diff options
author | jar@chromium.org <jar@chromium.org@0039d316-1c4b-4281-b951-d872f2087c98> | 2009-06-30 16:31:54 +0000 |
---|---|---|
committer | jar@chromium.org <jar@chromium.org@0039d316-1c4b-4281-b951-d872f2087c98> | 2009-06-30 16:31:54 +0000 |
commit | e695fbd6d62e9967d6256cd0667eca2fb2bb918d (patch) | |
tree | 58e196031e02447b31bd9a4182b8fb3f4d7ea81b /chrome/common/chrome_switches.cc | |
parent | 5c7c19d83d201fa23d5097cce108c033ece5c63b (diff) | |
download | chromium_src-e695fbd6d62e9967d6256cd0667eca2fb2bb918d.zip chromium_src-e695fbd6d62e9967d6256cd0667eca2fb2bb918d.tar.gz chromium_src-e695fbd6d62e9967d6256cd0667eca2fb2bb918d.tar.bz2 |
Create A/B test of SDCH
To do this, I needed to add the feature that ALL FieldTrials that are
established in the browser process are forwarded and established in
the corresponding renderer processes. This then allows both DNS impact,
as well as SDCH inmpact (and any other field tests) to be studied
at the same time in a single binary.
This checkin also establishes a pattern that when we're doing A/B tests
via a histogram such as RequestToFinish, that we produce names for
all groups, rather than leaving one group as the "default" or "empty postfix"
group. This is critical for naming various sub-groups when a multitude
of tests are taking place at the same time.
BUG=15479
r=mbelshe
Review URL: http://codereview.chromium.org/150087
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@19595 0039d316-1c4b-4281-b951-d872f2087c98
Diffstat (limited to 'chrome/common/chrome_switches.cc')
-rw-r--r-- | chrome/common/chrome_switches.cc | 20 |
1 files changed, 10 insertions, 10 deletions
diff --git a/chrome/common/chrome_switches.cc b/chrome/common/chrome_switches.cc index cd9da9e..be103f9 100644 --- a/chrome/common/chrome_switches.cc +++ b/chrome/common/chrome_switches.cc @@ -473,16 +473,16 @@ const wchar_t kDisableAudio[] = L"disable-audio"; // is cleaned up and playback testing completed. const wchar_t kSimpleDataSource[] = L"simple-data-source"; -// Some field tests may be performed in the browser, and the randomly selected -// outcome needs to be propogated to the renderer to appropriately modify the -// histogram names that will be tested. This command line argument is only -// parsed by the renderer, and consists of a field test name, and a forced -// selection of an outcome. For example, if a field test "DnsImpact" has -// selected "_disabled_prefetch" as a current test, then the render should be -// passed the command line: -// force-fieldtest=DnsImpact/_disabled_prefetch -// The renderer will then force said named field test to exist, and will force -// the selected outcome to have the indicated text value. +// Some field tests may rendomized in the browser, and the randomly selected +// outcome needs to be propogated to the renderer. For instance, this is used +// to modify histograms recorded in the renderer, or to get the renderer to +// also set of its state (initialize, or not initialize components) to match the +// experiment(s). +// The argument is a string-ized list of experiment names, and the associated +// value that was randomly selected. In the recent implementetaion, the +// persistent representation generated by field_trial.cc and later decoded, is a +// list of name and value pairs, separated by slashes. See field trial.cc for +// current details. const wchar_t kForceFieldTestNameAndValue[] = L"force-fieldtest"; // Allows the new tab page resource to be loaded from a local HTML file. This |