i was not so much seeing this part of the thread as what to build .... but
what were some of the constitute components and driving factors of the
operational infrastructures (aka was it possible for government to mandate
stale, static certificates even if it made no economic sense in a rapidly
evolving online world).

we've had somewhat related activity in the standards privacy working group.
the surface analysis was to take the existing privacy regulation and
legislation and codify it.

the behind the scenes analysis from 1999 was that driving factors in
privacy related regulatory and legislative activity was
1) identity theft and
2) (institutional) denial of service.

There would continue to be a lot of regulatory and legislative activity as
long as there was identity theft and/or denial of service happening
(basically some fundamental economic driving issues). Some amount of this
activity suspended in the wake of 9/11 but didn't disappear. In the recent
march timeframe, the prediction was a lot of the regulatory and legislative
privacy related activity would start to see a lot more action by the summer
.... which appears to be coming to pass.

Which then somewhat gets things back to the subject line of confusing all
kinds of things with identification.

The x9.59 scenario with respect to being agnostic with respect to privacy
is that the integrity of a payment transaction can be significantly raised
at the same time removing any ancillary need for shared-secrets and/or
privacy information in conjunction with the payment.

There was a reference to GSA (a government entitty) resorting to bilaterial
contracts with all of the individual entities (TTP CAs and relying parties)
in attempt to provide stale, static certificates some legal foundation.
Rather than forcing all relying parties to have individual contracts with
each and every TTP CA ... they effectively made all of the TTP CAs agents
of the GSA (via contract) and then every relying party had contract with
GSA. This addressed the requirement for N times M individual contracts (as
in the discussion of some parts of the world ... which scales poorly in
situation where N times M equals 120 billion).

--
Internet trivia, 20th anv: http://www.garlic.com/~lynn/rfcietff.htm

[EMAIL PROTECTED] 6:29/2003 12:51 pm wrote:

I would not try to build a single system that could handle value transfers
for
regular business use and for government payments.

The major reason is the legal liability.  Business contracts typically
involve
civil penalties.  Government mandates, and our responses to them, typically

involve criminal penalties.  In the business case identity is seldom
necessary
for transactions that do not involve real-estate.  In fact the increasing
concern for privacy somewhat mandates that user's can limit the data
transfered
about themselves.  This is where account-based transactions should be
targeted.  In the government case identity is nearly always required by law
or
regulation, and privacy is typically not available.

I believe that payments from purchasers to merchants is the problem that we

have some chance of solving here.  Government payments will be mandated and

will probably not be designed for any of the purposes that business
desires.

Let's focus on what we can effect.


Reply via email to