Questions on mesostics and the software Cage used

Hello all,

I'm finishing up a mesostic generating program that I'm using for a
larger project. While researching for it, I came across a number of
claims about how Cage generated them.

I get that the process evolved over time. But what I'm most interested
is in the software he used to generate mesostics.

If I'm reading everything correctly, it looks like Jim Rosenberg wrote
the program MESOLIST for this purpose. I am having a difficult time
finding details about the program and hoping some people here can
help.

1. Is it available anywhere for downloading or studying?

2. Did the program generate the wing words?

3. Did the program follow only the 100% rule or allow for the 50% rule also?

4. I've read that Cage chose to leave out some wing words. Is this
true for how he used the program or was that just something he did on
the mesostics he generated by hand?

5. Right now my program chooses up to seven words on either side of
the spine word following the 100% rule for all the words. Next I will
have it limit the number of words randomly but they will still be in
order (ie, anywhere from 0 to 7 of the preceding/following words). A
mesostic generator I see online
(http://vyh.pythonanywhere.com/psmeso/) chooses randomly from all the
words between spine words following the 100% rule. My approach will
generate fragments that make sense since it can potentially string up
to 15 words in a row from the source text while the online one won't
do so. Which approach more closely matches Rosenberg's program?

I'm really a terrible programmer (I'm a composer and not a
programmer), so the fact that I've gotten this much working probably
means I won't change how it works on a fundamental level but I would
like to have an idea about how closely my approach matches the
software Cage used.

Thanks everyone,
Dave Bellows