prtChannelMagicCookie object

prtChannelMagicCookie object

prtChannelMagicCookie object

JK Martin jkm at underscore.com
Mon Aug 26 13:43:00 EDT 1996


Jeff Dunham's comments (via Bob Pentecost) are very insightful and,
if nothing else, a good overview of the complexities of trying to
come up with a complete and coherent architecture to solve all known
problems.


At one point, in response to Dave Kellerman's statements about the
benefits and pitfalls of incorporating the NSM MIB as a partial
solution to the problem, Jeff stated:


> Danger!  Need to keep it simple!


Exactly.  However, given the problem(s) we're trying to solve here,
"simple" doesn't seem to be an alternative if what you're trying to
accomplish is (as Jeff states):


   "...a hierarchy that describes the protocol-path of the data from
    the frontplane to the backplane (and for that matter, all the way
    to the formatter:  PS vs. PCL, etc)."


I totally agree that this kind of design is needed *really* solve the
complete problem, yet "simple" is about all the PWG can do right now.


Regarding Jeff's comment about the object name:


> This is a nitty comment, but "MagicCookie" is goofy.  I know its a "Unix
> thing", but the MIB should be a professional document, not one that
> talks about "Magic Cookies".


I believe Dave Kellerman used the name "prtChannelMagicCookie" merely
as a placeholder until the PWG came up with a better name.  However,
to date, no one has proposed a better name.  (I wish someone would!)


During the last PWG meeting (in Portland) when this topic was discussed,
at one point I had stated I was "withdrawing my request" for adding this
kind of info to the MIB, since the resolution always seemed to flounder
indefinitely.  However, the group came back and made a firm statement
that something (anything!) should be done to accomodate a very simple
capability for a generally pervasive problem.  As a result, the group
came to RAPID CONSENSUS on the concept of a simple "MagicCookie" idea,
and Dave Kellerman did a fine job in capturing that thought in his
proposal.


So, it looks like it could come down to a philosophical question of
whether the PWG is either willing to support a simple pragmatic approach,
or insist on a 100% solution comprising a fully elegant architecture.


	...jay



More information about the Pwg mailing list