In https://bugs.webkit.org/show_bug.cgi?id=175926, we found out that we weren't testing release build of Ember, and that release build of Ember is 3.5x-4x faster than debug build of Ember on Chrome/Safari. When we decided to use a release build of Ember instead of a debug build, we thought the difference was negligible due to an incorrect measurements we made in the past where we were basically testing the same debug builds twice. Since 4x difference indicates a debug build of Ember has a very different performance characteristics compared to a release build, we should keep the current version as EmberJS-Debug test.
Created attachment 319023 [details] Adds Ember-Debug-TodoMVC
Somehow the patch doesn't show the files I'm copying presumably because I directly used svn cp. Note that this is the current version of the test we have for Ember so there is no need for qualifications, comparisons between browsers, etc...
Comment on attachment 319023 [details] Adds Ember-Debug-TodoMVC r=me
Committed r221205: <http://trac.webkit.org/changeset/221205>
<rdar://problem/34088816>