🚀 go-pugleaf

 In <1992Nov13.012545.29228@news.eng.convex.com>
 gardner@convex.com (Steve Gardner) writes:


>In article <1992Nov12.230445.25742@cactus.org> ritter@cactus.org
>(Terry Ritter) writes:
>> Computer cryptography makes it possible for those who accumulate
>> information to avoid the due-process search which is expected
>> when people accumulate things.
>    Good memory that is sufficiently selective during a trial
>    makes it possible for those who accumulate information to
>    avoid the due-process search which is expected when people
>    accumulate things.
>
>    Do you see what is wrong with this sentence?  Think about it.
>    Then . . . repeat after me: Information is not a thing.
>    Information is not a thing.

 Actually (wearing my hat as a computer engineer), I directly dispute
 the statement that "Information is not a thing."  I am unaware of
 any concept of "information" which is not held by some "thing."
 Paper is a thing.  Silicon is a thing.  Capacitors are things.
 Magnetic domains exist in magnetic material which is a thing.
 Molecules, atoms and electrons are things.  Brain cells are things.
 About the only way to store information in "no thing" is by
 electromagnetic waves, but if that information is ever to be used
 (Q: Is unused information "information"?), it must somehow *affect*
 a *thing*.  Information is *always* associated with things.

 To return to the point, society has agreed that a person should
 not be compelled to testify against his or herself for any reason.
 (Presumably, this was intended to eliminate some of the abuses of
 torture which occurred in Spanish, English and American societies
 of that period.)  But this fact provides no basis for arguing that
 society does not indeed lay claim to any and all *other*
 information, under legal warrant.  The Constitution testifies that
 society does have such a claim.

 This is the way our society has been.  Cryptography and computers
 change this in a fundamental way.  We might argue that this
 cornerstone of our society is not very important, or perhaps that
 it *should* be changed, but we cannot argue that a sea-change
 is not about to happen because of cryptography.


>> The difference is that what was once rare and easily breached by
>> the authorities is now on its way to becoming common and
>> impenetrable.  Cryptography really is different from a wall safe.
>    It will still be impenetrable for law enforcement in the
>    case of criminals.  Unless you make the "crime" of an illegal
>    encryption device (ie. a computer with th appropriate software)

 I believe most readers will recall that I have clearly stated,
 several times, ***FOR THE PURPOSE OF DISCUSSION***, two approaches:


    1. Require users to register their keys in advance; presumably,
       creating or transmitting enciphered information which could
       not be accessed by a registered key would itself be a crime.

       This could be surprisingly tricky to prove.

       This implies a key-registration bureaucracy, and also
       supports wiretapping for content without warning the
       target.

       It inconveniences 99% of the users, simply to allow access
       to the other 1%, who would not use registered keys anyway.

       Misused, it would also allow "fishing expeditions."


    2. As an alternative, I have proposed that each user be
       responsible for providing keys, after the fact, in response
       to a legal warrant.  Failure to do so would be a crime.

       This would be easy to prove, although there would always be
       some percentage of honest mistakes which juries would have
       to address.

       This would eliminate the key-registration bureaucracy, but
       *would* require users to archive their keys without fail.
       This would inconvenience all users to some extent, and
       would make the use of cryptography a serious issue.

       It would prevent wiretapping for content (but not traffic
       analysis), but would warn an entire conspiracy the instant
       a single individual was approached.

       "Fishing expeditions" would rapidly become apparent.


 Note that neither of these cases make computers, cryptographic
 software, or cryptography itself, illegal.


>    comparable to the most heavily punished felonies recognized by
>    the law criminals commiting those felonies will use it and they will
>    consider the risk quite acceptable.  The only people who will be denied
>    the right to some privacy from big brother will be decent
>    law abiding citizens.  I don't understand why you can't see
>    this.  Is it really that hard a concept?!?

 I see very well that criminals would not register keys, and would
 not provide access to enciphered information.  This would be a
 crime in itself, and if this were the only thing the authorities
 could prove, it would at least be something.  Probably, failure to
 disclose enciphered information would be a fact disclosed to the
 jury in the larger case, and so would have an effect there as well.

 The defendant would still have the right to not testify against
 (him/her)self without prejudice.  But failure to "open his/her
 books" *could* be used against him/her.  (I am an engineer and
 not a lawyer; I *assume* that testimony about "destruction of
 evidence" can be used in a larger case now.)

 I do note that the traditional protections accorded to notes taken
 by the press may (if enciphered) need to be excepted, or handled
 on a case-by-case basis.


 *IF* we find that society *does* have an unexpected ultimate
 "right" to access private information, we may want to find some
 least offensive way to provide such access, as opposed to being
 forced to accept whatever proposal the 3-letter agencies would
 prefer.

 ---
 Terry Ritter   ritter@cactus.org