Bielliptic curves of genus 3 in the hyperelliptic moduli (Q386413): Difference between revisions

The authors study the moduli space of hyperelliptic genus \(3\) curves which are \(2\)-covers of an elliptic curve. Their purpose is to give equations for this space and models for the curves in terms of dihedral invariants [\textit{J. Gutierrez} and \textit{T. Shaska}, LMS J. Comput. Math. 8, 102--115 (2005; Zbl 1110.14025)] and also Shioda invariants [\textit{T. Shioda}, Am. J. Math. 89, 1022--1046 (1967; Zbl 0188.53304)]. This space is actually the moduli space of hyperelliptic genus \(3\) curves with automorphism group isomorphic to \((\mathbb{Z}/2\mathbb{Z})^2\) and has then been studied in previous articles by the first author. Hence a large part of the article is a recollection of results which have already appeared in a similar form. Reviewer's remarks: Unfortunately, the article is stained with mathematical mistakes and misleading references to the literature. Although it is not the purpose of a review to do this job, we want to list some of the most important ones below. p.~389: In [\textit{T. Shaska}, Commun. Algebra 42, No. 9, 4110--4130 (2014; Zbl 1427.14056)], the first author computes the invariants for binary octavics and their relations. Let us point out (see the footnote p.~603 of [\textit{R. Lercier} and \textit{C. Ritzenthaler}, J. Algebra 372, 595--636 (2012; Zbl 1276.14088)]) that Shioda's wonderful work actually only contains two minor typos. p.~390: In this article, the authors deal with the descent of a hyperelliptic genus \(3\) curve with automorphism group isomorphic to \((\mathbb{Z}/2\mathbb{Z})^2\) over a (at most) quadratic extension of its field of moduli. This result was already obtained in the article [\textit{R.~Lercier, C.~Ritzenthaler} and \textit{J.~Sijsling}, ``Fast computation of isomorphisms of hyperelliptic curves and explicit Galois descent'', in: \textit{E. W. Howe} and \textit{K. S. Kedlaya} (eds.), ANTS X. Proceedings of the tenth algorithmic number theory symposium. Berkeley, CA: Mathematical Sciences Publishers (MSP). 463--486 (2013)]. Theorem 3: The equation \(F_1\) is not defined in [\textit{T. Shaska}, Commun. Algebra 42, No. 9, 4110--4130 (2014; Zbl 1427.14056)] and the others do not correspond to the equations of the locus studied in the article (they are simply the expressions of Shioda invariants \(J_i\) in terms of coefficients of a general form). p.~402: The conjecture made by the first author in 2003 about the field of moduli being a field of definition for hyperelliptic curves with automorphism group of order greater than \(2\) has been proved wrong by \textit{B. Huggins} [Math. Res. Lett. 14, No. 2, 249--262 (2007; Zbl 1126.14036)] even for hyperelliptic curves of genus \(3\) with automorphism group isomorphic to \((\mathbb{Z}/2\mathbb{Z})^2\). Still, the authors pretend to show that the field of moduli is a field of definition in the present case (but finally no proof is given). Lemma 6: The authors claim that the given model of the curve is defined over its field of moduli. Such a claim was already made in Lemma~10~iv) of [\textit{J. Gutierrez} et al., Lecture Notes Series on Computing 13, 109--123 (2005; Zbl 1121.14021)]. With an explicit counterexample, it was shown in Remark 4.17 of [\textit{R. Lercier} and \textit{C. Ritzenthaler}, J. Algebra 372, 595--636 (2012; Zbl 1276.14088)] that this claim is wrong. Although the equation is a bit different in the present article, one can check that the counterexample still works.