From: "E. Allen Smith" <EALLENSMITH@ocelot.Rutgers.EDU>
Date: Mon, 3 Feb 1997 22:18:12 -0800 (PST)
To: roach_s@alph.swosu.edu
Subject: Re: Dissolving Choke Points
Message-ID: <01IF00D6DQPC9AN1BM@mbcl.rutgers.edu>
MIME-Version: 1.0
Content-Type: text/plain


From:	IN%"roach_s@alph.swosu.edu"  "Sean Roach"  4-FEB-1997 00:47:40.30

>On your idea for fast implementation,  this could be hacked by would-be
>censors.  A message comes into the list, the censor sees that the post is
>from a regular enemy.  The censor then copies the header information onto a
>new message, one containing garbage, or snippets from old posts, reads the
>new post to see if it is acceptable, and if it is, adds a new header,
>probably only the time would be changed, and transmits it out.  If the would
>be censor was the sysop of a machine near the origin of the message, and if
>the censor operated the censorship either by bots, or monitored the feed 24
>hours a day, (not likely), then the actual message would get to few.

	I would suggest that the cure for this problem is to have what is
examined be a cryptographically secure hash instead of (or as well as)
the message ID. The hash should be over the body of the message, and possibly
the sender and Subject line.
	Unless Lance has driven up the prices at Infonex a lot, I'd be willing
to support such a server on an Infonex account. I'd _greatly_ appreciate help
getting the thing started, preferably as a group endeavour (to spread the
load (i.e., keep infonex from being overwhelmed), keep me from temptation,
and make it less likely that a legally troublesome message would go through
an account that I might be blamed for).
	-Allen