commit 0356db4d4e9807699e162f97f8a9d86dc8c6926b Author: Lunar lunar@torproject.org Date: Fri Sep 6 18:11:15 2013 +0200

Use Selenium with the Tor Browser instead of plain Firefox

If we want Torperf to gather realistic data, it makes more sense to use the browser recommended by the Tor Project to access the web — the Tor Browser — instead of Firefox.

As the Tor Browser is based on Firefox, driving other Firefox versions should not be too complicated, but the primary target should be the Tor Browser.

This also will also give us a framework to test the performance of Tor Browser (present and future) patches. --- 2013/torperf2/torperf2.tex | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/2013/torperf2/torperf2.tex b/2013/torperf2/torperf2.tex index b3eb955..1c643c0 100644 --- a/2013/torperf2/torperf2.tex +++ b/2013/torperf2/torperf2.tex @@ -207,7 +207,7 @@ stem or another parsing library.

\subsection{High priority experiments}

-\subsubsection{Alexa top-X websites using Selenium/Firefox} +\subsubsection{Alexa top-X websites using Selenium and the Tor Browser}

One major reason for rewriting Torperf is to make its results more realistic.% @@ -216,10 +216,10 @@ It was suggested to track down Will Scott's torperf-like scripts, make them public if needed, and do a trial deployment somewhere.% \footnote{\url{https://trac.torproject.org/projects/tor/ticket/7516%7D%7D An experiment making these more realistic measurements should use -something like Selenium/Firefox to control an actual browser to make +something like Selenium to control an actual Tor Browser to make requests. -As a variant, this experiment should be run with a Firefox that uses -optimistic data.% +As a variant, this experiment should be run with other Firefox versions, e.g. +that one that uses optimistic data.% \footnote{\url{https://trac.torproject.org/projects/tor/ticket/3875%7D%7D

\subsubsection{Static file downloads}