From: Bill Stewart <stewarts@ix.netcom.com>
Date: Tue, 3 Oct 95 11:37:47 PDT
To: Carl Ellison <cme@TIS.COM>
Subject: Re: Certificate proposal
Message-ID: <199510031838.LAA27571@ix2.ix.netcom.com>
MIME-Version: 1.0
Content-Type: text/plain


At 10:21 AM 10/3/95 EDT, Carl Ellison <cme@TIS.COM> wrote:
>The whole issue of CRLs is on shaky ground with me.  I think it's gotten
>lost in debates about how to distribute them offline (or perhaps via
>e-mail) and have them work.
No surprise; it's a tough job.  I don't expect my laptop to be
continuously on-line when I'm reading mail, and even on-line use
needs to be more realistic than saying "Everybody must use X.400".

>CRLs are like the old credit card or check stop lists which used to be at
>every supermarket checkout station.  They aren't there any more.
>Checkout stations are now on-line.
>I see nothing wrong with having a "certificate" which says ``certificate
>available online only at xxx@yyy.zzz''.

Interesting.  Aside from the excessive cost of looking up cards in paper
booklets,
which can be avoided by having checkstands validate on the store's backroom
computer system, on-line verification lets the card company not only
refuse stolen cards, but also dynamically refuse cards that have reached
their credit limits, which you can't do on a slow push-based offline system.

However, it's _very_ tough to spoof a credit-card verification system, 
because the checkout device uses phones or private networks to reach the 
authorization company, so the response that comes back saying yes/no/stolen
can be real dumb.  On the internet, the response needs to be signed,
though I suppose it could say "Key sssss at xxx@yyy.zzz authorized
key uuuuu today yyyy/mm/dd/hh:mm:ss, valid for up to $500", and you'd then have
to validate the key that signed it, etc....  On the other hand, you could
have the cert require multiple confirmations, e.g. both the bank and the
user have to authorize this use.

>>3) Neither PGP nor X.509 (as documented in the RFC1422 and PKCS#6) have any
>>	   mechanism for additional information other than cramming it into
>>	   the username, but supposedly X.509 Version 3 includes something?
>Yup -- and it's ugly.  It counts on a defined OBJECT ID to define the

OBJECT IDs solve two problems - one is that you need some kind of format (yuk),
but the other is that fields to have mutually agreed on values to be meaningful,
and central registration is an easy way to implement it, as long as there's
a simple way to register things.  I hope it at least supports an OBJECTID
with parameters, e.g. "CreditLimitUSDollarsBankFoo integer"  rather than needing
excessively many OBJECTIDs "CreditLimit3700USDollarsBankFoo"?  As you say below,
there's certainly no need to use ASN.1 formats instead of readable ones...

> What's invalid is the assumption that
>there must be a relationship (or even a person) in physical space *before*
>one can have a relationship (or a person) in cyberspace.

Yeah.  I've decided, as an experiment, to start signing keys for pseudonyms,
though I haven't settled on how to deal with unauthenticated signatures
for realspace people (in the one case where I've been asked, the person
didn't have any independent signatures from other people, so so far I've
declined,
but I may re-evaluate and just do uniqueness.)

#---
#                                       Thanks;  Bill
# Bill Stewart, Freelance Information Architect, stewarts@ix.netcom.com
# Phone +1-510-247-0664 Pager/Voicemail 1-408-787-1281
#---