Om Malik has an interesting piece today about Amazon web hosting.  

I'd love to see them do it, then I could move my archived static sites up to S3 and forget about them. 

But they're missing a key feature for web hosting -- index files. 

Suppose I put a website on S3 and map a domain to it: 

http://hello.newsriver.org/ 

Try going there, you'll get an XML error return. 

But... it does have an index file. 

http://hello.newsriver.org/index.html 

All Amazon has to do is make the first URL work, which is how web servers work by default, and boom they're in the web hosting business. 

Without it, all we can use it for is to store images, MP3s, JavaScript snippets, archives.