[thelist] PHP Search Engine

sbeam sbeam at onsetcorps.net
Fri Mar 17 13:11:41 CST 2006


On Thursday 16 March 2006 05:32 pm, Paul Bennett wrote: 
> Thanks for everyone's feedback on my [opinionated] posts. It's been 
> good to get actual user feedback from a couple of search software 
> users.

I agree with you, they are all terrible ;) seriously, have not invested 
hundreds of hours in configuring them (nor will I), but having used 
htdig, mnogo and swish-e - the search results are just not what we are 
used to from the big G (or even Yahoo, msn, inktomi...) As you say, 
text that is clearly in a <h2> can't be found. Or search for some 
common term used on the site, and you get all the site pages in 
apparently random order, no real relevance.

Which is why it would be really nice if some SE provided an XML-based 
API, so you could just place a GET/REST request of some kind, and parse 
and display the results to your liking. So you wouldn't have to 
recompile or install anything, or have a "search appliance" taking up 
cabinet space, or do any configuration. And as it turns out...
http://www.google.com/apis/index.html

But this is just a beta for non-comm use only and <1000 queries/day. I 
know the ad-based-revenue reasons why all SEs don't do this, but if 
people were paying around $.01 per query for something like this it 
might turn profitable....

So anyone know any other SEs that have anything similar?


-- 

# S Beam - Web App Dev Servs
# http://www.onsetcorps.net/



More information about the thelist mailing list