From: tmp@netcom.com
Date: Mon, 4 Apr 94 08:12:58 PDT
To: cypherpunks@toad.com
Subject: wrong place at wrong time...
Message-ID: <199404041513.IAA11198@netcom9.netcom.com>
MIME-Version: 1.0
Content-Type: text/plain


hello, whoever this `detweiler' person is you really seem to hate him.
i appear to have been in the wrong place at the wrong time. would you
like me to change my login name to something different so you don't
confuse me with him? i think i heard it only costs $10 or so. yes,
i do post from denver but i thought i heard detweiler was posting
from somewhere else in colorado. (of course, geography is pretty 
meaningless in cyberspace.) yes, i am something of a newbie to this
list. i heard about it out in talk.politics.crypto or something like
that. i have also been reading up on your remailers from the public
domain stuff out there (soda.berkeley.edu, etc.)

the issues of `detweilering' do bring up some interesting questions. 
it seems to me that merely charging for access to remailers is not
a sufficient deterrent (although it would have some effect). consider
this argument: top-notch explosives can be very expensive, but that
doesn't prevent terrorists from using them. in fact, it just has
the effect of deterring the use of them by `casual' terrorists.

wouldn't it be simpler to put limits on message propagation through 
a remailer? for example, limiting the volume of mail from a given
source address?

other interesting issues raised by `detweilering': hal finney forbids
detweiler from using his remailer at any time. but (hello? is anyone
home?) remailers operate on precisely the principle that names are
irrelevant. how can you forbid anyone from doing something when you 
have no way of *identifying* them? how do you know what their mail
aliases are? i wouldn't be surprised if hal finney censors *me*
from his mailers just because he *suspects* that i am detweiler (ouch!).

i think everyone here needs to be reminded why names were *invented*--
so that people can attribute actions to various individuals. it is
`social ettiquete'. now, in some cases you may say that there is no
need to attribute a name to an action (such as a posting) but then
it seems you have no right to complain about `detweilers'.

consider this-- suppose that i really hate t.c.may (this is just an
example). i would like to screen my mailbox of everything he writes.
but if he has dozens of imaginary aliases on netcom.com, each individually
pretending to be a real person, how can i do this? it would be impossible.
and don't say that the cost of multiple addresses would be a deterrent.
what if he is a bored millionaire with nothing better to do? what
if detweiler is a bored millionaire as well? apparently (from what i
can determine) he seems to be endlessly entertained by harassing 
cypherpunks. and the criteria `that's an annoying post' is obviously
not a very effective way of identifying him.

so, `cypherpunks', you might consider the following problem. is it possible
to have a system where there are multiple aliases allowed to a given person,
and comparison between aliases (does alias1 == alias2) is permitted, but
anonymity is generally preserved? one possibility is to remove screening
one level from the users. that is, i tell a central server `stop sending
me all posts generated by [alias1]'. if certain people suddenly disappear
from my view that would be highly incriminating but not conclusive that
they are all the same person.

p.s. if detweiler is using the initials `tmp' (which he is judging by
the FAQ piece that was reposted here) does anyone know what they
stand for?

p.p.s. can anyone tell me of recent press or media accounts of cypherpunk
activity? thanks.

p.p.p.s. try not to be so paranoid about `detweilers'. seems to me  that
if all he does is bang on a keyboard he is probably quite harmless. but
then again, that's all that mitnick and morris did <g>