Recently we implemented resumable download feature at our web site. Through this feature user can pause a download, and resume at any time back. This feature also supports download accelerators.
Is this a big deal? Certainly not. We used to server trial downloads in a public accessible directory and web server automatically has this feature. However, we want to our user to get *versioned* filenames, as it will be easy for our users to tell which version they downloaded. Implementing resumable downloads in a scripting language is not a trivial task.

Resumable download in Firefox

Resumable Download in Internet Explorer