Some memory leaks and other bugs in elliptic curve point counting have been fixed.
The choice of field representation for GF(2k) for small k has been improved.
A bug with the AutomorphismGroup of a lattice has been fixed. The group has the same generators, order and base, but the set up of the strong generators has been corrected. The bug was found as the group gave an incorrect FPGroupStrong. Note that there was no problem with the generators or order of the group computed.
A bug with the map returned by PCGroup of a trivial matrix group has been fixed. Bug reported by Mark Stather.
A bug where the IsConjugate function for subgroups of GrpPC sometimes returned an incorrect conjugating element (along with a true result) has been fixed. The problem only applied to conjugation of subgroups, not elements.
A memory management problem that sometimes caused a crash when doing arithmetic in GPC-groups has been fixed. Bug reported by Mark Stather.
The returned element from TransversalElt has been made to be of type GrpPermCoxElt, where before it could have been either that or GrpPermElt. Bug reported by Markus Grassl.
A bug with constructing commutators of elements of Lie groups has been fixed. Bug reported by Sergei Haller.
A bug in IsIsometric (taking sequences of forms) has been fixed (reported by Gabi Nebe).
See Matrix Group item to do with AutomorphismGroup of a lattice.
Bugs in Rank and Eigenspace for matrices over euclidean rings with zero divisors have been fixed (reported by Sergei Haller).
A bug in HenselLift when lifting roots of degree 1 polynomials has been fixed.
A bug in Factorization for polynomials over the integers has been fixed (reported by Scott Murray).
A bug in Resultant and Discriminant when the input contains non-trivial content has been fixed (reported by Michael Stoll).
A problem involving quotients of algebraic numbers by integers has been fixed. While the division was correct, it could induce arbitraryly wrong results later on.
Negation of elements of a differential ring has been fixed.