On Sat, 2011-01-29 at 04:16 +1300, RussellMc wrote:
> > I think you're misunderstanding how google works.
>=20
> No.
> I'm understanding how they SAY it works and to some extent how it works.
>=20
> They SAY that a leading - should exclude entries with that string.
>=20
> Once upon a year it did.
> Now it doesn't.

Actually, did you actually SEE it not work?

I just tried it. Ignoring the "About 496,000 results" vs. the "About
1,380,000 results" the actual results (when clicked through) are
correct: the second search resulted in pages with the words "the fsm"
but NOT "flying spaghetti monster".

Google uses MANY techniques to reduce the workload on it's farms. It's
not at all surprising that different techniques are used to chop down
how many hits are returned based on the KIND of search you perform.

My "guess", is that your simple search of just "the fsm" results in a
super fast but limited search, which only returns 496,000 hits. The
second search spawns a more compute intensive search (lots of shortcuts
on the first search can't happen when you've got the AND NOT in there),
and ends up returning more results.

I don't know the why exactly, but my very limited experience in search
large datasets tells me something like this isn't out of the realm of
possibility.

To prove this point, try the following search:
the fsm
-"somethingcompletelygibberishwithrandomcrapattheendalskdjfoiahs08vh"

Note the results: "About 1,380,000 results"

See? The exact same "hits" as the second search. It's the change in
search algorithm that's resulting in the change in hits, NOT the actual
search terms.

The major thing though is all of this is moot: NOBODY is going to check
the 496,000 results the first search returned and find that isn't
enough, so who really cares?

Google is not broken, it's just smarter then us. I for one welcome our
google overlords... :)

TTYL=20

--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.