Is it intentional that pages that don't exist no longer return an error message?

Previously a request for something like: http://en.wikipedia.org/w/api.php?action=parse&page=testest&format=x...

Would return some friendly xml error that I could parse: <errors> <error code="missingtitle" info="The page you specified doesn&#039;t exist" /> </errors>

Now it just returns an empty page. Is there some new way to detect if an invalid page has been requested or is this simply a bug?

Thanks,

Robert