The past few days I have noticed several instances where
interrupting GAP left me with a crippled version of GAP.
One more instance happened a moment ago: I said
gap> PrintRec(N);
where N was some permutation group
[B.t.w., PrintRec does not occur in the on-line help,
only in the chapter About Gap]
and this produced much more output than I liked.
So, I gave ^C. Afterwards, N was partially mangled:

gap> N;
~
    gap>
gap> PrintRec(N);
~gap>
gap> Size(N);
729
    gap> Print(N);
~gap> c in N;
Error, Record: right operand must have '~.operations.in'

Let me take this opportunity to ask a question.
Do there exist facilities to regard an elementary abelian group
as a vector space (and vice versa)?
Do there exist facilities to go back and forth between a (permutation)
group p^n:G with elem. ab. subgroup p^n and a matrix group G
(with matrices of order n over GF(p))?

Andries Brouwer