From: stewarts@ix.netcom.com
Date: Sat, 20 Dec 1997 04:50:05 +0800
To: Robert Hettinga <schear@lvdi.net>
Subject: Re: Lions and Tigers and Micromoney, Oh, My...
In-Reply-To: <199712182121.QAA04394@users.invweb.net>
Message-ID: <3.0.3.32.19971219123619.0070897c@popd.ix.netcom.com>
MIME-Version: 1.0
Content-Type: text/plain



At 4:21 pm -0500 on 12/18/97, William H. Geiger III wrote, 
though not in this order,

>> I have *PAID IN FULL* for my Inet usage!! What bits I send over the Inet,
>> how many bits I send, and who I send them to is NO ONE's BUSINESS but my
>> own!!! -- Last Anarchist of the Inet.

You've paid for your end of the connection, but you haven't paid for mine,
and you especially haven't paid for getting your bits to my eyeballs.
Neither has Spamford.  I'm usually willing to read your words, but not his.
Hashcash-mail is a tool for users to use to automate limiting the email 
they receive to mail sent out by real individuals who actually want to 
communicate with them, throwing away mail from spammers.

Whether it will catch on or not is a social issue.  There are telephone
answering devices that prompt for a password and either hang up on callers
who don't have one, forward them to the answering machine, or just
give them a low-priority ring.  I don't use one, and nobody I know
currently does, because they annoy the callers you care about
more than they annoy telephone spammers, who have lists of thousands of
other people to spam if you're not interested.  On the other ham,
if PacBell offered distinctive ringing like most Bell companies do,
I'd use it - not for splitting types of callers, but for my fax machine...

>> Note: In Adams proposal for hashcash only charges the user CPU cycles. The
>> incentive for wide implementation of hashcash is going to be a real ecash
>> based system where the implementors can make $$$ off it.

It would be interesting to find useful problems that could be used 
instead of hash collisions.  Factoring numbers can be distributed well,
and it's possible to tell quickly that X is a factor of Y, but unfortunately
it's not possible to tell quickly that the user unsuccessfully checked some
range of potential factors rather than just pretending to have done so.
Similarly, if the user finds the key K | C=DES(M,K), you can tell quickly,
but you can't tell if the user checked a range K1..Kn except by
checking it yourself (slowly.)  But maybe there's some useful variant that
can be checked quickly.

>> Well IMNSHO hashcash mail sucks!! It opens up the pandora's box of usage
>> based charges for everything done on the 'net. What will be next? FTP
>> sites charging hashcash for DL's? WebPages charging hashcash per hit? DNS
>> servers charging per lookup? Routers charging per packet?

Digicash, maybe, though not hashcash, since hashcash is just a timewaster
that's easy to verify and only takes a long time for high-volume users.
As far as routers charging per packet goes, we've had X.25 networks which
charge by the kilopacket for a long time; it's much clumsier and more
expensive
to operate than a flat-rate network.  One of the reasons the Internet has 
overtaken X.25 so dramatically is that flat-rate pricing encourages people to
use and provide free services, increasing the base of customers and the
size of the flat-rate service they buy -- the ISPs also win.

And many web pages _do_ charge for reading them - usually the charge is
the time it takes to download and (read or) discard an advertising banner;
users are generally much more willing to pay that cost than cash,
though there are some services that give you abstracts of articles for free
and charge you money for the full version (either charging per item or 
charging per month for "premium" service.)

DNS charges are paid for by the domain name holder who wants to be found.
TLDs and SLDs are funded by the domain name registration fees in some places,
and unfortunately by Your Tax Dollars in other places, or by DNS fees
laundered by the government through their subcontractors.
Lower-level domain name service is paid for by the SLD holder or agents.
If you don't want to pay to publish your DNS name on the servers,
you can always call yourself http://192.257.23.42/ or whatever,
as many spam sellers do.


				Thanks! 
					Bill
Bill Stewart, bill.stewart@pobox.com
PGP Fingerprint D454 E202 CBC8 40BF  3C85 B884 0ABE 4639