My company uses a proxy that automatically scans downloaded files for viruses. The user is forwarded to a page showing the download and scan progress. When it's done you can download the file. It's highly annoying but scanning all (there are whitelists too though) incoming data is non-negotiable requirement.
While this is only "annoying" when a user downloads something through a browser it is a show-stopper for scenarios like "apt-get" or eclipse updates - both will obviously just fail if they are retrieve with the proxy's progress page.
Are there any ways to improve this ? The non-browser downloads could be identified through the useragent id, but how to handle them ? Just not checking files for those downloads is not an option.