Stephen Davidson wrote:
[1] I think it's fair to say that the origins of the
CA market were a case study in a pure anti-competitive
market.  Legislation was proposed and pushed through
by CAs in some places that created a barrier to entry.



I don't buy this. Legislation was proposed to allow digital signatures to have legal equivalence to a witnessed pen and ink signature.


Originally, legislation was proposed (and passed)
to have far more ramifications than just that.
Certainly, "legal equivalence to a signature" was
the headline intent, but it went way beyond that.

(We are talking circa 1994-95 here, so this is all
ancient history.  It's only place is in understanding
the origins of protocol design around that time, as
these forces were present and pervasive.)

Now, luckily, the whole concept was questioned, and
the original "model legislation" fell out of favour.
By 1997, most legislative researchers were aware of
the debate, and legislatures passed something closer
to useful, which is that "a digital signature should
not be denied purely on the grounds that it is a
digital signature."  Those laws also often included
a gap-filler clause that stated that the government
had the authority to license CAs if it thought fit,
but left all rules up to later.  (I recall reading
the California proposal, it was a page long.)  This
was quite wise, as it blocked any attempt to force
through the franchise model at a later time.

(I do not recall what Bermuda's legislation said
precisely, but I do recall that it had avoided the
trap.)


> To do so, the
CA needs to prove that the user has been properly authenticated, and that
the CA operations are secure (ie fake IDs can't be generated).


Question is, does it work?  That we can (and do)
debate.  A more precise question is whether this
can shown in court?

( It's now fairly well established in the crypto
community that digital signatures are not
acceptable, prima facie, as equivalent to the
pen&ink human signature, and laws that try and
enforce that won't survive in court.  IOW,
non-repudiability is now a repudiated concept. )


Is this a barrier to entry?  Don't think so -- it makes the technology more
useful.


The "deliberate" barrier to entry was the original
model, not the later generations, which squashed
the automatic creation of the franchise.

When I was researching this question over the
weekend, I found one State's estimate that the
cost to the CA of the original model as $3m, which
rates as a barrier to entry, in my book.


iang _______________________________________________ mozilla-crypto mailing list [EMAIL PROTECTED] http://mail.mozilla.org/listinfo/mozilla-crypto

Reply via email to