Singular-value decomposition of the basis vectors can return
singular values that differ by 8--10 orders of magnitude. The
default value of thresh (1e-5) would discard several of these smaller
values.  That's a problem because the _inverses_ of the singular
values are actually used to produce the solution in _svd. Discarding
the largest singular values didn't seem to help either. Since it
wasn't entirely clear why some of the singular values were being
discarded, I commented out that line of code, and the associated
documentation, so now we keep all of the singular values.

I also added a detailed explanation of what fitwarp2d, _svd, and
_mkbasis actually do.  And of course, some tests.

If somebody with a lot of linear algebra experience is convinced
of a a good reason to discard some of these singular values, we
can revisit this (hence why I haven't removed the code, just
commented it out for now).

…stamble) portable.

Previously the first element of a list provided to 'cat' was used
to set the type of the output piddle.  Now it uses the "highest"
type if passed a mixed-type list.

Also added some tests for this new behavior, and improved the
documentation for cat, append, and glue to better highlight the
differences between then.

There seems to be a problem with threads and Windows and freeing
memory.  Previously the warp2d test would crash on Strawberry Perl
with a "free to wrong pool" error.

I followed the advice from BrowserUK in this PerlMonks thread:
http://www.perlmonks.org/?node_id=742205 to create a pass-through
routine to free the interolation kernel's memory.  Thanks for asking
the similar question 9 years ago, Rob!

These changes don't cause any tests to fail on my Mac, we shall see
on the AppVeyor.