Re: Methods in the NIM requirements

[Juergen Schoenwaelder <schoenw@ibr.cs.tu-bs.de>]

>>>>> David Harrington writes:

David> I would consider it inappropriate to start from scratch with an
David> information model and not consider all the IETF standards work
David> that needs to be able to map to that information model.

Yes, this is IMHO very important and I was arguing along this line
before. However, we also need to look into the future and I think that
we need methods to significantly simplify interactions. Sure, this
makes algorithmic mappings more complex and we need at some point to
answer the question whether

(a) we try to map to existing technology which may turn out to be in
    the general case either undoable or at least too nasty to explain
    to the outside world, or

(b) we evolve the technology itself along the creation of the NIM to
    make the mappings simpler.

The answer to this IMHO very fundamental question is the time scale of
this work item we have in mind. I believe that any successful NIM work
(which at some points reaches the break even point) will take years
since it also implies work on the various technologies we have. If we
strictly do (a) and limit NIM to the least common denominator, then I
am sure we will restart in a few years from now.

We need to move ahead (making those steps we think we understand).

/js

-- 
Juergen Schoenwaelder      Technical University Braunschweig
<schoenw@ibr.cs.tu-bs.de>  Dept. Operating Systems & Computer Networks
Phone: +49 531 391 3289    Bueltenweg 74/75, 38106 Braunschweig, Germany
Fax:   +49 531 391 5936    <URL:http://www.ibr.cs.tu-bs.de/~schoenw/>