4 comments

[ 4.8 ms ] story [ 14.9 ms ] thread
Part of my project requires me to look at how different websites do search. Some websites (ridiculously) don't allow you to access their search without doing a POST (allmusic.com). Other sites do, but tag along 1,000 different GET parameters. For these sites, I usually strip off all the unnecessary stuff and see what happens.

For Bloomberg, this is what I got. It looks like they've cloned Google Search from 2005, bit for bit, and injected their own search results inside of it. Any ideas why?

Maybe they have a Google Search Appliance? (I don't really know. Just wild ass guessing. I'm pretty sure the appliance has been around that long, though.)
HTTP headers suggest GWS/2.1 (Google Web Server 2.1). The GSA seems like a pretty keen estimation.
Looks that way. Then again, the default behavior from using the search box from the home page also has a few more parameters on the URL, which result in skinning the results.