NODE 4288ab1cRe: State Dept Response to my second CJ request
Hal <hfinney@shell.portal.com>Wed, 11 May 94 23:01:31 PDT
There is a problem with these "hair splitting" approaches to avoiding the
ITARs (they accept the book; they reject the disks, so we ask to send some-
thing that is halfway between the book and the disks, etc.). There is a
well-known fallacy (whose fancy name I don't remember) which says that even
though night and day change gradually from one to the other, and you can't
really draw a line separating night from day, that doesn't change the fact
that night is different from day.
We may establish that hitting someone with a baseball bat is against the
law, and hitting them with a feather is not; then we proceed to ask whether
hitting them with a pillow is against the law, and so on. At some point
the law is forced to make an absurd decision that hitting someone with item
X is illegal while hitting them with Y is not, but X is almost the same as
Y.
Does this prove that no amount of assault is illegal? No. It just means
that lines are not always easy to draw.
In the same way, it is not easy to draw a line between a book which is
protected by the first amendment and a program which a person can sit
down and run to get military grade cryptography. But that does not
lead to a strong legal argument that all cryptographic software is export-
able, IMO.
Hal
NODE 9a75323dRe: State Dept Response to my second CJ request
m5@vail.tivoli.com (Mike McNally)Thu, 12 May 94 05:51:40 PDT
Hal writes:
> In the same way, it is not easy to draw a line between a book which is
> protected by the first amendment and a program which a person can sit
> down and run to get military grade cryptography. But that does not
> lead to a strong legal argument that all cryptographic software is export-
> able, IMO.
Though I agree that the feather/pillow/stick/club scenario is
unrealistic, I disagree that it applies in this case. The ITAR
regulations are being enforced around a situation that's essentially a
technological accident.
The difference between an exportable piece of software printed with
ink on a page and one in human-readable ASCII on a diskette is defined
solely by the state of technology. If, tomorrow, some company began
selling a $99.95 scanner with built-in OCR translation software, then
there would really be no difference whatsoever. To return to the
original analogy, it would put the "court" in a position of having to
declare an assault with an oaken bat illegal, but one with a hickory
bat OK.
There is a similar lack of distinction between source code and machine
code. If I introduce a computer system whose primary interface
includes a C interpreter, then in some ways the source code *is*
machine code.
--
| GOOD TIME FOR MOVIE - GOING ||| Mike McNally <m5@tivoli.com> |
| TAKE TWA TO CAIRO. ||| Tivoli Systems, Austin, TX: |
| (actual fortune cookie) ||| "Like A Little Bit of Semi-Heaven" |