The Market for Crypto--A Curmudgeon's View

Eric Hughes eric at remailer.net
Tue Nov 29 22:36:06 PST 1994


Let me be REAL clear about this.  The immediate proposal is to mark
and possibly delay unsigned messages to the list.

The proposal does NOT include bouncing messages or preventing use.
These options are acknowledged as possibilities for the future.  They
are not on the table right now.  I, unlike the gov't, will warn you of
your impending doom.

   From: tcmay at netcom.com (Timothy C. May)

   Not to trivialize this proposal by frivolously insulting it, but
   consider a mailing list that decided to delay/bounce any messages that
   were not written in TeX, or in Acrobat, or whatever. 

I don't think you are frivolously insulting it, but I do think you are
ignoring the basic distinction I made about the difference between
measures which prevent use and measures which do not.  The use of the
syntax "delay/bounce" denies exactly this distinction.

   [...] to delay/bounce any messages that
   were not written in TeX, or in Acrobat, or whatever. How would people
   react who lacked these capabilities, or preferred to use alternatives
   (like simple unadorned text), or who merely object to an enforced
   standard?

I have two answers, one for delay, the other for bounce.

1.  For delay or other non-preclusive measures, those who do not use
the valorized feature can still use the list.  They get signalled in
some fashion that use of the valorized feature is desired.  I consider
this primarily a communication mechanism.

I wish to communicate to everyone one the list that using digital
signatures is something that I want everyone to do.  In particular,
that means that you, the current reader of this message, are one of
the people I want to use digital signatures.  Rhetoric is not as
effective as a policy embedded in software that people interact with.
Doing is more effective than hearing.

2.  For bouncing or other preclusive measures, those who do not use
the valorized feature can't participate in the discussion.  This would
in many situations be counterproductive, but in others, say, an
experimental group discussing design in Acrobat, absolutely vital.
As this is not germane to the actual proposal, I leave off here.

   But that Eric [...] has some notions of
   what people _ought_ to be using does not seem to be enough to
   effectively bar those who helped form the Cypherpunks group (many of
   us) just because they choose to communicate in one particular way.

I want you, Tim May, to use digital signatures.  There, that's
explicit and verbal.  I do understand if your software doesn't
cooperate.  I've been there.  I'm not (to repeat) talking about a
proposal to eliminate you from the list.

Does a mark or a delay constitute an "effective bar" from
participation on this list?  I think not, although I'm entertaining
arguments.

   If some flavor of PGP is mandated, I expect I'll unsubscribe (as I
   can't stand reading but not posting...lurkers obvious feel otherwise).

Whoa!  We went from an effective bar to an actual prevention there.
That's not what I'm talking about.

And I'm not tied to PGP by any means.  You want to make a digital
signature with some other piece of software?  Fine.  I'll add it right
in.

   Absent a compelling reason, a market reason, why bother with someone's
   notion of ideological reasons?  

I'm not a libertarian (neither big L nor small l), and I don't find an
identity between compelling reasons and market reasons, as apposition
implies.

The implementation of function at the server is a communication
between me, Eric Hughes, the implementor of that nasty shit, and you,
the participant in the cypherpunks list, that I want you to use
digital signatures.  Now, because of my position as de facto list
maintainer, I can do this and you can't.  I've got the bully pulpit,
and while I've not used it much, I am beginning want to spend some it
on urging crypto deployment and usage.

Not all is lost for erstwhile communicators.  One could write a filter
to look for unsigned posts and pipe them off through a suitably hacked
'vacation' filter which would send them a missive (but not too often)
encouraging the use of cryptography and which would include pointers
to software.  This kind of communication is similar in form but not in
scope to what I've proposed for the list.  In fact, if someone were to
bundle this kit up, I suspect it might receive fairly wide use.

   [...] perhaps I'd insist that all posts be paid
   for in digital cash...or bought, or whatever. 

You hypothetical includes an insistence.  Mine does not.

   Again, I thought the proposal was to ultimately reject non-signed
   articles?

There's a very explicit disclaimer to contrary in the original.  To
paraphrase, it acknowledged the possibility of rejection but removed
it from immediate consideration.

   Speaking of this, it's already pretty clear who signs and who doesn't.
   What could be clearer than "----BEGIN PGP SIGNED MESSAGE---"?

What about random headers with things like:

X-Signature: none
X-Warning: Cryptography Non-User
X-Heckle: Yo! Too _good_ to use crypto?
X-Lazy: Jeez, Eric's even got a Unix box at home and _still_
        isn't signing?
X-Bozo: God, Tim's been on this list for over two years and
        he still doesn't sign his posts?
X-Traitor-To-The-Cause: <insert From: field contents here>
X-Cryptography-Impaired-And-Proud:

[For the satire impaired, note the use of the phrase "satire impaired"
at the beginning of this sentence.]

   If the proposal is to stamp a scarlet
   letter on non-signers, it seems overly harsh, somewhat petty, kind of
   insulting, and not needed.

A scarlet letter is a reasonable apt analogy, except the intent is not
to create outcasts.  Harsh?  I still fail to see that.  Petty?  What
trivial mattr is being blown out of proportion?  Insulting?  I'm sure
some people can take it that way.

Not needed?  Perhaps not, but I may _want_ it.

Eric






More information about the cypherpunks-legacy mailing list