Re: ATM Foucault test accuracy

[Mel Bartels <mbartels@efn.org>]

We can do another round of "mirror mirror on the wall, who judges the
figure best of all?" if you people wish.  We need 2 mirrors: 1 that is
closely figured with minimal error, and one that has medium grade
errors.  That way we can test the hypotheses that "__________ <-fill in
your test method that you wish to employ> can accurately diagnose small
and medium grade errors for the purposes of figuring and mirror
performance rating".

We also need people signed up and a time schedule that is adhered to, so
that we can each have a couple of days to test, then ship the mirrors onto
the next person in line.  The time sched is important so that we can work
around vacations and the like.  You have to pay shipping to the next
person.

Only at the end will someone perform an interferometric test (hopefully
two of them).  In the first "mirror mirror on the wall", someone did
perform the interferometric test as their choosen test.  It is important
that you send test results to one person, and not discuss them amongst
yourselves, otherwise you get 'knowledge creep' of the mirrors, which
invalidates the hypothesis test.

This would be a good chance to see how you stack up with others.  Take the
mirrors to your local TM group, but caution them to not share with others
outside of their group knowledge of the mirrors.

Mel Bartels