Comment 8 for bug 234349

A link starting with "http:" is (correctly) referred to the browser application. Thunderbird itself is able to issue HTTP requests, e.g., for remote image content. Now one can conclude that it should be possible to directly resolve that link, but it's a bit more complicated than you may think:

 - the link may end in ".pdf", but this doesn't guarantee that the server
   will actually send application/pdf back (it may be just a download page);

 - the server may request authentication, what should Thunderbird do then?

So, what you are requesting would imply that Thunderbird resolves the HTTP link itself first, connect to the server, and verifies that what it gets back actually represents something different than a web page or authentication request. Then, it would be treated like an attachment based on the stated MIME type, downloaded to /tmp and opened with the registered application.

That functionality exists already in the SeaMonkey suite, which combines various applications (including mail and browser) into one. For Thunderbird, given that it is primarily a mail/news/feeds application, the question is how much browser "intelligence" should be provided beyond remote content.