Re: Methods, Inheritance, Exceptions, etc. (was: Re: Methods in SMIng ?)

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

>>>>> Andrea Westerinen writes:

Andrea> Also, we were supposed to be working on requirements at this
Andrea> stage and then solutions.  It has always been considered poor
Andrea> software engineering to define a solution and after the fact,
Andrea> document its requirements to justify that solution.

A project with a requirements list where nobody has at least an idea
how to realize the requirements is likely to fail or to be endlessly
delayed. We are supposed to deliver the language specification in
September and the mappings in November.

You claimed that it is possible to map generic method invocations to
existing SNMP operations. So please provide a justification for your
claim. (I can only think of ways that are really ugly and which nobody
ever would like to implement or use. But perhaps I am just not smart
enough - that is why your help is needed.)

Note that I am not asking for _the_ solution for SMIng. But if nobody
here on the WG list can provide a convincing idea how to realize
methods on top of the existing protocol operations in a way that is
actually implementable and usable, then I prefer to not put methods on
the requirements list.

/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/>