---
title: "Magma V2.24-9 Patch Notes"
canonical: "https://docs.magma-maths.org/patchnotes/2.24-9.html"
source: "https://docs.magma-maths.org/patchnotes/2.24-9.md"
---

<a id="magma-v2-24-9-patch-notes"></a>
# Magma V2.24-9 Patch Notes

## Algebras

- A problem with the map returned by `VectorSpace` for subalgebras of group algebras has been fixed.  Reported by S. Humphries.

## Conics

- An occasional problem with an uninitalized array with Lagrange’s method over number fields has been fixed.

## Elliptic Curves

- The [`FaltingsHeight`](../ArithmeticGeometry/EllipticCurvesOverQAndNumberFields/intro-2.md#function-faltingsheight-crvell-fldrat) in the case of curves with non-integral j-invariant has been corrected. It now agrees with `FaltingsHeight2` though neither is normalized as per Deligne. Also, the Handbook gives an incorrect formula this. Reported by A. Kresch.

## Fuchsian Groups

- A problem with parameters in [`FuchsianGroup`](../ModularArithmeticGeometry/ArithmeticFuchsianGroupsAndShimuraCurves/ArithmeticFuchsianGroups.md#function-fuchsiangroup-algquatord) has been fixed.

## Galois Groups

- Primes specified by the parameter `Prime` to `GaloisGroup` are now checked for being unramified. Reported by J. Voight.
- Some extra printing has been removed from `GaloisGroup` computations for reducible polynomials. Reported by D. Krumm.
- [`HilbertIrreducibilityCurves`](../GlobalFields/AlgebraicFunctionFields/galois.md#function-hilbertirreducibilitycurves-rngupolelt) will now return curves whose defining polynomials use the print names of the Parent of the input polynomial. Reported by D. Krumm.

## Graph Theory

- Tuples of the form \<vertex, set-of-vertices> are now properly handled by the network constructor.  Reported by W. Bosma.

## Groups

- A bug in the computation of the Frattini subgroup of a matrix group has been fixed. When the matrix group was a p-group, a proper subgroup of the Frattini subgroup may have been returned. Bug reported by Martin van Beek.
- The representation of ATLAS group 2J2d2 contained in GL(448, 7) has been corrected. Bug reported by E. O’Brien.

## Hyperelliptic Curves

- A bug was fixed in the Selmer machinery, which occasionally caused wrong results for `RankBounds` to be returned. A preconditioning attempt to use a better model in `MonicModel` was accidentally replacing the curve by a nonisomorphic one. Reported by J. Doyle.

## Lattice over Number Fields

- A problem with `Sphere` was fixed. Reported by M. Kirschmer.

## Modules

- A bug in `IsSurjective` for Hom-modules has been fixed.  Reported by D. Craven.

## Number Fields

- Factorization of polynomials over orders of number fields or their field of fractions has been fixed for large degree orders.
- Setting up of coercions between fields in `CompositeFields` has been improved to reduce failures of applying these coercions. Reported by J. Voight.
- The computation of `Valuation`s when the ideals are of an order not known to be maximal has been improved. Reported by L. Liu.
- A crash with the number field sieve (quadratic case) has been fixed, involving multiple divisors of the discriminant in the range around 1000 causing a buffer overflow. Example: 218467444968689005796305.

## Polynomial Rings and Groebner Bases

- The function `Degree` has been fixed to return -1 when a grading is present and the input polynomial is zero.  Reported by M. Zieve.
- Some bugs in local polynomial rings have been fixed. `Coordinates(I, f)` in the local polynomial situation now returns C and d so that `&+[C[i]*B[i]: i in [1 .. #B]] eq d*f` where `B := Basis(I)`.  Problems reported by G. Blanco.

## Series Rings

- Constructing a homomophism from an extension of a series ring previously did not allow a coefficient homomorphism to be provided. This is now allowed. Reported by J. Klüners.

## Sparse Matrices

- A crash in `Rank` for sparse matrices over GF(2) has been fixed. Reported by S. Goldstein.

## System

- Following a bug report by S. Humphries, a memory management problem has been fixed. This was to do with the ^^ constructor.