MicrosoftPrototypeCrawler
Microsoft has launched a web spider. I was visited on May 29th, it requested robots.txt like a good spider and then hit some random set of pages across this server. It also made lots of totally bogus requests to things like http://minutillo.com/steve/weblog/http. It seemed to generate more 404s than actual valid requests.
It came from 131.107.163.47 and its user agent was MicrosoftPrototypeCrawler (How's my crawling? mailto:newbiecrawler@hotmail.com).
They’ve been keeping track of this spider for a while at Webmaster World. I guess it used to ignore robots.txt, and triggered some of their tricky spider traps.
Microsoft already has search at MSN but the actual results are outsourced. Recently they said they want to have their own search engine. Some product manager there said he thought it shouldn’t be too hard to make one better than Google. Maybe this is the beginning of that. Personally, I think they might consider fixing the search engine at MSDN as a warm-up problem first.