I spent a few nights in lovely Paris last week, attending the Journées Nationales d’Informatique Mathématique at Paris Diderot university.

Many computer science fields in France are organised in national, cross-institutional groupes de recherches, not completely unlike the ACM SIGs in the US. The TCS group (“Mathematical Informatics”) encompasses some 400-500 people, as far as I understand.

As in most of Europe, theoretical computer science in France includes many more fields more than US-style theory of computation. Amazingly, they meet once a year for two days, and give well-attended talks to each other. The 2012 meeting had 170 registrants, an impressive number.

This struck me as particularly noteworthy after just attending SODA, where the various subfields of algorithms become increasingly fragmented and estranged, to the point of hostility and mutual incomprehensibility.

At the Paris meeting, a steering committee selects a number speakers from the various working groups in the GDR IM, who give meaty, 45-minutes talks to a general TCS audience: Computational logic, computational geometry, distributed systems, process calculi, extremal graphs, ….

In addition, the meeting includes two 1-hour invited talks recruited outside of the French GdR. Ashwin Nayak talked about communication complexity, and I used to opportunity to present an overview of zeta transform algorithms and applications, culminating in our SODA 2012 result from last week. [slides]

Thanks to everybody who attended, and to the nice organisers for putting me into a disarmingly charming hotel in the middle of the Latin quarter, where you couldn’t swing a dead Marsipulami without hitting a comic store. I had a splendid time.