From: tcmay@netcom.com (Timothy C. May)
Date: Thu, 12 May 94 13:16:23 PDT
To: cypherpunks@toad.com
Subject: Archives, FAQ, and Why Things are the Way they Are
Message-ID: <199405122016.NAA25186@netcom.com>
MIME-Version: 1.0
Content-Type: text/plain


(This message has two parts. First, a discussion of archives, the FAQ,
etc. Second, why encryption and remailers have been such successes and
why things like digicash and other more exotic protocols have not.)


I received half a dozen requests for ftp posting of articles, for
archive sites, etc. This was in response to my comment about "golden
oldies."

We are in a tower of Babel these days. Dozens of computer platforms,
from mainframes to workstations to Macs to Windows to Amigas. Dozens
of mail programs (hence the fragmented support for something so basic
as PGP), dozens of newsreaders, dozens of etc.

The point: There is no easy solution to the problem of how newcomers
can come up to speed on this list. Some miscellaneous points:

* Yes, a FAQ would help. I am working on one. A long one, as is my
wont. I hope to have a version out soon for comments and further
submissions. I expect to either put it up for ftp at my site, or at
the soda/csuu(?) site, or to just mail it to folks who request to be
early commenters (I don't want the early versions ricocheting through
cyberspace).

* However, many people are demonstrably unwilling or unable to read
the sci.crypt FAQ (else why some of the questions we get?) or to
acquire Schneier. ("What's Schneier?" is one of the FUQs--Frequently
Unanswered Questions.)

* My best articles (in my opinion) and the best articles of others are
scattered in 200 folders/directories, arranged thematically. I use
Eudora's sorting capabilities to sort the mail into topics I think are
related (such as Crypto/Technical/One Time Pads). Then I index the
articles with "On Location," which allows me to pull up each article
that has key words, such as "uranium" anywhere in the message. My own
personal research tool. 

I mention this because I have no plans to collect these articles
(either mine or those of others) and put them up for ftp access.
Just too much work for too little gain.

* Newcomers should be patient. Read the list for a while. I notice
that one newcomer has announced that he is leaving the list after 3
days (!!) because his question on analog encryption went unanswered.
Oh well.

* The "Cypherpunks Archives" comes up from time to time. While we all
have our own personal archives (I have 40 MB of Eudora files devoted
more or less to Cypherpunks), many believe a browsable archive of
_all_ posts would be nice.

- this has some downsides: inspection by prosecutors and the like for
seditious, infringement-inducing posts, etc. (Yes, they could
subscribe and see much the same things, but making a browsable site
accessible to one and all seems risky, given the political climate.)
Also, the 50 or 100 MB (rough estimate) of Cypherpunk mail would be
unreadable except in dribbles and dabs. (The Bible is only 10 MB!)

- and recall that Cypherpunks has no budget, no organized structure,
and no means of accomplishing such goals as making archives available
except insofar as the volunteer efforts of folks like Hugh Daniel and
Eric Hughes go. And the toad.com machine is John Gilmore's personal
machine, for which we should be thankful we have any access to at all.

- a real archive, maintained by real people, would require time
commitments and budgetary commitments I don't see materializing
anytime soon. 


* Meanwhile, we face the "tower of Babel." Only text messages, like
this one, cut across all systems, all mailers, all readers, and can be
encrypted (into _other_ text blocks, which is why some
interoperability exists at all).

Many things that are "possible" in the malleable and Protean world of
computers simply never gets done. For while many things are
"possible," time and energy limits mean these things don't get done.

To the Unix jocks who send me their idea of helpful messages,
suggesting that I use MIME-compliant agents and HTML URLs for the FAQ
(or somesuch...), or that a few pages of perl would fix these
problems....thanks, but no thanks! (I get a fair number of messages
suggesting that my complaints about foo would vanish if I gave up the
Macintosh and adopted the One True Way, be it BSD, or XWindow, or
perl, or whatever.)

(I do have access to "lynx," a stunning character-based form of
Mosaic, and I've been cruising around webspace with this. But I intend
to distribute the FAQ as a simple text doc, not as a
WWW/Mosaic/HTML/URL/lynx thingamajig. I know this will be
disappointing to some fraction of you, but we can't all be pioneers
with arrows in our backs. I actually have some expectation that WWW
and Mosaic are the Next Big Thing, and that groups like our list will
eventually migrate to webspace, with Cypherpunks being a virtual
meeting place in webspace. But not in the next year or two.)

AND NOW FOR SOMETHING COMPLETELY DIFFERENT,...

One final word: My point about only the text block being the lingua
franca of e-mail and the Internet (with a few images and sound files
gaining strength) is an important clue to what works and what doesn't
work in the world of crypto:

* simple encryption works because everybody conceptually understands
the concept of the encrypted message, which remains a text block.

* remailers have worked for the same reason: everybody understands the
idea of readdressing a letter, and the underlying data structure for
the system remains a text block.

* other protocols, like digicash, reputation servers, anonymous
markets, and so on, have languished because of the tower of Babel
effect--too many layers of protocol communication, negotiation, and
cross-platform incompatibilities. And the "semantics" of these
protocols are far from clear. (Try playing the roles of Banker,
Customer, and Shop in a 3-entity digital cash protocol, with messages,
signatures, blinded signatures, and "money objects" flowing back and
forth. It gets confusing, even to those who've pored over the Chaum
papers. Now try to _automate_ the protocol to run with little human
intervention on a mix of platforms, e-mail systems, etc. This is the
"protocol problem" that I happen to think needs a _lot_ more work. A
big C or perl problem will not necessarily be the solution.)

So, our early successes (use of encryption and remailers) is not too
surprising, and that's why these successes came early on.

What's next is a much harder problem.

--Tim May

-- 
..........................................................................
Timothy C. May         | Crypto Anarchy: encryption, digital money,  
tcmay@netcom.com       | anonymous networks, digital pseudonyms, zero
408-688-5409           | knowledge, reputations, information markets, 
W.A.S.T.E.: Aptos, CA  | black markets, collapse of governments.
Higher Power: 2^859433 | Public Key: PGP and MailSafe available.
"National borders are just speed bumps on the information superhighway."