Re: Identification of documents in Web applications

On 3/9/2011 10:40 AM, Nathan wrote:
> FWIW, I agree with Ashok, I don't really want to be developing client side
> applications which require a server for every view/state change

Of course not, please read my note. What I'm suggesting is that we need, 
over time, to evolve clients to the point where changing at least the query 
parms, and perhaps other parts of a URI, do not necessarily result in 
server access.  Once that's done, ? becomes superior in my opinion. In 
addition to what one might perceive as more architectural complexity with 
#, there's a significant practical drawback, which we see clearly with 
things like Twitter #!: they pretty much can't work right from 
non-Javascript clients, and are quite hard to make crawlable. ? seems far 
superior from that point of view.

Noah

Received on Wednesday, 9 March 2011 16:32:56 UTC