Ben Laurie wrote: >Quibble: you have the protocol backwards. The client sends the cached ETag >with its request. If it matches the current ETag the server doesn't bother >to send the page. Oops. Right. Thank you for the correction.