Well, its much better than the early eap release. But it /is/ still way
way slower than the text runner. For example, with the text runner (on
3.x) its takes about 2-3s for the first test dot to appear. With eap
(813) it takes about 12-13s for the first test tree-item to appear.
And again, since our entire test suite (of 2000+ tests) takes less than
7s to run completely, the graphical runner overhead is well over 100%.
Having said that, running single test cases, the performance difference
is not too noticable. (For those of you who will ask "But how often to
really run /all/ tests?" -- the answer is hundreds of times a day!
So, I again ask, will it be possible to chose the test runner? Or is
there some time scheduled for tuning the new graphical runner? Or is it
only us that really cares?