# More matroids in Maastricht

I’m just back from a week in Maastricht, at the * Third Workshop on Graphs and Matroids*, which is the third (and sadly last) of a series of biennial workshops all of which have been organised essentially single-handedly by Bert Gerards. If you’ve ever tried organising a conference, you’ll know that running a 70 person conference and introducing every one of the 57 talks is not something most people undertake on their own.

The conference attracted many, if not most, of the “*household names*” in graph and matroid structure theory and included several of my existing and hopefully future collaborators, so despite the distance from Perth, it’s one of the highest priority events in my conference calendar. And given that I hate the tediousness, discomfort and inconvenience of flying long-distance in economy that’s saying something.

There were six longer talks, given by Paul Seymour, Paul Wollan, Carsten Thomassen, Jim Geelen, Jim Geelen and Jim Geelen, and 51 shorter talks so, with no parallel sessions, it made for a packed and tiring program. Jim got three spots because he was giving an overview of the Matroid Minor Structure Theorem which is a huge multi-decade project with Geoff Whittle and Bert Gerards with the ultimate goal of giving a structure theorem for minor-closed classes of matroids representable over a finite field, analogous to the Graph Minor Structure Theorem of Robertson and Seymour.

Overall, Jim did an excellent job of giving a sweeping overview of the project, what the structure theorem will look like, the results proved so far and their consequences, and what needs to happen next. I particularly appreciated the effort he made to give just enough definitions and detail for everyone in the audience to to get a sound intuitive notion of the theorem, while shielding us from the bucketloads of incredibly complicated and highly technical details that underly the proof.

Unfortunately, quite a number of the other talks did not treat the audience so tenderly, and were delivered in the more traditional *“take no prisoners”* style in which the speaker launches into 20+ detailed slides covering their latest proof. Slides densely packed with definitions – not leavened by illustrative examples, context or motivation – and with each new definition building on one or more of the previous ones leaves me lost after the first few minutes.

If I have a choice of seeing 15 minutes explanation of * what* the problem is and

*it is interesting (even just why the*

**why***speaker*finds it interesting) and a 5-minute sketch of

*they are doing it, versus a full 25 minutes of the gory details of a proof, then I’ll always choose the former. But probably I’m just slow, or grumpy, or most likely both.*

**how**The session in which I spoke (the “*graveyard shift*” on late Thursday afternoon) had Dillon Mayhew, Charles Semple and Steve Noble all giving the sort of talks that I like. Obviously, I tried to give such a talk myself, and it seemed to go down well, but you can judge for yourselves as the slides are here (Maastricht2012). (Note that the PDF contains some peculiar horizontally squashed pictures. I noticed that the projector was messing up the aspect ratio and stretching everything out horizontally and as nobody could fix it, I had to pre-squish the photos in the PDF so that it would appear correctly after being stretched.)

Of course, the talks are only a modest part of why anyone goes to such conferences, with the most important part being the interaction with everyone else there. Catching up with old collaborators and friends, meeting new people, finding out who has students starting or finishing, making plans and nutting things out on the spot. If all goes according to plan, we should have a small paper on binary matroids in a few weeks, as some of the final pieces of the puzzle we were working on fell into place during one such interaction. On top of that, there were more discussions about `sage_matroids`

, which is a computer package under development with the ultimate goal of implementing matroid functionality into Sage (www.sagemath.org). Just as John’s fining package (FinIng) adds special-purpose finite geometry functionality to GAP, we hope to add some effective, extendable and programmable matroid tools to Sage. This project is really being driven by Stefan van Zwam and Rudi Pendavingh, who have already done an immense amount of work implementing a substantial number of things, but lots remains (docstrings, anyone) before it can be submitted to Sage, and much tutorial material needs to be prepared if we hope it to be used by hesitant programmers.

Maastricht itself is a gorgeous town in which to have a conference. Known primarily for European treaties, the famous “bookstore in a church” (Google *Selexyz Dominicanen*), and Andre Rieu (the soft pop-classical violinist made wealthy by his carefully targeted *“mums and grans” *demographic), it is actually a picture-postcard European town.

