On 2007-10-05 05:38, ken carlberg wrote:
I don't recall when was the last (Diffserv-based) QoS talk at NANOG or
similar operator-rich meeting. (Sure, there is the tutorial, but it
doesn't count.)
I would be concerned if outside groups spent time arguing "foo" is bad,
or if they advocated other positions to the same issue. But I tend to
feel quite uncomfortable with litmus tests based on inactivity of other
groups/people. My personal view is that advocates of that line of
reasoning place a bigger burden on themselves in providing specific
in-depth arguments.
Seems like a potential indication that most typical ISPs aren't
working on or interested in this, this stuff is so trivial, or that
coordination is not necessary.
i appreciate work that is trivial because its generally simple, easy to
accomplish, and leads to fewer interoperability issues. as for ISPs,
its fascinating the disparity of how quiet and talkative they are
depending on what side of the NDA you are on :-)
In any case, if Pekka is correct, that's *exactly* why this
draft and RFC 4594 are needed - to lay a minimum foundation on which
ISPs can build operational practices and SLAs.
It's always been clear to me that voice and video would be the main
drivers for uptake of diffserv, and Marshall's comments confirm
that. As that type of traffic grows, ISPs won't have any choice.
Guidnace from the IETF seems entirely appropriate.
Brian
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf