- From: Mike Meyer <mwm@contessa.phone.net>
- Date: Tue, 24 Sep 1996 20:37:56 PST
- To: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
> As I believe Jeffry Mogul has already presented, it would break the > assumptions of some very important caches if they can't assume idempotence > of GET. This is true. On the other hand, the functionality is sufficiently important that I routinely build applications that do non-idempotent GETs when I'm building for a captive audience. >From what I can tell, the easiest way to provide this functionality is by providing a mechanism for a response to say whether or not the request that it is in response to is idempotent or not. Possibly extensions to HTML - adding a METHOD attribute to the A tag, and allowing FORM elements in the tag - would do the trick. However, that feels like a kludge to me. If everything involved is HTTP/1.1, then there isn't a problem. The transition period could well be interesting, but that's not unusual when providing new functionality to an old mechanism. Other Since this feature only applies to the RESPONSE, the person generating the response has the responsibility (sorry) for making sure that the caches/clients in the chain are all HTTP/1.1, just like they now have the responsibility to make sure that the GET is idempotent (or that it not being idempotent won't break things for any users). I've not given the proxy/caching in 1.1 to know whether the above constraint is necessary, but it should certainly be sufficient. <mike
Received on Tuesday, 24 September 1996 20:49:48 UTC