Narrow cobbled streets, old churches, bridges over the Maas and picturesque squares with shady umbrellas over the tables of the numerous welcoming restaurants. In fact, the ** incredibly** numerous restaurants! Maastricht seems to have far more restaurants, and far too many mid-to-high end shops, and to be far more crowded in the town centre than is reasonable for a town of its population. So we could only guess that, along with being a regional centre, it must be visited by a huge number of tourists. But as there’s not actually all that much to do other than appreciate the beauty and the architecture and to eat and drink in the many restaurants, most of the tourists seemed to be local European tourists of the staider kind, rather than the sun-sand-and-sex seeking lager lout variety.

Like so many European towns, it is wonderfully pedestrian- and cycle-friendly and pretty much everyone cycles or scooters everywhere. I just love seeing a stylishly attired elderly matriarch sail serenely past on the heavy upright city bikes that give their riders such wonderful posture. (Mind you, I rented a bike and did a 25km ride and by the end of it, riding back into a stiff headwind, I wasn’t so happy seeing the elderly matriarchs rocketing past, leaving me labouring in their wake.)

Almost everyone stayed in the same hotel as last time, the rather eccentric Hotel Mabi which is right near the centre of town (conveniently located adjacent to a few “*head smartshops*” for those so inclined). A converted old cinema, it is decked out in cinematic paraphernalia and somehow conveys a sort of plush red velvet feel with a modern touch. Included in the rate is Limburg pie and coffee each afternoon, and an hour-and-a-half of wine and olives in the salon in the evenings. This has the pleasant side-effect that everyone starts the evening in the salon, has a drink or two and a chat, and then breaks up into groups heading for dinner. It’s like having a reception every evening and vastly increases the amount of informal interaction and mixing at the conference. I was also intrigued by the Dutch honesty – when wine time was over, the receptionist came to the salon and pulled a curtain over the three wine fridges and that was all that was needed. And the curtain seemed to be sacrosanct – the three unlocked wine fridges, each holding maybe 12 bottles, seemed to remain untouched until the next day. Yet on the other hand, when I rented a bike, I was given about three locks and basically told that anything not chained down in multiple ways to an immovable object would certainly vanish. In Australia, I think I’d rather take my chances leaving a bike unlocked than 30 bottles of wine behind a curtain!

I like to try as many local things as I can when overseas, and so asked Stefan, Bert and Rudi about Dutch specialities. One of these is licorice, which comes in an enormous range of different shapes and either hard or soft, sweet or salty. To get the most authentic Dutch experience, I tried the hard “*dubble zout*” which is hard and doubly extra-salty. All I can say is that Dutch kids must be pretty tough if they grow up where something with approximately the taste and texture of a leather boot soaked in brine is viewed as a treat. Still, Rudi said that it was often used medicinally for sore throats, and sure enough the trace of sore throat that I woke up with one day vanished after assiduous application of plenty of “dubble zout”, and after that it sort of grew on me. The next delicacy on the list was “*maatjes”* which are a raw salted herring, about 10 cm long, which are traditionally eaten by holding it by the tail, and lowering it into your upturned mouth. Geoff, Steve and I had one each (2 euro per herring) it was quite nice, but surprisingly rich and substantial and one was enough.

Not quite a Dutch highlight, but certainly a Maastricht highlight, is a visit to the famous Take One beercafe, which is owned by a rather eccentric beer fanatic. He selects the right beer for you after asking a few questions about what style of beer you fancy. If you answer sensibly regarding hops, malt, fruitiness etc, then all is well and a suitable beer appears, but if you answer foolishly, with something like “a dark beer” then you get a rebuke along the lines of “if you want dark, then you need a paint store – dark is not a taste”. Along with the beer, a bowl of raw peanuts is provided with the strict instruction that shells are not to be left on the table or in the ashtray but must be thrown or dropped onto the floor.

So, all in all, an excellent week. But I can see why Bert has called it a day – the conference has grown so much that it must take a huge amount of work, and if it were to run again, some hard decisions would need to taken about how to make it manageable. Better to finish on a high.

This means that we’ve got two years to convince Stefan that Princeton is the ideal spot for the next “*Maastricht conference*” – let the lobbying begin!

Oh for the love of God make it Princeton. 🙂

I just got far enough down my RSS feed to see this post… You’ll have a bit of convincing to do, because I rather like the Netherlands as conference location, for obvious reasons 😉

Now, back to sage.matroids…