RE: [IP-Optical] RE: GMPLS Last Calls

["Mannie, Eric" <Eric.Mannie@ebone.com>]

Hello George,

I think that Dimitri wanted to say that GMPLS is suitable for a large set of
scenarios (not models in your understanding). Anyway, I think that you
understood very well what he wanted to say...

Rgds,

Eric

-----Original Message-----
From: George Newsome [mailto:gnewsome@lucent.com]
Sent: Wednesday, May 30, 2001 3:29 PM
To: Dimitri Papadimitriou
Cc: ccamp@ops.ietf.org; mpls@uu.net; ip-optical@lists.bell-labs.com;
tsg15q11@itu.int; t1x15@t1.org
Subject: Re: [IP-Optical] RE: GMPLS Last Calls


Dimitri,

I'm probably just showing my ignorance, but I'm rather curious about
your statement that GMPLS is not based on a model, "Because from the
beginning it is a model independent specification." 

Likewise you state that by defining a profile, GMPLS can be applied at
any interface. Together these statements implies that GMPLS can meet any
model and any set of requirements.

This sounds too good to be true, and I am reminded of the old advice
that when something looks too good to be true, it probably isn't.

Could you clarify what a model independent specification is, and how a
profile can enable any requirements to be covered?

Thanks

	George Newsome

Dimitri Papadimitriou wrote:
> 
> .......................
> We don't taylor a
> protocol suite based on the model to which we have to apply it since
> if the model change - you have to re-design your protocol ! -. Today
> GMPLS can be accomodated to any UNI/NNI interfaces coming from
> wherever you want. Why ? Because from the beginning it is a model
> independent specification ! By defining the profile you desire you
> can apply it to the OIF UNI / ASON UNI (if still needed) / etc. Also,
> don't spend to many time at the OIF to define the architecture it is
> already done...
> 
> Thanks,
> - Dimitri.
>