IM URL Syntax
But the problem is, according to Brent, who is researching it, there is no standard way to refer to an account on an instant messaging service. It's funny because we didn't know that when we did tcp.im in Radio and Frontier, we just went ahead with this format: service://screenname/, and no one seemed to object (if they did we didn't hear from them).
In keeping with the nature of URLs, wouldn't something like:
im:aim/YourAIMHandle or im:jabber/YourJabberID
im:/aim/YourAIMHandle or im:/jabber/YourJabberID
be more appropriate? It's likely at some point there will be convergence in the IM space and then specifying individual, proprietary services as the protocol portion of a URL will seem silly. Rather, the point is that you want to send an instant message to a particular host (service) for a specific path (user).
Here's little-known and oft-disputed historical tidbit regarding URLs. Anyone who's familiar with parser generation will notice and question all the extra tokens in a HTTP style URL. Why is it that the protocol HTTP is separated from the host name by a three character token, "://"? As I recall from the original discussions about URLs being had with the guys at CERN in '91-'92, there WAS a reason. In the original URL scheme, protocols were separated from paths by a ":" token, which is why we see URLs like "telnet:localhost" without all the spurious "/" characters.
For HTTP, it was assumed that content may reside on different networks. Anyone who remembers the early Internet (pre-NREN, 1985 or so) will recall that there were LOTS of different, fragmented networks out there. CS-Net, CERFNet, BayNet, SesquiNet, etc. all served their little corners of the world. Some were TCP/IP, some were X.25, some were IBM-based, and all talked through gateways. The space between the slashes in a HTTP URL was originally supposed to be where the gateway or network name was specified. So a URL might have looked like "http:/csnet/www.somewhere.edu/file.html" if the modern Internet hadn't sprung out of its commercialization in 1991.
The short portion of the long story is that the URL syntax I proposed above for instant messages falls nicely into the original (if not overcome by events) syntax for URLs as envisioned by the creators of URLs in the first place.