The 11th International Conference on Concurrency Theory was held at University Park, State College, Pennsylvania, U.S.A from the 22nd to the 26th of August 2000.
State college is a small university town in the middle of nowhere. The location for the university was chosen, I am told, to be the geographical centre of Pennsylvania. That means, it is at least 3 hours drive from Philadelphia, New York, Pittsburgh, or for that matter from any place you can think of. Very thoughtfully, Pennsylvania has also built a large state penitentiary close by.
The University is well known for its football team, the Nittany Lions. Nittany by the way is the name of a hill near State College. Nittany is also the name of the hotel I stayed in, the name of every second restaurant in town, the name of all the barber shops in town, the name of every second street in town and so on. I hope you get the idea.
Inspite of the rural setting, the university town itself has a lot of life. There are a lot of bars and restaurant and night clubs, most of them (of course with names containing the substring ``Nittany'') full of undergraduate (underage?) students. Michiel Von Rosch, a Dutch student visiting Stony Brook, found much to his disgust that he cannot quench his thirst without an ID card verifying his age. He alleges that most bars never checked the IDs of the women who went in. But then I have no complaints, as I look a healthy 30 and was never ``carded'', some of the bars here do have excellent collection of beers from all over the world.
The conference was very well organised. Catuscia Palamidessi and Dale Miller had done a great job in organising the conference and the associated events (including an excursion involving cave exploration, a banquet and a reception.)
What about the business part of the conference? Well, it is about 3 months since I attended the conference (and strangely it all seems very hazy now.). So, let me issue a statutory warning: the talks may not have been as good or as bad as they are painted here.
First, some general philosophy. I think that concurrency as an area has existed rather respectably for decades before model-checking and verification became the rage. There are many forums where the emphasis is verification, for instance CAV and TACAS. But looking at the list of invited talks and tutorials at this years CONCUR one wonders if CONCUR is slowly turning itself into another verification conference. I shudder.
There were 4 invited talks. Natarajan Shankar gave the first invited talk. He spoke about the integration of model-checking and theorem-proving. He argued, quite effectively, that it is possible to combine them in a manner where both of them play a nontrivial role. Roughly speaking he suggests the use of model-checking to gather global information and deduction to propagate this information locally. His write-up in the conference proceedings is quite lucidly written.
Ed Brinksma gave a talk titled ``Verification is Experimentation''. Some of you might remember a talk by Juris Hartmanis at a FSTTCS in Bombay. Well, need I say more?
Gene Stark spoke on his work on analysing Probabilistic I/O automata. The talk did convey that he has developed a powerful machinery based on linear algebra and the theory of Markov processes and used it effectively om analysing these automata. But then the talk was sketchy (and it had to be, given the depth of the subject and the time he got to present his work) and so it was difficult to understand any of the details.
Alberto Sangiovanni-Vincentelli (oh, I love this name. If everyone had names like this it would have been so easy to write my one page report) gave a rather animated talk on the design of embedded hybrid controllers (in particular for automobiles) He is definitely a good lecturer, in the style popularised by sunay morning preachers, and though the topic was a bit light-weight I did enjoy the talk. His point, that if we want to do ``useful'' research we must first talk to the end-user and identify the problem (which he says is the hardest part!) and then without any idealogical baggage attempt to solve it to the satisfaction of that user, is obviously correct. Which of the formalisms are useful and which are not will be determined in this process and not apriori. So simple yet so rarely followed. Ugo Montanari, introduced Alberto as one who has been part of many start ups worth about 1 million dollars. Clearly theoreticians are clueless. Alberto started his talk with ``Ugo, not 1 million. 1 billion, 1 billion.'' Add the classic Italian accent and you know how it sounded.
The conference also had 4 invited tutorials. I attended two of them. Rajeev Alur presented a tutorial titled ``Exploiting Hierarchical Structure for Efficient Formal Verification''. May be it was the sumptuous lunch or may be it was the late evening, but I did not quite catch the point there. C. R. Ramakrishnan gave tutorial on the use of logic programming in model checking. C.R. also gave a talk at one of the workshops following the conference (It was called MTCS, T standing for ``timed'' I think) and without a doubt that talk got the best slides award. It was the closest to animation in slides I have seen. Well, talking of slides, P.S.Thiagarajan gave the first contributed talk in the conference, with handwritten transparencies. Inspite of the transparencies, it was a very nice talk and I enjoyed it thoroughly (hope I got my brownie points here.)
I thought it was going to be the only ``handwritten'' talk in this day and age, but low and behold, some of the other speakers not only had handwritten transparencies but also had prepared them with (almost) invisible ink. (For the dense, I prepared my transparencies on slidetex this time around.) There were two other invited tutorials that were co-scheduled with these two and so I did not attend them.
There were not many memorable contributed talks. I liked the talk by Kim Larsen titled ``The Impressive Power of Stop Watches''. He showed that timed automata with stop watches have the power of linear hybrid automata. Tevfik Bultan presented decidability results for reachability and emptiness for several extensions of reversal bounded counter automata. James Liefer, a student of Robin Milner, presented what seemed like a very nice talk. I say ``what seemed'' because I understood very little. Purush Iyer gave a rather entertaining talk titled ``Reasoning about Probabilistic Lossy Channels''. He showed that determining ``reachability with probability 1'' is undecidable for probabilistic lossy channel systems. There were a couple of other talks on probabilistic systems: Anna Philipou presented a decidable weak bisimulation for probabilistic systems and Roberto Segala gave a fairly bleak and un-understandable talk. Thiagarajan mentioned that the talk by Remy Morin titled ``Pomsets for Local Trace Languages'' was interesting (I had to miss this talk as it was presented in one of the parallel sessions.)
I attended a lot more talks than the just ones listed above, but then ...In any case, I would have attended a lot more but for the fact that I had carried a copy of ``Interpreter of Maladies''. That is a great collection of short stories and if there is one thing I can recommend after coming back from CONCUR 2000, that is without a doubt Jhumpa Lahiri.