David Shaw wrote:
On Fri, Jan 21, 2005 at 03:00:01PM +0100, Adrian von Bidder wrote:
Perhaps 'rush in' an ID (or a couple of IDs, as per Vlad Miller's mail) with
reference to it being the sole legally useful algorithm for some
applications in russia, with all details to be ironed out in another
document?
I'm taking a wait and see attitude with regards to GOST itself, but I
am against sticking in an ID without details. Whether in 2440bis or
in another document, we should either do GOST or not do GOST. The
halfway thing doesn't really help anyone.
A few months back we actually removed some 'IDs without details' from
the draft.
So the test might be that even though 2440bis does
not describe the paramaters/details, it should have
a reference to where they *are* definitively stated,
to the extent that any implementation can create an
OpenPGP compatible plugin with that algorithm.
I must admit, if these details are not laid down
somewhere, I'm unsure what the point of adding
the Id into 2440bis is. If there is a risk that this
could end up pointing to separate implementations,
then that would indicate that we had put the cart
before the horse?
iang
--
News and views on what matters in finance+crypto:
http://financialcryptography.com/