Re: Anonymous Remailers at work

[ichudov@algebra.com (Igor Chudov @ home)]

Peter M Allan wrote:
> > From: Bill Stewart <stewarts@ix.netcom.com>
> > I've recently run into a couple of business problems at work
> > that could be solved by (slightly modified) remailers.
> 
> > 1) Manager performance review, suggestion boxes, and questions to
> > visiting honchos - there are several departments that are using
> > "email to the secretary who'll take your name off and forward it"
> > to handle this problem.  Remailers are an obvious solution.

It is a little funny solution. What prevents me from writing five positive
performance reviews about myself, anonymizing them and sending to my boss?

What prevents someone from writing a very negative performance review
about someone else and forwarding it to their bosses (has been done
numerously:)?

A system analogous to anonymized voting may be useful in this case.
Simplified for a real-life office, it may be the following. Suppose we
have N workers. The secretary's program generates N random numbers and
publishes ONLY their SHA hash values. It also prints the numbers
themselves on separate pieces of paper. Secretary puts these pieces into
a hat so that the numbers are not visible.

Workers take one number each. Since they witness the procedure of taking
numbers they know that they have their anonymity.

Then they send the reviews of their peers to the management or publish
them in internal newsgroups, of course anonymously, attaching the 
numbers given to them for verification.

Since the list of checksums is (or may be) publicly known, there is
little way to cheat by double voting, and there is little way to find out
who wrote what.

Of course in their peer reviews workers should beware of using
cliches such as "Mr. X is a lying homosexual Sovok forger". :)

	- Igor.