Re: [lisp] LISP Interworking: Proxy Egress Tunnel Routers

["Joel M. Halpern" <[e-mail]>]

I think that there is a distinction that is important to keep in mind.
First, some assumptions I make:
For LISP to work, it has to be deployable.
Deployability includes the presence of suitable incentives for 
deployment.  That clearly includes appropriate balance of financial 
incentives and disincentives.

For many parts of this system, that is straightforward, and does not 
involve business models.

For PTRs and PETRs there does appear to be a question about what the 
incentives for deployment are going to be.  Who will deploy them, and why?

It is quite reasonable, and I think necessary, to discuss what possible 
incentives there are, who needs to deploy the things, and why they might 
do so.
The distinction I draw is between that and "a thorough set of possible 
business models."  We are not trying to explore the space of business 
models, or determine whether this is compatible with every one.  We do 
not even have to do a full economic analysis confirming that there is a 
business model that covers everything.

Phrased differently, I think we need to convince ourselves that the 
business model has enough solidity that it may succeed.  After that, the 
experiments should help us tell who is interested, and what the actual 
costs are, etc...  Our initial evaluation needs to look at overall 
incentives and overall costs, not details of business models.

Yours,
Joel


Robin Whittle wrote:
> Hi Sam,
> 
> Thanks for this.  Can you or anyone else point me to the RFCs or
> whatever which give guidance on discussing "business models" within
> an IETF WG?
> 
> Irrespective of what those restrictions are, I think there's no point
> in devising a scalable routing solution without a thorough set of
> possible business models to match everything which needs to be
> adopted, changed deployed etc.
...