[ZODB-Dev] Browser send HEAD is weird?

[Toby Dickenson]

On Friday 06 Sep 2002 3:50 pm, Chris Withers wrote:
> Andrew Kuchling wrote:
> > On Fri, Sep 06, 2002 at 12:57:19PM +0100, Chris Withers wrote:
> >>Isn't that how a lot of browser caching works as HEAD responses are
> >>supposed to be cheap to generate?
> >
> > I doubt it, because doing first a HEAD and then a GET would require
> > two TCP round trips.  Caches and browsers (correct ones, anyway)
> > should be doing a GET with an 'If-Modified-Since' header.  See sectio=
n
> > 14.25 of RFC2616:
> > http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html .
>
> Fair enough :-)
>
> So, what are HEAD requests used for then?

Things that never need the response body, or which will fetch it in a=20
different way if needed. This includes link checkers, web crawlers, downl=
oad=20
accelerators, etc.