On \((k,n)\)-arcs (Q557095): Difference between revisions

In this short note the author constructs new \((k,n)\)-arcs in \(PG(2,q)\), with \(k\) roughly \(q^2/d\) and \(n\) roughly \(q/d\), for any divisor \(d\) of \(q-1\). (A \((k,n)\)-arc is a set of \(k\) points of \(PG(2,q)\) whose secants contain at most \(n\) points, and some secant contains exactly \(n\) points.) To do this, the author uses algebraic curves and the Weil bound. The construction goes roughly as follows: one considers a suitable homogeneous polynomial of degree \(d\) and looks at the set of points where the value is in a fixed coset of the \(d\)th powers of \(GF(q)^\times\) (in fact, one chooses the coset in such a way that this set has maximal cardinality), and removes the lines contained in it. This is a rather ingenious construction, and a lot of questions arise, some of which also stated by the author in the paper under review.