W3C home > Mailing lists > Public > ietf-http-wg@w3.org > July to September 2014

Re: consensus on :query ?

From: Poul-Henning Kamp <phk@phk.freebsd.dk>
Date: Sun, 20 Jul 2014 21:03:23 +0000
To: Phil Hunt <phil.hunt@oracle.com>
cc: Mark Nottingham <mnot@mnot.net>, HTTP Working Group <ietf-http-wg@w3.org>
Message-ID: <32892.1405890203@critter.freebsd.dk>
In message <2E9D36EF-978F-49E4-8A23-22CD3536680A@oracle.com>, Phil Hunt writes:

>There is a broader issue of how to handle queries in HTTP...
>
>Do as a separate header against any http method as suggested?
>
>Or...
>
>Do as a new http method eg QUERY to reduce overloading of POST
>and other methods. 

Uhm, I think you're misunderstanding.

My proposal is to transport the part of the url after the first '?'
in a separate pseudo-header (:query), just line we have already
split the url into :scheme, :authority, :method and :path.

The main argument is that this will improve compression, since
the :path part is much more constant than the :query part.

:query parts are and can be used with all :methods.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk@FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.
Received on Sunday, 20 July 2014 21:03:47 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 30 March 2016 09:57:09 UTC