Missing you, Brian, I was contemplating a private message.
I recall that when I joined this list a previous time, it happened
that Ludwik was discussing the experimental determination of a radio active
dust wiped from a terminal or TV screen, by tracking its half life.
In the unlikely way that events have, it happened I had a survey meter in
arm's reach, and an alcohol wipe to take a specimen.
Now THAT was list value, I thought.
In slight return, I can now mention that I have been sampling a non linear
regression program suited to W95 and NT which takes one's model equation
and fits parameters. This doubles as a least squares fit (of course!) plotter
and as a solver, of sorts. NLR shareware in crippled form is free for 30
days trial, or about $60 (I think) as sellware.
This seems like a handy tool.