Re: draft minutes from the sming interim:

[Frank Strauss <strauss@ibr.cs.tu-bs.de>]

Hi!

David> There are several examples of RFCs that contain MIB modules that
David> the ADs know are illegal. And they are published anyway! Why,
David> because illegal stuff was put it, it wasn't checked, agents
David> and managers used it, and to change it to make it legal would
David> be a bigger impact than to fix it. Directives would be one way
David> to flag such illegal usage.

I don't agree. `Legalizing' such illegal constructs by compiler
directives in future MIBs would be bad style. It would be the wrong
sign to MIB authors. It would confuse MIB readers more than
necessary. It would blow up the language unnecessarily. And probably
it would not even be a win for what MIB compilers could do with a MIB,
except turning error messages into warning messages.

What I would like to support is a way to supply annotations to MIB
definitions with added value.

For Standard MIBs I agree that we should not allow any formal
annotations.  I agree that we should not stress the compelxity of the
SMI language more than necessary.  But I believe that vendors often
have additional needs where efficient MIB design and implementation
go hand in hand. I think, the SMI as a standard language should not
be limited to just Standard MIBs. Today there are much much more
vendor MIBs than Standard MIBs and I'm sure that this won't change in
the future. Hence, I vote for a standardized way to support language
extensions, but without standardizing the extensions themselves.

 -frank