Linkage

Good article, terrible headline (\(\mathbb{M}\)). Bill Gasarch rants about several recent instances of clickbaity, inaccurate, and overhyped media coverage of theoretical computer science topics. I suspect the answer to his question “is it just our field?” is no.

Vox on the sexist backlash against astronomer Katie Bouman (\(\mathbb{M}\), see also), of black hole image fame, after she was cast by the media in the “lone genius” role typically reserved for men and untypical of how science actually happens.

Mathematical sign language (\(\mathbb{M}\)). Hearing-impaired eletrical engineering researcher Jess Boland discovered that weren’t enough technical terms in British Sign Language to cover the mathematics she uses in her work, so she’s been creating new ones as well as promoting the ones BSL already had. Katie Steckles interviews her for The Aperiodical.

Regular polygon surfaces (\(\mathbb{M}\)). Ian Alevy answers Problem 72 of The Open Problems Project: every topological sphere made of regular pentagons can be constructed by gluing regular dodecahedra together. You can also glue dodecahedra to get higher-genus surfaces, but Alevy’s theorem doesn’t apply, so we don’t know whether all higher-genus regular-pentagon surfaces are formed that way.

Czech president Miloš Zeman “has repeatedly used presidential powers to block the professorships of political opponents” (\(\mathbb{M}\)). Charles University is now suing to allow their promotions to go through.

Planar point sets determine many pairwise crossing segments (\(\mathbb{M}\)). János Pach, Natan Rubin, and Gábor Tardos make significant progress on whether every
\(n\) points in the plane have a large matching where all edges cross each other. A 1994 paper by Paul Erdős and half a dozen others only managed to prove this for “large” meaning \(\Omega(\sqrt{n})\). The new paper proves a much stronger bound, \(\Omega(n/2^{O(\sqrt{\log n})})\) (Ryan Williams’ favorite function).

Why asymptotics matters (\(\mathbb{M}\), via): because if you don’t pay attention to it you get problems like this slow quadratic-time process creation bug in Windows 10.

Snap cube puzzle. The cubes have one peg and five holes; how many ways can you snap them into a connected \(2\times 2\) block with no pegs showing? See link in discussion thread for spoilers.

There’s lots of reasons to be unenthusiastic about newly-official-presidential-candidate Biden involving multiple instances of poor treatment of African-Americans and women, but here’s another more techy reason: his first major fundraiser as a candidate closely involves anti-net-neutrality lobbyists from Comcast (\(\mathbb{M}\)).

SAO/NASA Astrophysics Data System updates its user interface (\(\mathbb{M}\)). The ADS is a useful database of papers in astronomy and related fields. From comments on their post, the new UI is very slow. It is unusable without JavaScript. And it “sends the users’ personal identifying information to at least 5 third-party companies”. This is progress?

I ask for a reference for an easy fact about divisibility representations of partial orders (\(\mathbb{M}\)). The MathOverflow community isn’t very helpful, preferring instead to simultaneously complain that it’s too trivial and explain why it’s true to me as if I didn’t already say in my question that I thought it was trivial.`

Cannon-Thurston maps for veering triangulations, whatever those are. Henry Segerman posts some pretty pictures from his joint work with David Bachman and Saul Schleimer.

Newsletter of the IMU Committee for Women (via). Includes an interview with Marie-Francoise Roy and the announcement of the book World Women in Mathematics 2018.

Network coding yields lower bounds (\(\mathbb{M}\)). Lipton and Regan report on a new paper by Afshani, Freksen, Kamma, and Larsen on lower bounds for multiplication. If algorithmically opening and recombining network messages never improves fractional flow, then \(O(n\log n)\) circuit size for multiplication is optimal. But the same lower bound holds for simpler bit-shifting operations, so it’s not clear how it could extend from circuits to bignum algorithms.