Accept extension (was: Re: Server-side data conversion ....)
Mon, 26 Jun 1995 14:48:54 +0100 (BST)

Message-Id: <>
Subject: Accept extension (was: Re: Server-side data conversion ....)
To: (Jon Wallis)
Date: Mon, 26 Jun 1995 14:48:54 +0100 (BST)
In-Reply-To: <> from "Jon Wallis" at Jun 26, 95 07:52:01 am

Jon Wallis wrote:
-->  [other people's stuff snipped]
--> This all seems to be getting out-of-hand - HTML is now getting so complex
--> (by comparison with HTML "1" anyway) that it is no longer fulfilling its
--> original purpose (or at least it's harder to see the wood for the trees).
-->  [snip]
--> I think most of the action *should* be server-side (i.e., only send what the
--> browser wants or can handle), but this requires a start-up dialog between
--> browser and server to establish what the browser *can* handle (not too
--> difficult to establish).

How about extending the Accept header to pass this information.  This
wouldn't require as much effort as adding a new header surely, and would
fit neatly into the syntax and semantics of Accept thus:

(modified from section 5.4.1 of the March 1995 IETF HTTP/1.0 draft)

Accept: "Accept" ":" 1#(
            [ ";" "q" "=" ( "0" | "1" | float )]
            [ ";" "mxb" "=" 1*DIGIT ]
            [ ";" "colordepth" "=" 1*DIGIT ])

giving rise to headers such as:
    Accept:  image/png; q=0.8; colordepth=8, image/png; q=0.7

which would be interpreted as: "if you have an 8 bit colour PNG, send it;
otherwise send me any PNG you have".

It seems to me that the colordepth preference could also apply to video
types as well as image types.

Stewart Brodie
Dept. Electronics & Computer Science, Southampton University, UK.   <-- running on my Risc PC