Until now, even though customers were able to configure an Amazon S3 bucket as a website and host static HTML pages but they didn’t function like a website. A visitor attempting to access the root of the website (e.g. http://www.mywebsite.com/) would see the list of objects in the Amazon S3 bucket instead of the website’s home page. Also, if an error occurred, users would see an Amazon S3 error message instead of a website specific error message.
Amazon has now added support for root and custom error documents to address these issues. Users can now add an index.html pages to the root of their website and also add custom 404 error pages, making it possible to host entire websites on Amazon S3 without the need of a web server.