On Wed, 30 Jul 2003, Tod Harter wrote: > Oh, I understand, but either a request IS or IS NOT idempotent, and the fact > that a client sends a HEAD request is NOT what determines that! > > The point is that HEAD is fairly useless with dynamic web applications. Its > just the nature of the beast. Proper up-front cache-control is where you > really have to deal with things. Sometimes you can use HEAD to your advantage > though. I guess the upshot is that (getting back to the original point ) > there is no one proper way to do it, some HEADs should replay the entire > request, others shouldn't.
If HEAD was a very quick short-circuit in AxKit it would bypass generation steps. That would give you different headers (e.g. text/xml instead of whatever content-type you generate) than if you did the equivalent GET. So in effect AxKit has to run it's whole processing cycle. If the semantics should be different in actual code terms (e.g. in XSP) then that's up to the developers of the XSP page to deal with. --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
