Ira,
This isn't exactly a stone, more like a pebble, but, how may more bit
vectors do we need to add before we throw up the white flag and say this
attribute method of MIB specification was not a good idea? I can
immediately come up with at least one more useful bit vector. This one will
let us know which attributes actually are currently multi-instanced. Of
course none of the bit vectors is ever going to answer the question "What is
the meaning (significance, point, ???) of each instance without any given or
know previous knowledge?".
Mike
> -----Original Message-----
> From: Ira McDonald [mailto:imcdonal at sdsp.mc.xerox.com]
> Sent: Saturday, June 26, 1999 3:54 PM
> To: Joe.Filion at usa.xerox.com; Mike.Elvers at usa.xerox.com> Cc: Angelo.Caruso at usa.xerox.com; hastings at cp10.es.xerox.com;
>jmp at pwg.org> Subject: RE: JMP> FW: Document Action: Job Monitoring MIB - V1 to
> Informat ional
>>> Hi Mike,
>> (don't throw stones yet...) What if we added a bit vector to EACH
> job that says 'one or more instances of this job attribute are
> CURRENTLY instantiated on this job'? Maintaining this bit vector
> (a parallel subset of the support vector at all times) doesn't
> seem that costly to an agent and it *seems* to help with your
> concerns.
>> Thoughts?
> - Ira McDonald
>>