--- title: "Magma V2.26-11 Patch Notes" canonical: "https://docs.magma-maths.org/patchnotes/2.26-11.html" source: "https://docs.magma-maths.org/patchnotes/2.26-11.md" --- # Magma V2.26-11 Patch Notes ## Algebraic Geometry - An error check in the intrinsic [`ProductProjectiveSpace`](../AlgebraicGeometry/Schemes/schemes-ambient.md#function-productprojectivespace-rng-seqenum) for an empty sequence has been added. ## Algebraic Number Fields - `Magma` level printing of towers of algebraic number fields has been fixed. Reported by A. Previtali. - A fix has been made to the embedding improvement in V2.26-10. Reported by A. Pacetti. - `IsQuadratic` applied to an order of degree 2 which is not a radical extension has been fixed. Reported by J. Swenberg (comm. by J. Voight) and T. Keller. - `div` of elements is now possible in degree 2 fields which are not radical extensions. ## Algebras - A bug in reduction of the final Groebner basis computed for some types of ideal of a free algebra has been fixed. Reported by G. Brown. ## Arithmetic Geometry - A bug in `Magma`-level printing of points on Jacobians of hyperelliptic curves has been fixed. Previously, in some cases evaluation of the resulting string would fail. - An old bug in ReducePoint() (and ReducePointNF()) has been fixed by Michael Stoll. The effect of the bug was that it could violate the specification of the Mumford representation when the reduced curve has one point at infinity, but the original curve has two points (geometrically). - Michael Stoll fixed a bug in his code for computing Mordell-Weil groups of genus 2 Jacobians over the rationals. - The intrinsic [`LocalTwoSelmerMap(2)`](../ArithmeticGeometry/EllipticCurvesOverQAndNumberFields/over_anf.md#function-localtwoselmermap-rngordidl) has been revised to work properly for the rationals. - Triggered by a bug report by Daniel Hast, a number of bugs were found and fixed in the implementation of the Chabauty method for both elliptic and hyperelliptic curves by Michael Stoll and Danial Hast, with input from Nils Bruin. - A fix was made to code handling a special case of the height machinery. - Fix the intrinsic `LSeries` when applied to an object of type GrpHeckeElt so that it ignores the precision parameter in the case of the trivial character. - A verbose print message in [`FourDescent`](../ArithmeticGeometry/EllipticCurvesOverQAndNumberFields/mordell-weil.md#function-fourdescent-crvhyp) was accidentally being printed always. This has been fixed. Reported by R. Rathbun. ## Graph Theory - A bug where the `Converse` function returned an undirected graph rather than a digraph has been fixed. Bug reported by Akihiro Munemasa. ## Group Representations - `Magma` level printing of G-modules no longer uses the string “G” for the group, and instead properly prints a Magma-evaluatable representation of it. - A bug in [`BasicAlgebraOfEndomorphismAlgebra`](../Algebras/BasicAlgebras/algebras.md#function-basicalgebraofendomorphismalgebra-modrng) where the dimension of the resulting algebra was wrong has been fixed. Reported by B. Sambale. - A crash in [`BasicAlgebraOfGroupAlgebra`](../Algebras/BasicAlgebras/algebras.md#function-basicalgebraofgroupalgebra-grpperm-fldfin) has been fixed. Bug reported by D. Benson. ## Group Theory - Bugs in [`IdentifyGroup`](../FiniteGroups/DatabasesOfGroups/smlgps-database.md#function-grpdata-identifygroup) have been fixed. One bug (an apparently infinite loop) reported by D. Roe, another crash by an unknown user. - A crash in `Complements` for a pc-group has been fixed. Bug reported by T. Dokchitser. - The function [`NumberOfSmallGroups`](../FiniteGroups/DatabasesOfGroups/smlgps-database.md#function-grpdata-numberofsmallgroups) now returns 0 when the given order is not in the database. Previously this was a runtime error. This change was requested by several users. - The [`SmallGroupProcess`](../FiniteGroups/DatabasesOfGroups/smlgps-database.md#function-smallgroupprocess-rngintelt) code has been made more efficient in the case where there is no predicate given. Report from S. Glasby. - Homomorphisms defined from an abelian group to another group type now allow calculation of the inverse map. Reported by T. Dokchitser. - A bug in constructing homomorphisms with domain an automorphism group of a group has been fixed. The bug caused a erroneous runtime error reporting bad images for the hom constructor. Bug reported by D. Roe. - A crash when computing `FPGroup` of an automorphism group of an infinite abelian group has been adjusted to become a runtime error. Computing the fp-group is under consideration. - A crash in [`AutomorphismGroupSolubleGroup`](../FiniteGroups/FiniteSolubleGroups/automorphism-group.md#function-automorphismgroupsolublegroup-grppc) has been fixed. Bug reported by D. Roe. - A crash when evaluating a homomorphism with codomain a `GrpLie` has been fixed. Bug reported by E. O’Brien. ## Linear Algebra - A bug where constructing random matrices over the finite field `GF(2^30` occasionally gave invalid entries has been fixed. ## Magma: Distributed Computation - A `RetainWorkers` parameter has been added to [`DistributedManager`](../MagmaLanguage/Parallelism/par-user.md#function-par-distributedmanager) to allow worker processes to be retained for a future call. - [`DistributedManager`](../MagmaLanguage/Parallelism/par-user.md#function-par-distributedmanager) now also allows the inputs to be specified with a `Process`. - A crash that happens sometimes when using verbose printing with [`DistributedManager`](../MagmaLanguage/Parallelism/par-user.md#function-par-distributedmanager) has been fixed. ## Magma: Intrinsics - The intrinsic `CreateProcess` has been added to simplify making processes. Simplified versions that work on sequences and lists have also been provided. ## Modular Forms - An occasional crash in computing the dimension of Hilbert cusp form spaces has been fixed. Reported by E. Costa. - A slowdown in computing Hecke operators of cusp form spaces has been fixed. ## Polar Spaces - The function `IsIsometric` has been extended so it can take user-defined maps. Issue reported by M. Rodgers. ## Representation Theory - The [`SaveCharacterTable`](../RepresentationTheory/CharactersOfFiniteGroups/irreducible.md#function-savecharactertable-seqenum-algchtrelt-monstgelt) command has been extended to also save the underlying group (when there is such a group). Reading the resulting file creates the group, its conjugacy classes, and the character ring containing this data as it sets up the saved character table. This improvement requested by D. Roe.