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

Re: Header Table and Static Table Indicies Switched

From: Jeff Pinner <jpinner@twitter.com>
Date: Sat, 2 Aug 2014 08:55:43 -0700
Message-ID: <CA+pLO_hMuQ9osACw1bcGDUd+orw6bsY8m4i-oCAKE8hdiuN2pg@mail.gmail.com>
To: Greg Wilkins <gregw@intalio.com>
Cc: HTTP Working Group <ietf-http-wg@w3.org>
Looking at request on an internal upload API that 12 internal header fields:

On the first request I measure no change in compression efficiency
(not surprising as all header names are unique so nothing is being
reused).

On the second request I measure a size reduction of 8% when switching
the header table to occur before the static table.

On Sat, Aug 2, 2014 at 8:31 AM, Jeff Pinner <jpinner@twitter.com> wrote:
>> If static header names are more frequent than custom ones (as you would
>> expect them to be), then I think this result will be pretty much the same on
>> most data sets.
>
> This is my point, they are only more frequent on "public" data sets
> since you are getting web browser requests that have to interop with
> every website. For APIs or internal requests that is very referenced.
>
> And for those that do appear, they will be indexed and thus get a lower index.
>
> Consider this, for a client encoding requests, there is no reason to
> use the index space for response header names that will NEVER be sent.
> Putting the header table below the static table moves those request
> headers that are indexed below the static table, pushing the response
> names to a higher value.
Received on Saturday, 2 August 2014 15:56:10 UTC

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