W3C home > Mailing lists > Public > ietf-http-wg@w3.org > January to March 2013

Re: bohe implementation for compression tests

From: Mark Nottingham <mnot@mnot.net>
Date: Fri, 18 Jan 2013 17:52:59 +1100
Cc: James M Snell <jasnell@gmail.com>, "ietf-http-wg@w3.org" <ietf-http-wg@w3.org>
Message-Id: <90D04E0D-A284-4CB4-B5DF-3B7CCA0F262D@mnot.net>
To: Martin Thomson <martin.thomson@gmail.com>
OK, I've started to record CPU time in my refactor branch:

* TOTAL: 1012 req messages
                  size  time | ratio min   max   std
   http1       830,970  0.05 | 1.00  1.00  1.00  0.00
  simple       320,883  0.05 | 0.39  0.07  0.92  0.24
   spdy3        85,492  0.06 | 0.10  0.03  0.66  0.08

* TOTAL: 1012 res messages
                  size  time | ratio min   max   std
   http1       424,075  0.04 | 1.00  1.00  1.00  0.00
  simple       176,216  0.12 | 0.42  0.11  0.95  0.12
   spdy3        80,706  0.07 | 0.19  0.04  0.68  0.09


https://github.com/http2/compression-test/tree/stream-sep


On 17/01/2013, at 12:21 PM, Martin Thomson <martin.thomson@gmail.com> wrote:

> On 16 January 2013 17:11, Mark Nottingham <mnot@mnot.net> wrote:
>> Getting there, although you may need a small truck to haul the grain of salt that will accompany it...
> 
> Even if it just means running the sample set n times using 'time', it
> would be nice to get ballpark figures.  Even if the errors are
> enormous.  For instance, a python implementation of delta is probably
> not that good to compare to an optimized gzip implementation, but it
> might still let us know that deltav3 is better than deltav2.
> 

--
Mark Nottingham   http://www.mnot.net/
Received on Friday, 18 January 2013 06:53:27 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 18 January 2013 06:53:29 GMT