[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Methods in the NIM requirements


In your thinking, there are some cases where attributes are best interface
and other cases where methods are the best interface. Are these mutually
exclusive, or overlapping?



> -----Original Message-----
> From: Harald Tveit Alvestrand [mailto:Harald@Alvestrand.no]
> Sent: Wednesday, April 19, 2000 2:58 AM
> To: Weiss, Walter; nim@ops.ietf.org
> Subject: RE: Methods in the NIM requirements
> At 02:16 19.04.2000 -0400, Weiss, Walter wrote:
> >If I understand you correctly each interface should be 
> represented in a
> >method. Therefore, each interface should be represented as a 
> methods. In
> >addition, each interface should also have a set of 
> attributes that provide
> >an alternate interface. Is that what you are saying?
> I don't understand the grammar above - "interface in a method"?
> but generally, whatever works for the model in question.
> in some cases, methods only work best, IMHO.
> in some other cases, methods + attributes work best, IMHO.
> I still haven't seen a case which I wanted to model with 
> attributes only.
> When USING the model, one can think of implementing the methods as 
> attributes that have side effects (such as SNMP "magic" 
> variables), or 
> implementing the attributes as accessor methods (which in 
> many programming 
> languages (not SNMP) give you much more power of access 
> control, syntax and 
> consistency checking).
> But the tool used for modelling shouldn't enforce an implementation.
>                        Harald
> --
> Harald Tveit Alvestrand, EDB Maxware, Norway
> Harald.Alvestrand@edb.maxware.no