
PS: this is an architecture issue, I don't know how to solve it
(ask the IAB to move diffserv to historic?)
--------------------------------------------------------------------
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]
--------------------------------------------------------------------
