Ram A Moskovitz <[EMAIL PROTECTED]> writes:
> Are you sure OCSP didn't come out of the IETF as an effort to
> 'standardize' the VeriSign / Microsoft certificate status service
> that was launched as an extension to ActiveX?

it may have ... but i know there were people starting to discuss some
sort of real-time rube goldberg contraption that attempted to preserve
the facade of an offline operation with throwing in minimum of online
operation.

i was getting hit with wouldn't it be a modern marvel to convert the
payment infrastructure to certificates ... and then pointing out to
them, that conversion to an offline certificate-based operation would
actually represent regressing the payment infrastructure by at least
20 years.

when we were doing this thing with this small client/server startup
that wanted to do payments
http://www.garlic.com/~lynn/aadsm5.htm#asrn2
http://www.garlic.com/~lynn/aadsm5.htm#asrn3

we had to do due dilligence on the major operations that were going to
be issuing these things called SSL domain name certificates in support
of the operation. we were constantly pointing out that most of them
were actually doing "certificate manufactoring" (a term we coined at
the time) and hadn't actually bothered to implement a real PKI that
actually administered and managed the infrastructure. furthermore, the
payment infrastrucutre had learned at least 25 years earlier that
revokation lists scaled extremely poorly.

the following is from the PKIX WG minutes apr 7-8 1997

Sharon Boeyen presented the work to date on Part 2 regarding the use of
LDAP and FTP for retrieval of certificates and CRLs and the requirements
for and specification of an Online Certificate Status Protocol (OCSP).

DISCUSSION

1 - Should we consider splitting the document into two separate ones,
since the OCSP is a new protocol definition which may require
significant more review and discussion than the LDAP and FTP profiles?

Resolution: Although we agree that OCSP may require additional review,
the document will remain a single draft and we will re-address this
issue, if the OCSP discussion is such that it will require a longer
review period and impede progression of the remainder of the document.

.....

I have the original email ... but it can also be found here
http://www.imc.org/ietf-pkix/old-archive-97/msg00316.html

some comment about the lead architect for ocsp was from valicert
http://www.rsasecurity.com/press_release.asp?doc_id=334&id=1034

in addition to ocsp ... about the same time there were some other
infrastructures looking at various gimicks to improve the revokation
process. note in the following announcement ... they were almost
quoting me word-for-word about how archaic the CRL process actually
is.

Date: Wed, 29 Oct 1997 21:05:12 -0800

SUBJECT: ON-LINE BANKING/ VALICERT TACKLES FLAW IN E-COMMERCE SECURITY

American Banker via Individual Inc. : A group of Silicon Valley
entrepreneurs has set out to correct a flaw in the digital
certification process that many Internet experts have been counting on
to make Internet commerce secure.

The solution, called a certificate revocation tree, is the property of
Valicert Inc., a Sunnyvale, Calif., company formed last year and
officially opened for business this week.

In a sign that Valicert may be on to something that could bring added
security to Internet transactions, three vendors in the data
encryption field have given endorsements, and Netscape Communications
Corp. has made a provision for Valicert's technology to "plug in" to
the SuiteSpot server software.

The advent of Valicert indicates that digital certification-a
cryptographic technique that is believed to be on the road to broad
public acceptance through Internet security protocols such as the
credit card industry's SET-needs further refinement.  "Today there is
no way to know if a certificate is valid at the time of a
transaction-it is known only that the certificate was valid at the
time of issuance," said Joseph "Yosi" Amram, president and chief
executive officer of Valicert.

