Re: Comments on draft-morin-l3vpn-mvpn-considerations-01.txt

[Eric Rosen <erosen at cisco.com>]

Thomas  raises  a lot  of  issues,  which  deserve discussion  (probably  in
separate threads), and which will take a bit of time to respond to.

But there're two issues I want to respond to right away:

Eric> I would  like to  see the  draft offer much  stronger support  for its
Eric> recommendation to implement outsourced  RPs, or else to eliminate that
Eric> recommendation.

This is addressed at the text from section 4 of the draft:

        "It is therefore the recommendation of the authors
        that implementations should support a co-located RP model"

Thus it seems  to me that the draft recommends that  the co-located RP model
be implemented, and I think I am within my rights to oppose the inclusion of
that recommendation.

Thomas> Everybody having read the draft can see that it actually not only say
Thomas> that the RP-colocation feature is useful but also that an implementation
Thomas> should _not_ require using this feature

That's a  completely different issue, I  never said that  the draft requires
the co-located RP feature to be deployed.

Thomas> your  underlying suggestion  that co-authors  of mvpn-considerations
Thomas> state  that "the  outsourced-RP  feature should  be implemented"  to
Thomas> simplify things for some particular vendor(s)

I apologize if my words led you  to make that inference, but that was not my
intended  implication.   I  do  think  that  the  outsourced-RP  feature  is
primarily a  simplification for the  vendor, and really just  a complication
for the  SP and its customers.   However, my assumption is  that the authors
have  an  honest disagreement  with  me, not  that  they  have any  ulterior
motives.

Thomas> operators have just a sincere  interest in trying to identify a best
Thomas> candidate for becoming a standard (vs. a set of "profiles")

At no time did  I question the sincerity of the authors,  nor have I accused
anyone  of attempting to  spread FUD,  of telling  lies, or  having ulterior
motives.  I hope you will agree that ad hominem attacks of that sort have no
place on this ML.  I suppose  I have accused people of not fully considering
all the issues, but that does seem appropriate for this ML.

Thomas> If you happen to disagree with this perspective

I don't disagree about the sincerity  of the authors, but I don't think it's
either possible  or desirable at this  point to dispense with  the notion of
profiles.