On Sun, 03 Jun 2001, Ian Clarke wrote:
> On Sun, Jun 03, 2001 at 05:03:29PM -0400, Benjamin Coates wrote:
> > >From [EMAIL PROTECTED] (Ian Clarke)
> > >> This system will certainly require
> > >> some kind of web of trust rankings system to provide useful search results.
> > >Not nescessarily, both Gnutella and Napster get along without them.
> > I was under the impression that spam was a rather serious problem on gnutella,
> > despite it being a non-anonymous protocol where spammers can be
> > filtered/retailiated against, and have to invest a fair amount of bandwidth.
>
> Not in my experience, however through implementation of my "unrequest"
> idea for searches could combat that problem by permitting collaborative
> filtering on search reply quality.
>
> Ian.
The problem I have with unrequests is a question of when to use them.
Say I make a request and I get back some good answers and some bad
answers. Should I unrequest to make sure the bad don't propogate at the
expense of keeping the good results from propogating as well? What if
there's only one bad answer? "one bad apple..." If people were trying
to spam the searching, I could reasonably expect bad responses in all
but the most specific, obscure searches. Then the only way to stop bad
metadata would stop good metadata as well.
Thelema
--
E-mail: [EMAIL PROTECTED] If you love something, set it free.
GPG 1536g/B9C5D1F7 fpr:075A A3F7 F70B 1397 345D A67E 70AA 820B A806 F95D
PGP signature
