From: Greg Broiles <gbroiles@netbox.com>
Date: Tue, 7 Jan 1997 16:41:17 -0800 (PST)
To: Pierre Uszynski <pierre@rahul.net>
Subject: Re: Sandy and I will run a cypherpunks "moderation" experiment in Jan
Message-ID: <3.0.32.19970107164034.006b6044@mail.io.com>
MIME-Version: 1.0
Content-Type: text/plain


At 09:37 AM 1/7/97 -0800, Pierre Uszynski wrote:
>Rich Graves <rcgraves@disposable.com> very correctly mentions:
>
>> 1) Moderator liability and anonymous posting.
>
>I agree that this is actually a critical problem with a filtering
>moderation scheme. Such a scheme appears to provide the capability to
>filter out possible "copyright violations" posts. From what I remember
>of the Netcom/CoS case (without going back to the sources), that may
>mean more liability for the reviewers (and the operator of the
>machine). That's a major point against simple filtering moderation.

I agree that this raises the spectre of liability for messages passed on,
but I'm not sure it's a big problem. I see three broad categories of
information which the moderation liability scheme may suppress:

1.	Copyrighted items such as newspaper/magazine articles
2.	Secrets which are being revealed (e.g., the alleged RC4 source, the
Mykotronix trash stuff) 
3.	Defamation

As to the loss of (1), I'm not heartbroken - much of this information is
already placed online by its owners, and can be referred to with a
hypertext link or a reference instead of being posted in its entirety.
Also, some people (who I don't feel like singling out) currently provide
access to third-party copyrighted information, but in a discreet manner.
The "copyright violations" poster(s) could provide access in this way,
unless they're committed to making John Gilmore and the remailer operators
take the heat for someone else's actions. 

The loss of (2) is regrettable; but Usenet (and other unmoderated forums)
are still available for hit-and-run disclosure of secrets. Also, it's
unclear to what extent (2) will be lost. Given the recent California and
Federal statutory changes strengthening trade secret protection, I think
it's useful to be careful - but it's arguable that neither the alleged RC4
code nor the Mykotronix trash stuff would have been a trade secret
violation. Also, let's weigh the value of what'd be lost (how many of these
messages do we really get?) against the negative value of the crap we're
currently subjected to, and the (speculative) value of posts we don't get
because their authors have left the list because of disgust and annoyance.

I'm not going to lose much sleep over (3). I can't call to mind a single
instance of interesting or useful defamation that I've seen on the Net.

>ichudov@algebra.com (Igor Chudov @ home) responded:
>
>> [in another forum]
>> After long thinking, moderator board has come with the following 
>> solution:
>>
>> 1) We do not know for sure if a certain post violates
>>    some copyrights or not
>> [and more in the same line 2,3,4,5]
>
>Would any of this have mattered in Netcom/CoS?

My hunch is that liability for infringement will focus on defendants'
actions after they knew or had reason to know that a particular act of
copying or distribution was a violation of the copyright owners' rights.
Copying data from one place to another is the essence of the net; holding
owners of machines (or moderators) responsible for violations they couldn't
have stopped is nonsensical. 

I predict (but cannot cite cases so holding) that moderators/listowners
will be held liable for copyright violations they could have detected and
stopped with the exercise of some level of care; but that they will not be
held liable where no reasonable amount of effort on their part could have
prevented the copyright violation. And I think "reasonable" will depend on
the facts at hand. (All of the contributory infringement cases I'm familiar
with have involved a defendant who was aware of, if not actively assisting
with, the underlying infringement.)

>Instead, a system that would forward reviewers' opinions *after the
>fact* does not have any of this problem. And we have already mentioned,
>it is also more powerful (real time initial feed, easy multiple
>feedback feeds, fully compatible with anything else...) although it
>does not reduce bandwidth requirements.

Can you name a software package which runs under Windows or the Mac OS
which automatically processes reviewers' opinions against a mailbox of
incoming mail? Better yet, can you name a Eudora plug-in which does so?
Abandoning an imperfect but workable solution because it's possible to
imagine that someone will create something better someday strikes me as
silly. Comparing theoretical software/strategies with currently implemented
software/strategies is comparing apples and oranges. 


--
Greg Broiles                | US crypto export control policy in a nutshell:
gbroiles@netbox.com         | 
http://www.io.com/~gbroiles | Export jobs, not crypto.