Unlikely, for performance reasons. Super Search performs a much more detailed scan, and hits the database much harder than Search does. By keeping the common case the less expensive (albeit less powerful), we keep the site responsive.

I sometimes use Google to search the Perl Monks archive. It works pretty well... they have the hardware we don't.

How about using Google's API within a search box here?

The way i read their agreement, that might be OK, though limited to 1000 searches a day.

The search power of Google combined with the favorable signal-to-noise ratio here could be mighty fine indeed!

There are lots of threads here related to Google, it's API and WWW::Google, just use the search box on "google api"

just a thought...

This might not be a bad idea. You could cache the results so that if you were at the limit, you could just show the slightly staler cached results. Perhaps even do the search for each word of the query (after removing common words like the) rather than the whole string and then merging the combined results for display.

-Lee

"To be civilized is to deny one's nature."

I don't think that this is a good idea. Relying upon a crippled, experimental program which if it succeeds will most likely become commercial isn't a good idea for PM with its amount of traffic, limited budget, and emphasis on open source software.

A fully operational Super Search will be a very useful tool. One would not want it beholden to Google's commercial wims.

()-()
 \"/
  `                                                     

If I look for "regexp parameter" I want all titles that contain BOTH "regexp" AND "parameter" at the same time. If I want to look at nodes that contain one of these, I can do two queries.

Don't use search exactly because of this.

Bye
 PetaMem