[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: the principle of least astonishment



In article <[email protected]>,
Eli Brandt <[email protected]> wrote:
: > From: [email protected] (T. William Wells)
: > For my service, given what it is for, the presumption should be
: > anonymity. For the personals groups, perhaps the presumption
: > should be the other way around. There is, almost certainly, no
: > one right answer.
:
: What are your thoughts on solutions which do not do either of these
: alternatives?  For example, several people have discussed systems
: involving two sets of addresses.  These avoid both problems
: (unexpectedly failing to anonymize / unexpectedly anonymizing a
: message with a sig), at the cost of some complexity.

I haven't really given this much thought because it hasn't been
relevant to my server. There are a number of problems that I
think fall into the same category of inadvertent disclosure. All
of them are "operator error" in a sense. All of the anonymous
e-mail services are hacks added onto the existing e-mail services
and require significant attention to detail if one is to not
inadvertently give away one's real identity. It's as if you had
to type in the RFC822 headers yourself for each message; even
though they're quite simple, you're going to mess up reasonably
frequently. And just once is sufficient to destroy one's
anonymity. The multiple address thing doesn't address this at all
so I don't think it will help.

Alas, I really don't have the spare time to work up my thoughts
on where the e-mail system ought to go. All I can really say for
sure is that the whole thing should be rethought from the ground
up.

: > Either the perpetrators are so intellectually lacking that they
: > do not see what they are doing or they are so intellectually
: > dishonest that they do. In either case, I am utterly disgusted.
:
: Aren't you detweiling a bit here?

I don't think so. Detweiler and those like him simply react; they
do not examine where they are coming from nor are they willing to
do so. Their fault is not that they reach strong moral
conclusions nor that they express them but that they so lack
respect for others that they will not examine the positions of
others nor attempt to determine where the line of "it isn't my
business to tell them what to do" should be drawn.

(BTW, cypherpunks wins big when it loses Detweilers; just look at
Objectivism's reputation for why.)

:                                    I don't think the situation really
: warrants "utter disgust"...

Technical problems can often be solved by "try it and see" but
systemic people problems rarely can or should be. A wrong
solution is often just too costly; ask the folks in the ex-USSR
what they think of that "try it and see". People problems require
careful thought if the results of one's actions are not to lead
to misery and death.

When people resort to rather obviously flawed modes of reasoning,
disgust is quite appropriate, except when stronger responses are
warrented.