Re: A proposal for Host header

• Messages sorted by: [ date ][ thread ][ subject ][ author ]
• Next message: Mark Nottingham: "Re: WG meeting in Washington?"
• Previous message: Roy T. Fielding: "Re: HTTP to "Standard"? When?"
• Maybe in reply to: Josh Cohen: "A proposal for Host header"
• Next in thread: Larry Masinter: "RE: WG meeting in Washington?"
• Reply: Larry Masinter: "RE: WG meeting in Washington?"

In message <BFF90FB6CF66D111BF4F0000F840DB850BCBBB91@LASSIE>, "Josh Cohen (Exch
ange)" writes:
>I would like to propose:
>
>1) If a client is issuing a 1.1 request and the client has obtained
>positive knowledge, possibly through an out of band mechanism, that all
>proxies and the origin server in the request path are 1.1 compliant or 
>better, that it may omit the host header when absolute URIs are used.
>
>2) Clients are permitted to use absoluteURIs when talking to 1.1 servers.
>
>Does this seem reasonable ?

Of course not.  Try it with any HTTP/1.1 compliant server now and you
will get a 400 response.  Those servers won't disappear just because
an RFC is modified.

This part of the standard will never change without a version bump.  Never.

....Roy

• Next message: Mark Nottingham: "Re: WG meeting in Washington?"
• Previous message: Roy T. Fielding: "Re: HTTP to "Standard"? When?"
• Maybe in reply to: Josh Cohen: "A proposal for Host header"
• Next in thread: Larry Masinter: "RE: WG meeting in Washington?"
• Reply: Larry Masinter: "RE: WG meeting in Washington?"