[Web4lib] Google limit of 1,000 results

[Lars Aronsson]

Roy Tennant wrote:

> Sure. Given that Google's ranking algorithm tends to weight 
> pages that important sites point to more heavily than others, 
> there are times when I want to see what someone says about a 
> topic that 1) isn't being pointed to by anyone -- perhaps 
> because it is too new, for example, or 2) isn't linked to by an 
> "important" web site.

Sorry, I still don't see any realistic example here.

For the phrase "Academie Parisienne des Inventeurs" Google brings 
up 15 hits and the Wikipedia article (that I wrote two weeks ago) 
is not (yet) among them.  Limiting the result list to the first 10 
hits would only lose three duplicates of an Italian text (two 
copies of the same text are among the first 10 hits) and two 
expired eBay auctions.  There might be 5 other web pages on the 
topic that Google haven't found, and 50 people in Paris who know 
more but haven't yet written webpages about this.  Finding those 
people would be something, but Google doesn't really help me. This 
is the end or edge of the web.  And looking beyond it is much more 
interesting than looking beyond the end of Google's truncated 
result lists.

Setting up a blog hosting site or a wiki is a good way to make 
people write more webpages on odd topics, full of links to less 
known pages.  And it's good economy too, since people write blog 
entries on Blogger.com (now owned by Google) for free, while 
Google's engineers work for a salary.

> My point is merely that Google is good -- very good -- at some 
> important but limited tasks. It really sucks at others.

The web is limited.  I don't see how any other web search engine 
could do much better, even in theory.


-- 
  Lars Aronsson (lars at aronsson.se)
  Aronsson Datateknik - http://aronsson.se