Paul Benjamin asked about semigroups of partial 1-1 functions in GAP.

It is correct that the present release (3.3) of GAP does not provide
specific support for such structures, and I want to add that the plans
for the next release do not include anything for such structures
either.

Therefore we see the plan, to use the possibilities of creating new
domains using records as described in section 1.28 of ("About Defining
New Domains") of the manual, as the only realistic present
possibility.

We have been asked several times in the GAP-forum, but also privately
if we have plans to extend GAP to support structures that in one way
or other generalize the notion of groups. Therefore I would like to
use this letter in the GAP-forum to state clearly that in Aachen (for
lack of manpower as much as for any other reason) we have no such
plans in the foreseeable future. At the same time we will be open for
discussing inclusion of such structures if other groups would do the
work and we suggest that those, who consider any such ideas, use the
GAP-forum to make contacts. In particular it might be discussed to
which level of generality of such structures one should go.

Joachim Neubueser