When editing large files, mupacs takes a lot of time
for harmless operations such as inserting two
characters between a pair of parentheses or finding out
that a certain paren has no matching one. EMacs is not
interruptible in these cases, which may easily take
more than 30 minutes.
I believe the main problem is that a lot of
regexp-searches are performed over the whole buffer to
find seldomly used keywords such as "axiom". It helps
to narrow the buffer, but there seem to be cases
(unfortunately not reproducible yet) where the
narrowing is undone by mupacs.
I suggest introducing a variable defining a maximum
search range for all regexp-searches.