Re: [lisp] Capabilities Type LCAF - a proposal

[Roger Jørgensen <[e-mail]>]

On Wed, Nov 20, 2013 at 1:41 PM, Noel Chiappa <[e-mail]> wrote:
>     > From: Damien Saucez <[e-mail]>
>
>     > I would have a question, if there is capability, it means that there is
>     > a possibility to meet no capability so in this case what is replied?
>     > ...
>     > If 0 locators then it is a negative reply but conceptually this is not
>     > a negative reply. If no answer, then the requester will keep continuing
>     > sending requests.
>
> Good points.
>
> In general, I don't think we should keep tweaking Map-Request and Map-Reply
> messages 'because they are there'. If we need to extend semantics, let's do
> it right.

Keep tweaking existing things to support all sort of new idea only
lead to complexity, and more problem somewhere down the road.

But I do see the need for that Dino describe.


>     > If a new message, what kind of message?
>
> The has been some discussion about adding a new message-type to fill a
> variety of roles which involve transferring information around (e.g.
> dynamically loading configuration information such as default PxTRs into
> small xTRs, a la DHCP, instead of having to manuall configure them).
>
> This would seem to naturally fall into that general classification?

I would support doing this as a new message in the form of a framework
to support existing idea, new idea and future needs. Sort of a version
2 extension. You don't need it, but if you support it an use it you
get access to a whole range of cool and fancy extension of LISP :-)



-- 

Roger Jorgensen           | ROJO9-RIPE
[e-mail]          | - IPv6 is The Key!
http://www.jorgensen.no   | [e-mail]