Path: utzoo!utgpu!news-server.csri.toronto.edu!clyde.concordia.ca!uunet!samsung!usc!ucsd!ucbvax!CHEETAH.NYSER.NET!mrose
From: mrose@CHEETAH.NYSER.NET (Marshall Rose)
Newsgroups: comp.protocols.tcp-ip
Subject: Re: Efficiency (or lack thereof) of ASN.1. Was: Re: What is the IAB?
Message-ID: <16293.644974816@cheetah.nyser.net>
Date: 9 Jun 90 23:40:16 GMT
References: <12058@asylum.SF.CA.US>
Sender: daemon@ucbvax.BERKELEY.EDU
Reply-To: tcp-ip@nic.ddn.mil
Organization: The Internet
Lines: 31

Karl - we've reached a circular discussion here.  The SNMP PDUs were
written in ASN.1 so as to be as fixed and predictable as possible.  That
is one method that was used in order to help out people optimizing their
encoders and decoders.

I suppose the PDUs could have been written in such a way as to use more
of ASN.1's expressive-ness, but the only benefit would be to be more
elegant in an academic fashion with no gain in functionality.  Of
course, the drawback would be that it would be nigh impossible to do
optimizations.

Sure, SNMP could have used a special-purpose, roll your own encoding
algorithm.  And sure, it could be faster than any form of ASN.1 known to
man; probably an order of magnitude faster, if you were very, very lucky.

And what would that cost exactly?  Well, let's see, it would be yet
another special-purpose thing useable only for one task; it would have
associated with it the usual debugging cost for the designers; it would
also have the usual start-up and training costs for the people writing
implementations.  So it would have a lot of hidden costs that would slow
development and hinder deployment.  (And to add insult to injury it
would still need to have things like ASN.1 object identifiers to provide for
decent naming and extensibility.)

Now I hate to be the one to point this out, but technological issues are
hardly boolean.  Any time someone says "X is better than Y", they have
neglected to state the criteria associated with that judgement.  As
such, there is more to consider than just raw speed when discussing this
issue. 

/mtr