He said that if not for the Valicert method of keeping revoked
certificates from being approved-it will be available in the form of a
tool kit for system developers, a server system, and a service from
Valicert-electronic commerce could collapse under the weight of
millions of digital certificates that cannot be adequately validated.
SET, the Secure Electronic Transactions protocol adopted by MasterCard
and Visa for on-line credit card transactions, illustrates the problem
in the extreme. SET requires issuance of digital certificates to all
parties to a transaction. They are the E-commerce equivalent of a
driver's license to verify a cardholder's identity or a certification
that an on-line merchant is what it claims to be.  The complexity of
processing transactions with those multiple certificates is widely
seen as slowing the adoption of SET.  But digital certificates have
already been issued by the millions through Netscape and Microsoft
Corp.'s Internet browsers. Verisign Inc.  and GTE Corp. are prominent
certificate vendors. GTE, Entegrity Solutions, and Entrust
Technologies, the leader in public key infrastructure systems, have
each agreed to some form of collaboration with Valicert.

Valicert's efforts can "expand the security infrastructure available
for commerce," said Tom Carty, vice president of marketing and
business development at GTE. "Given our focus on providing all of the
pieces of the infrastructure required to make Internet commerce
possible, it makes great sense for us to partner with Valicert to fill
in one of the most essential pieces of the infrastructure puzzle-the
digital credential checkpoint."

In a recent interview, Mr. Amram and Valicert chairman Chini Krishnan
said the problem is akin to what the credit card industry faced before
electronic authorization systems.

"A merchant would get a book, which came once a week or once a month,
full of bad credit card numbers, and credit cards presented at the
point of sale would have to be looked up manually," said Mr. Amram,
who joined Valicert in August after being involved in other high-tech
start-ups and in the Silicon Valley venture capital scene. "It was a
big hassle and it slowed down checkout."

The digital certificate equivalent of the hot-card list is known as
the certificate revocation list, or CRL.

Mr. Krishnan, the Valicert founder, said CRLs are "unscalable,"
meaning they become cumbersome, if not impossible, to manage as they
approach mass-market proportions. The lack of scalability "has posed a
barrier to widespread deployment," Mr. Krishnan said.  He claimed that
the invention of the certificate revocation tree brings a "1,000-to-1
advantage" that solves the problem of revocation and validation in a
tamper-proof and economical way.

"Developers need a cost-effective, one-step solution for building
applications that can check the validity of digital certificates," Mr.
Amram said. "By providing a clearing house network into multiple
certification authorities, and by delivering a robust technology
combined with a liberal licensing policy, Valicert will enable the
widespread development and use of applications that will make the
Internet and corporate intranets safe to conduct business."

"Certificates are the only way to deal with identity in any meaningful
way," Mr. Amram said. "They will take off in a big way. But
certificates without validation are like a car without brakes."

Mr. Krishnan said the development of Valicert's technology had "a lot
of rocket science elements," which is why it took the company 20
months to reach the launch stage. Enhancing its credentials, Paul
Kocher, a leading cryptography researcher, is credited with inventing
the underlying technology. Martin Hellman, a Stanford University
professor and half of the Diffie-Hellman team that invented public key
cryptography, is on Valicert's scientific advisory board.

Commercializers of cryptographic security have been intrigued by
Valicert's proposition. When he heard about it during American
Banker's Online '97 conference in Phoenix, Scott Dueweke, a marketing
manager in International Business Machines Corp.'s Internet division,
said, "They should call us."

Another expert, who asked not to be identified, said Valicert's
biggest problem is that it is a few years ahead of its time.  "The
market has fallen down with respect to revocation management, relying
on relatively short expiration dates" to minimize invalid
certificates, said Victor Wheatman, a California-based analyst with
Gartner Group, Stamford, Conn. "Valicert fills a void and hopes to
develop technology before the leading players move forward with their
own revocation capabilities."

Valicert's server and tool kit are available now, and its service to
certificate acceptors will enter field trials later this year, the
company said. The tool kit can be downloaded from the valicert.com Web
site free for noncommercial use and evaluation purposes. Application
development licenses are a flat $995 with unlimited sublicense rights.
The server can be deployed on corporate intranets for $9,995.



-- 
Anne & Lynn Wheeler | http://www.garlic.com/~lynn/
_______________________________________________
mozilla-crypto mailing list
[email protected]
http://mail.mozilla.org/listinfo/mozilla-crypto

Reply via email to