• Guest, The rules for the P & N subforum have been updated to prohibit "ad hominem" or personal attacks against other posters. See the full details in the post "Politics and News Rules & Guidelines."

how does google do it?

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Shalmanese

Platinum Member
Sep 29, 2000
2,157
0
0
I am pretty sure that google does not use any metatags but searches ALL the document for the text you want as it finds, and quotes it, from the page. It also has all the web pages cached which would be a huge job. I am also pretty sure it is not a meta engine but a plain text-based engine.

the major difference I am aware of between windows search and gfoogle search is that windows searches for *cat* whereas google searches for <space>cat<space>.

A sudden insight that just came to me is that they might have all the web pages ranked in poularity and display the time took to search the first 10 rather than all of them (though how would it be able to display the &quot;there are approx 8 billion webpages with the word cat&quot; thing

I know that it is really heavily optimised but I think it is almost reaching a physical limit (SCSI HD random access times are around 0.009 s so if it is looking at 10 results, that is 0.09 tho parrelism would help alot).

also, what kind of internet backbone would they have there? any chance of them setting up a ftyp site? :)
 

Rendus

Golden Member
Jul 27, 2000
1,312
0
71
It's a good mix of a good backend (SQL), massive quantities of hardware (they had 4,000 machines as of May 31st, 2000), and extremely good algorithyms (spelling, anyone? :)).

Mostly, it's a very highly optimized SQL backend I'm guessing.
 

thEnEuRoMancER

Golden Member
Oct 30, 2000
1,415
0
71
Hey, Shalmanese, just read the article. You are thinking in the wrong direction (it happened before to you, huh?)
 

Shalmanese

Platinum Member
Sep 29, 2000
2,157
0
0


<< Google stores all of the actual documents it crawls in compressed form >>





<< Another intuitive justification is that a page can have a high PageRank if there are many pages that point to it >>





<< Also, it is interesting to note that metadata efforts have largely failed with web search engines >>

(which I assume means that they are not running meta :)

where else have I been misguided in my assumptions?
 

Czar

Lifer
Oct 9, 1999
28,510
0
0
searching in windows and searching in unix is so different. When I was in my unix class we had a HUGE txt file and it took extremely little time to find anything in it, just with a simple command.
 

ASK THE COMMUNITY