Re: [hybi] [Uri-review] ws: and wss: schemes

[Erik Wilde <[e-mail]>]

hello mark.

Mark Nottingham wrote:
> I can't help but feel that if we follow the arguments given in this 
> thread, there won't be any need for FTP, or telnet, or any other URI 
> schemes; all will be subsumed by HTTP, and eventually we won't need any 
> URI schemes at all (because it's HTTP all the way down).

http://dret.typepad.com/dretblog/2009/07/the-last-uri-scheme-youll-ever-need.html 
was something i wrote recently based on another instance where people 
were arguing that from now on, all we'll ever need is HTTP URIs and 
string matching to detect special classes of URIs, that can then be 
handled differently, if an application has knowledge of the pattern that 
needs to be matched.

while the argument for string matching for "special URIs" has been 
repeated a number of times, i still haven't seen any example in real 
life where this happened in the way predicted: plain HTTP users 
discovering useful metadata or getting useful fallback behavior, and 
other applications doing string matching and implementing non-HTTP 
behavior. are there any examples where this has happened on any 
significant scale?

cheers,

dret.