test outputs are too verbose that it is hard to see what is failing

[rakyll]

See https://ci.appveyor.com/project/jessfraz/dep/build/52 for an example output of the tests.

Each test is logging all the events happens during the test rather than just printing out what has failed. It is impossible to find out what actually failed.

Log only the failure case with helpful context.

[sdboyer]

Heh, I was just making the same note in #155 . Yes, we definitely need to get the testing output under control.

[sdboyer]

Relevant to testing improvements - #282

[tro3]

I wired a "-logs" flag into the test harness branch I'm working on. Without -logs set, the stdio pipes will be silent. The WIP PR should be submitted today.

[sdboyer]

oh, we're super much better with this now - have been for a while! closing this out.