Arguing Crypto: The Engineering Approach

[kentborg@world.std.com (Kent Borg)]

Who knows how well it plays, but my faxed letter to Moakley on crupto
export took a different approach from my usual privacy tirades.

I tried to (calmly) argue that we need strong cryptography for
fundamental engineering reasons.

Data is so damn flexible.  This is both good and bad.  The bad is that
data can be capricious and flighty.  If our physical world were to
allow objects to appear out of no where and disappear again,
transmorgraphy beyond recognition, or become massively duplicated in
unknown locations, we would find it disconcerting--to say the least.
We would go to considerable lengths to keep physical objects
reassuringly in one place.  In fact, with some physical objects that
often do carpiciously vanish, we go to inconvienient extremes in hopes
we can prevent the vanishing.  Look at some of the anti-theft devices
people will put on their cars.

There are good engineering approaches which can force data to behave
itself.  Many of them involve cryptography.  Our government's
restrictions on crypto limit our ability to build reliable computer
systmems.  We need strong crypto for basic engineering reasons.


Note, my fax to Moakley was phrased (and spelled?) considerably
different from this posting.  I am still wondering how best to make
this argument.  Something I want to avoid is too strong a reliance on
"pulling rank": "We are professionals, we need these tools to do our
jobs, don't try to understand the reasons.".  Just using words like
"engineering" smacks of that enough, let's leave it at that.

One thing I like about this approach is that is avoids the kneejerk
positions the word "privacy" prompts.

-kb, the Kent who tries to sound reasonable


--
Kent Borg                                                  +1 (617) 776-6899
kentborg@world.std.com                                
kentborg@aol.com                                      
          Proud to claim 29:45 hours of TV viewing so far in 1994!