----- Original Message -----
From: "Brad Wyble" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Tuesday, February 25, 2003 6:05 PM
Subject: Re: [agi] really cool


> >
> > > They are not mapping to IP addresses, probably geography as Ben
suggests.
> > I went to the search window and intercepted searches done by other
people.
> > >
> >
> > Ockham's razor is useful for this one...I believe it's a simple
server-side
> > FIFO system...NB Google's not mapping [EMAIL PROTECTED]
> >
> > For additional practice encounters with simplicity try...
> > http://cs.bluffton.edu/~scoffman/trick.html ...but don't cut yourself
> > shaving ;-)
> >
> > Just my $0.02 worth. EGHeflin
>
> Nice illustration, although that's a fairly simple one to guess.
>
> In this case, I can't imagine a simple FIFO would work because it can't
scale.  You are essentially proposing that as soon as any person in the
entire world submits a voice request, everyone else in the world who's
viewing that page gets their search results.
>
> It would work for a very small number of users, but beyond a certain
number, your search results would only be available for hundreds of
milliseconds if not less.
>
> If Google applied the FIFO to a rough geography mapping, they would cut
the userbase into manageable chunks.  Mapping IP address to phone exchanges
isn't hard if you've got the data.  Cell phones are tricky though, but then,
it's possible that their network can reverse map to the cell network that
handles your call, rather than the phone number itself.
>
> I'm willing to believe Google can get this information.

Interesting idea...should be testable and in a number of ways.  But keep in
mind you want to be testing what Google is actually doing not what they
might be able to do with this or that piece of information or technology.

It seems to me that if both you and a friend can access nearly
simultaneously the same results from a 'speak search', e.g. say "AGI", then
the server-side interpretation is proved.  If you're happier with a
client-side interpretation try the system for a land v. based cell call and
compare results.

etc....I'm open to testing... ;-)

 Just my $0.02 worth. EGHeflin

>
> -Brad
>
>
> -------
> To unsubscribe, change your address, or temporarily deactivate your
subscription,
> please go to http://v2.listbox.com/member/[EMAIL PROTECTED]
>

-------
To unsubscribe, change your address, or temporarily deactivate your subscription, 
please go to http://v2.listbox.com/member/[EMAIL PROTECTED]

Reply via email to