RE: entstate-651 Notification Description Changes (Part II)

hi

Well, as I said, I can't think of a design pattern to make then
conditionally mandatory. Personally, I'd love to see them mandatory, but I
think we would need strong consensus to make that change at this point.

Do we have it?

Sharon

-----Original Message-----
From: Juergen Schoenwaelder [mailto:[email protected]] On Behalf Of
Juergen Schoenwaelder
Sent: Thursday, December 16, 2004 4:04 PM
To: Chisholm, Sharon [CAR:5K50:EXCH]
Cc: [email protected]
Subject: Re: [Entmib] entstate-651 Notification Description Changes (Part
II)


On Thu, Dec 16, 2004 at 11:04:24AM -0500, Sharon Chisholm wrote:

> Missed part of this ....
> 
> <Juergen>
> p) The notification are optional. Should there be any further 
> guideline under which conditions implementations should support them?
> </Juergen>
> 
> I don't think so. They probably should have been mandatory but 
> optional was more a political decision. I think it is an implement 
> them if you find them useful sort of thing rather than a design 
> pattern that implementations could test against.

I dislike totally optional notifications in standards. Either these
notifications are generally useful or they are useful in specific situations
which then need to be spelled out or they are not useful 
at all. How do I practically find out whether a given box supports these
notifications? Is the answer that I have to physically remove decives are
beg the vendor to give me a suitable agent capabilities statements? Sounds
all not very practical from a network management application writer point of
view...

/js

-- 
Juergen Schoenwaelder		    International University Bremen
<http://www.eecs.iu-bremen.de/>	    P.O. Box 750 561, 28725 Bremen, Germany