- From: Jonathan Borden <jonathan@openhealth.org>
- Date: Mon, 28 Jul 2003 15:45:21 -0400
- To: <www-tag@w3.org>, "Norman Walsh" <Norman.Walsh@Sun.COM>
Norman Walsh wrote: > > Think of it this way: the network is full of proxies. On my laptop, > I'm running a local proxy, so if I request http://www.w3.org/TR/, I > might get my cached copy and be none the wiser. Even if the request > escapes off my laptop, it may get served back by Sun's firewall proxy > and beyond that by other proxies on the network. By extension, I have > no trouble accepting that the web server running on norman.walsh.name > functions as a proxy for me when requests are made for > /knows/who#norman-walsh. Similarly, I have no trouble with Tim's > server proxying for him when requests are made for > http://www.tbray.org/ongoing/misc/Tim. > As far as I can tell for both of you guys, the HTTP request is displayed on a wireless terminal, and you both type in your HTTP responses on the keyboard. I guess having various caches is a really good thing. In all seriousness, this is a critical function of the current Web architecture. We think that when an HTTP URI is dereferenced, that some communication between the client and resource occurs and that somehow the resource itself emits a representation of itself. That is not the case. There need not be any direct communication between the client and the server identified by the DNS name in the URI. Requests can be redirected and responses cached. The URI really is being used as an identifier by which some information is obtained. From _where_ the information is obtained is not necessarily, or even preferably, from the resource. Jonathan
Received on Monday, 28 July 2003 15:45:29 UTC