On Sat, 2006-11-11 at 14:59 +0800, Adrian Chadd wrote: > On Sat, Nov 11, 2006, Ow Mun Heng wrote: > > > > You will still have the same TCP/IP protocol issues if not more. > > > The only problem it might solve is access those sites not happy when a > > > user changes IP when going from http to https. > > > > That's the only good thing about tproxy support? That seems a bit > > little.(to me anyway) > > It allows you to transparently intercept and cache web traffic > from clients who have real IP addresses. Its very useful for ISPs > who wish to deploy Squid (or any other kind of interception caching) > but don't want to break the "End to end" model a lot of applications > today expect. > > I've got plans to put an example or two of it in use up on the Squid > wiki after I finish this set of exams. I'll be deploying it for an > ISP in precisely this manner. > > Its not so useful for office setups behind a NAT as the clients have > no existing end-to-end public addressing in place. Now, this is more useful and precisely the answer i was seeking.. Thanks. > > > > > Adrian >
