The following reply was made to PR protocol/875; it has been noted by GNATS.
From: Dean Gaudet <[EMAIL PROTECTED]> To: Mark Herman II <[EMAIL PROTECTED]> Subject: Re: protocol/875: force-response-1.0 bug Date: Wed, 16 Jul 1997 13:02:06 -0700 (PDT) On Wed, 16 Jul 1997, Mark Herman II wrote: > telnet into www.jcount.com port 80, and type the following: > GET > /cgi-bin/counter2.cgi?secondary_exposure=true&increment=false&counter_id=30 > HTTP/1.1 > Host: www.jcount.com You are making an HTTP/1.1 request. If you were truly emulating what the browser was doing then you'd probably be making an HTTP/1.0 request and including the useragent. The force-response-1.0 does not change the response except for the HTTP/1.0 200 OK piece. The rest of the response is valid in the dialect that the request was in. The HTTP/x.y part of the response is NOT an indication of the dialect of the response. Apache's response to an HTTP/1.0 request is valid HTTP/1.0. Apache's resposne to an HTTP/1.0 request with force-response-1.0 is valid 1.0. Apache's response to an HTTP/1.1 request with force-response-1.0 is undefined, and as you've pointed out, invalid 1.0. But I doubt we'll be fixing that, since no 1.1 client should ever need force-resposne-1.0. Dean
