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

[Jon Saperia <saperia@jdscons.com>]

> 
> >>>>> Jon Saperia writes:
> 
> Jon> Juergen, it is because I think there is not a decent map for
> Jon> methods in the protocol operations, that I suggested that without
> Jon> coordination with EOS, an effort here on mehtods would be
> Jon> sub-optimal.
> 
> I think we are more or less in agreement.
> 
> I have always been arguing that adding methods to the SMI without
> specific protocol support does not make sense.

We agree.
> 
> Some people however argue that you can do methods with the current set
> of protocol operations, which I believe is really really ugly and
> where I would like the proponents of this idea to show at least an
> outline how this could work in practice.

I agree again. While I would like to have methods, I just do not see how
to do it very well without protocol support. If a proposal that works
can be made that is great - I just do not see how it would be done very
easily.

> 
> Regarding EOS: The EOS WG is not chartered to provide support for
> methods. So they first have to deliver what they are chartered for
> before method support may officially become an EOS work item (if I
> understand the message from the ADs correctly). Hence, I have been
> arguing to do methods later as an extension to SMIng in order to move
> forward on the stuff that is available/being worked on right now.
> 
Seems quite reasonable, I was not really making a charter change request
for either group. I did want to observe that methods especially need
help from the language and the protocol. On thing might be to let EOS
get going as you say, then in 6 months or so put this on everyones plate?
> 
> 

Thanks,
/jon
--

Jon Saperia		     saperia@jdscons.com
			     Phone: 617-744-1079
			     Fax:   617-249-0874
			     http://www.jdscons.com/