Trac'ing tests status

[Allison Randal]

Michael Hind wrote:
> 
> Here we have a situation where we have major failures in our HLL's and 
> we are not catching it in the tests.

HLL developers (like all users of moderately complex systems) have a 
habit of pushing Parrot to the limits, and using features in ways they 
were never intended to be used to make it do what they want.

When they work around the expected interface, it tells us where we need 
to develop new interfaces. When they touch on edge cases that aren't 
currently tested, it shows us where more tests are needed.

Allison