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

Re: Target times for MIB Doctor Review

Hi,


I believe David has a couple of points here that the IESG
should consider also (see inline).

On Jan 31, 2006, at 4:47 AM, David T. Perkins wrote:
[snip]

5) Unless you are keeping up with a technology,
   then when you review a MIB module for the technology,
   you must spend some time coming up to speed
   on the technology. This can be a difficult
   because the technology is new and/or changing
   and thus there are no books. It would be quite
   helpful if a MIB reviewer could be provided
   with high level documents and access to a
   technical expert (which may not need be the
   document author) to get a quick tutorial
   and have questions answered.


I guess this is crucial for a proper review. I mostly ended
up first trying to understand the overall technology and then
how that model was mapped onto a MIB definition.


What I've found most in reviews is
1) lack of "deep knowledge" needed to most appropriately
   write MIB modules. That is, the authors know what they
   want to do, but lack knowledge in how to express it with
   constructs in a MIB module.


In general I see indeed that the way people express themselves
in the DESCRIPTION is not optimal.


2) misunderstandings of the limitations SMI that are the
   result of the limitations of the SNMP protocol
3) messed up modelling with complex tables


This I beleive is a result of the MIB editors who understand
very well the technology they are working with but less the
MIB technology. Where MIB reviewers are opposite.

I believe it also might help to get MIB reviewers more involved
from the beginning of the MIB definition. This also gives a
MIB reviewer more time to perform reviews, make corrections
and learn the technology. I remember at some point in the past
'supervisors' were assigned for MIB modules in WGs. This has
for some reason not always worked optimal I guess, but it
also depends on the WG that lost focus of the MIB definition
needed.

therefore, I would suggest an earlier introduction of the
MIB reviewer to get him also up to speed.




Harrie