[Cryptography] Opening Discussion: Speculation on "BULLRUN"

Anne & Lynn Wheeler lynn at garlic.com
Sat Sep 7 15:52:52 EDT 2013

On 09/07/13 05:19, ianG wrote:
> If so, then the domain owner can deliver a public key with authenticity using the DNS.
>This strikes a deathblow to the CA industry.  This threat is enough for CAs to spend a significant amount
> of money slowing down its development [0].

unfortunately as far as SSL domain name certificate ... the domain name infrastructure is the
authoritative agency for domain name ownership ... the SSL domain name certification agencies
have to rely on the domain name infrastructure to validate true ownership for SSL domain name
applications. As I've repeatedly referenced ... this puts the CAs in catch22 ... they
need improved integrity of domain name infrastructure (attacks on ownership records of domain
name ownership and then being issued valid SSL certificate) ... which comes with lots of
DNSSEC ... but that also eliminates much of the need for SSL domain certificates.

as per prior reference about original working on SSL for electronic commerce ... at least for
the financial industry I've repeatedly shown that digital certificates were redundant
and superfluous. I also shown that at the time, the addition of digital certificates
increased the payload size by two orders of magnitude (besides being redundant and superfluous).
That apparently motivated the "compressed" digital certificate financial standard effort ...
trying to reduce digital certificates so that the payload bloat was only ten times (instead
of hundred times) ... in large part by eliminating all information that the processing
institution already had. I demonstrated that processing institution would have all
information and therefor digital certificates could be reduced to zero bytes ... so
instead of eliminating redundant and superfluous digital certificates ... it was possible
to mandate that zero byte certificates be appended to every transaction (it would be
possible to digitally "sign" a payment transaction for authentication ... and rely on
the individual's financial institution to have registered the person's public key ... w/o
having to increase the size of every payment transaction in the world by 100 times just
to transmit a redundant and superfluous appended digital certificate).

I like the interchange at panel discussion in early 90s ACM SIGMOD ballroom open session,
somebody in the audience asked what was all this x.5xx stuff about and one of the panelists
said it was a bunch of networking engineers trying to reinvent 1960s database technology.

there was some amount of participation by the information assurance directorate in financial
industry standards meetings. at various times there were references to rifts between IA
and SIGINT ... but for all I know that may be kabuki theater. I was fairly vocal about
any backdoors could put financial industry at risk for bad guys discovering the vulnerabilities
... and wanted KISS applied to as much as possible (and backdoors forbidden)

there are other agendas in much of this. at the start of the century there
were several "safe" internet payment products pitched to major merchants (accounting for 70%
of internet transactions) which got high acceptance. Merchants have been indoctrinated for
decades that a large part of interchange fee is proportional to associated fraud rate ...
and the merchants were expecting an order of magnitude reduction in their fees (with
the safe products). Then came the cognitive dissonance when the banks told the merchants that
rather than major reduction in interchange fees with the "safe" payment products ... there would
effectively be a surcharge added to the highest fee that they were already paying (and all the
safe efforts collapse).

Part of the issue was that the bottom line for large issuing banks was 40%-60% from these
fees and an order of magnitude reduction in those fees would be a big hit to
their bottom line (the size of fees in part justified by fraud rates). The "safe" products
going a long way to eliminating most fraud and commoditizing the payment transaction
business ... which would also lower the bar for entry by competition.

virtualization experience starting Jan1968, online at home since Mar1970

More information about the cryptography mailing list