On Mar 6, 2008, at 11:41 AM, Stephan Koops wrote:
> The idea is the following: A (human) user want to download a
> document. He prefers his own language (e.g.: "de"), but english or
> frensh is also ok. Because he also want to print it, he preferes
> PDF, but OpenOffice or Word is also good; if not available, he want
> the document in HTML.
> Then he type "http://examle.com/path/document.de.en.fr.pdf.odt.rtf.doc.html
> " in his browsers address line. That's very flexible for the user,
> and he must not check a lot of links.
> It's rigth, that we reinvent the flexible HTTP content negotiation
> mechanism with a new mechanism here. But on the other hand we offer
> another way, which is easy useable in a browser.
>
Some automated support for distinct URIs is one thing but this goes
quite a bit further. I can see that could be fun to experiment with in
an implementation but I'm wary about baking something like that into
the JSR.
Marc.
---
Marc Hadley <marc.hadley at sun.com>
CTO Office, Sun Microsystems.