From: SL Baur <steve@xemacs.org>
Date: Fri, 12 Sep 1997 16:08:17 +0800
To: Tim May <tcmay@got.net>
Subject: Re: Any talk of limiting _existing_ crypto?
In-Reply-To: <v03102809b03e747663d3@[207.167.93.63]>
Message-ID: <m2vi07t0vf.fsf@altair.xemacs.org>
MIME-Version: 1.0
Content-Type: text/plain



Tim May <tcmay@got.net> writes:

> If existing crypto is fully legal to use, then it could be years and years
> before the Freeh-Reno-SAFE outlawing has any significant effect.

Indeed.  How about existing software with decades of previous
existence that has hooks where crypto can be plugged in with fairly
trivial effort?

> Is there any reasonable interpretation of any of the SAFE or Pro-CODE bills
> that could make it illegal to use preexisting crypto programs ...

> If not, then our strategy should be to get the simpler, text-centric,
> crypto programs massively and widely deployed. Spend the year or so we have
> before D-Day getting crypto onto every CD-ROM being distributed, every
> public domain site, etc.

We (the XEmacs developers, but I write only for myself as XEmacs
maintainer) already distribute mailcrypt (Emacs lisp code to
integrate PGP with mailers and newsreaders implemented in Emacs
lisp).  It has been possible to integrate encrypted editing with
encryption of the user's choice transparently for some time.

> Better, perhaps, to leave the crypto at the "text edit" level, the
> ASCII level, where it can be dropped in cleanly to whatever program
> is current. (Also an old strategy, one with many advantages.)

I agree.  An Emacs-like architecture makes implementing something like 
this pretty straight-forward.