Kent and all,
Compleatly agreed here Kent. And this post of yours is not
excpetiong to the contribution of NOISE! >;)
Kent Crispin wrote:
> On Mon, Mar 15, 1999 at 03:33:09PM -0800, Dave Crocker wrote:
> > At 01:02 PM 3/15/99 +0100, Roberto Gaetano wrote:
> > >In other words, they succeeded in their goal.
> > >Which would not have happened if everybody could keep Fred Baker's attitude
> > >of ignoring them.
> >
> > I happen to agree with you, but the reality is that for a number of
> > reasons, these people are not ignored and their comments DO cause
> > problems. I'm a fan of having each participant do the filtering, but it
> > turns out not to be practical in this particular instantiation of reality,
> > so we need a procedure for dealing with it.
>
> As a practical matter, filtering cannot be made perfect, so some
> leakage will always occur -- someone could always change ISPs or
> change their name, and send a message. And it should also be noted
> that even the (hypothetically?) angelic Fred Baker does not ignore
> everything.
>
> I have often thought that it might be productive to think of this in
> more abstract terms, and try to approach the problem from an analytic
> point of view. At present, the discussions tend to degrade into
> political and social noise. So, maybe, if we deliberately try to
> keep things on a more abstract we can avoid this, and actually make
> some progress. Political and social issues are probably unavoidable
> in the long run, but perhaps we can discuss mechanism first.
>
> We could start by trying to describe the problem in terms of
> increasing the signal to noise ratio in a particular communications
> medium (mail lists).
>
> This medium provides gain (in a metaphorical sense) -- the reason
> that we participate in mailing lists at all is so that people will
> respond to our posts. So, to continue the metaphor, input of signal
> causes more signal to be generated. This (metaphorical) gain is true
> for "noise", as well -- input of noise causes more noise to be
> generated.
>
> A fundamental problem, then, is to provide some mechanism that
> discriminates between noise and signal. Such a mechanism would be a
> "noise discriminator" for a list. In concrete terms, it might be a
> written policy with human interpreters. Or it might be some
> complicated pattern matching mechanism supported through the list
> software.
>
> Obviously, no such discriminator can be perfect, so there will always
> be some noise in the system. (In fact, I think a good argument could
> be made that some noise is both necessary and desirable.)
>
> One work item might be, therefore, coming up with alternative methods
> for identifying noise. [Of course, a method might include
> characterizing individual posters by their individual S/N for the
> particular list -- a method with obvious political implications.]
>
> Once you can discriminate (however imperfectly) noise from signal
> (and bearing in mind that defintions of "noise" will change over
> time), then you have the question of how that discrimination can be
> applied to the medium. We might have another work item, which is to
> describe mechanisms for filtering noise from signal.
>
> In a general sense, we could call any such a mechanism a "noise
> filter" for a list. A human moderator is an example of a noise
> filter. Per person message quotas are another mechanism. We can
> imagine much more elaborate mechanisms implemented in software --
> filter recipes sent by participants to the list server...
>
> The particular abstraction I have been using here isn't the only one,
> of course -- there are obviously many different ways to think about
> this stuff. I just think that if we can explore the problem from an
> abstract point of view, for a while, we may get some useful insight
> that we can carry back to the ugly nitty-gritty social/political
> battleground.
>
> --
> Kent Crispin "Do good, and you'll be
> [EMAIL PROTECTED] lonesome." -- Mark Twain
Regards,
--
Jeffrey A. Williams
CEO/DIR. Internet Network Eng/SR. Java/CORBA Development Eng.
Information Network Eng. Group. INEG. INC.
E-Mail [EMAIL PROTECTED]
Contact Number: 972-447-1894
Address: 5 East Kirkwood Blvd. Grapevine Texas 75208
