Re: New draft available

[Ian Cooper <icooper@equinix.com>]

At 20:34 11/21/2000 -0500, Eric Dean wrote:
>On Tue, 21 Nov 2000, Markus Hofmann wrote:
>
> >
> > What exactly do you mean by "...client is proxied into a cache"? The
> > assumption here is that the client does NOT have to configure a proxy...
>
>I agree that the client does not have to...but there are many ISP
>implementations whereby the client is forced into such an arrangement.  In
>the event that the client is proxied in such a way, there are alternative
>non-DNS methods that are not described within this draft.

I've also been thinking about these sorts of arrangements.  This is one 
area where I could see CDNP and WREC/WEBI collaborating in the future. 
There's also the possibility of a client receiving (through something 
WPAD-like) configuration to direct requests straight to a surrogate that's 
known to be close.