why phantomjs failed to capture pages whose url not end with .html?

[mnz hz]

for example, failed in the following example. 

phantomjs rasterize.js https://discuss.atom.io/t/having-problems-with-atom-on-windows/14069 aaaa.png

Unable to load the address! 

but it's ok to capture pages end with .html

phantomjs rasterize.js http://phantomjs.org/api/webpage/method/open.html aaa.pdf 

so, is this a bug?