In his message, Ralf Dentzer asked:
A further problem I am faced with is the computation of the
ConjugacyClassesSubgroups(G). I didn't try this with GAP 3.2 yet,
but for some groups of order 256 it took several hours with 3.1.
Did these routines improve? What I am really interested in are
the normal subgroups of G; is there a simpler way to compute them?
ConjugacyClassesSubgroups can be *very* hard (especially for p-Groups),
since so many subgroups do exist. There also is a command 'NormalSubgroups',
which will only compute the normal subgroups (as closures of conjugacy
classes). In case the group still has lots of normal subgroups, one might also
consider computing the character table, and deriving the normal subgroups