In your previous mail you wrote:
> => I understand but I shan't support the c) option until the 5-tuple
> re-classifier monster is killed.
In fact the c) option is not using the 5-tuple classifier.
=> in your question you didn't specify if c) uses it or not.
So I do not understand why you link them together.
=> c) should be at least split in c)+PHB ID-like and c)+Appendix A-like.
I don't believe the first c) is a complete solution for diffserv,
and the second c) just triggers the issue with MF re-classifiers on the
5-tuple anywhere.
But the M-F Diffserv classifier is not different than the Intserv data
path classifier.
It uses the same fields, with the same type of capability to represent
range of values by wildcarding some of the bits, or an entire field.
=> devices are the same, ways to use them are very different.
How come you do not have a problem with the Intserv classifier?
=> yes because I can use for instance flow labels in order to
reduce an Intserv 5F classifier to a simpler one (making the assumption
for the discussion that Intserv is usable in the core).
Regards
[EMAIL PROTECTED]
--------------------------------------------------------------------
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]
--------------------------------------------------------------------
