Re: [gnugo-devel] ED66 considered harmful

[bump]

> I don't really see how this would help tracking progress. I don't mind
> starting a new strategy5.tst or owl.tst since the old ones are kind of
> overlarge anyway, but I don't see any advantage in splitting the
> smaller (or faster) specialized test suites like reading.tst,
> atari_atari.tst, filllib.tst, connection.tst, and a few more.

By more or less closing batches 1-4 we can easily distinguish
old known problems with new problems that might or might not
have been present before 3.2. Improvements in the pre-3.2 test
results definitely measure progress. On the other hand it may be
better to tune the new tests more aggressively since they
represent known recent problems. An old test could easily
represent a problem that will never appear on the board because
GNU won't play the moves leading up to it.

I did start an owl1.tst. We could move it back to first_batch
if you think it makes more sense to have it there.

> As examples of the problems it causes, I offer two test cases. The
> game records have been added to CVS, but I'm not sure which test suite
> to put the tests in. (nngs1.tst, strategy5.tst or a new nngs2.tst are
> the most likely alternatives.)

Let's not use nngs1.

Dan