🚀 go-pugleaf

 In quotable email, <mike@upolu.gsfc.nasa.gov> Mike Jones responds:


>This has little to do with sci.crypt

 I disagree.  Should society elect to limit the use of cryptography
 by law this will affect cryptographic design.  Systems may have
 to archive old message keys, for example, or even ciphertext.


>You and I and everyone else knows how to get that
>information. Put the alleged child molester on the rack and use
>heated irons
>until you have the information.

 Nonsense.  Two possibilities are presented:

 1) Require the use of registered crypto keys; if other keys are
    actually used (as we expect), this would itself be a crime;

 2) Require users of cryptography to reveal their keys (after due
    process); if they did not (or *could* not), this would itself
    be a crime.

 Nobody gets heated irons; this is rule of law.  The issue is
 whether a normal legislator could be convinced that this would be
 *bad* law.


>On a technical level, you know that a registered key will not
>guarantee the results can be decrypted. It may be XORed with an image of
>daisies.

 The question of whether a defendant has provided the desired
 plaintext can be addressed under oath by experts, and decided by
 juries.  Deciding issues of fact is what juries do.

 ---
 Terry Ritter   ritter@cactus.org