W3C home > Mailing lists > Public > ietf-http-wg@w3.org > July to September 2007

[RFC] Optional header negotitation

From: Stefanos Harhalakis <v13@priest.com>
Date: Fri, 17 Aug 2007 15:28:24 +0300
To: ietf-http-wg@w3.org
Message-Id: <200708171528.24688.v13@priest.com>

Hi there,

  After considering a lot your comments and other feedback I got regarding 
the "HTTP Information Request" proposal [1], I still believe that header 
negotiation is a good thing (tm) to have, so I'm resubmitting a request for 
comments to this list regarding this issue. 

  Let me explain the idea: It seems to me that web based applications can 
really use additional HTTP headers but we cannot propose them because of the 
extra overhead they will introduce (like the HTTP Timezone Header). So, one 
solution will be to find a way for supporting optional headers that will be 
send on demand. This proposal does exactly this.

  Example: A client contacts a web server and sends a GET request. Server side 
script (SC) is able to take advantage of header (H) (example: Timezone 
information) to adjust its content. SC examines the headers and finds out 
that the client did not send that header. Server side script sends an extra 
header like:

Header-Request: Timezone

and goes on sending the content. Client side browser:
a) Does not support "Header-Request" and ignores it, rendering the page
b) Supports Header-Request and the 'Timezone' optional header and resubmits a 
GET request with the optional header included (It would act as a refresh).

  I believe that this has great potentials because it can be used as the basis 
for easier future HTTP expansion, without wasting bandwidth.

  I'd appreciate any comments you may have, especially any objections.

Thank you in advance,
Harhalakis Stefanos

[1] http://www.nabble.com/-RFC--HTTP-Information-Request-tf3932478.html
Received on Friday, 17 August 2007 12:28:42 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 27 April 2012 06:50:15 GMT