Vital statistics

Official seminar home page
Organizers: Ralph Johnson (Ralph), Shriram Krishnamurthi (Shriram or Shri), Thomas Kühne (Thomas), Michael Sperber (Mike) (write us collectively at dagstuhl-06181-org@deinprogramm.de)
Dates: May 1 (Monday) through May 6 (Saturday), 2006

If you have attended past Dagstuhl events, you are used to arriving on Sunday evening and having seminars finish on Friday at noon. Our schedule is off by one because Monday, May 1, is a bank holiday in Germany. Therefore, we are being asked to arrive on Monday and leave on Saturday. The seminar proper will begin first thing Tuesday morning, and end at noon on Saturday.

We discourage arriving late or leaving early, especially if you are a speaker: If others have to listen to your talk, you should have to listen to theirs, too.

What are latently-typed languages?

We use the term to refer to languages that recognize that values have types, but do not force unique types on identifiers, variables and, more generally, expressions. These languages enforce safety (i.e., operations are never mis-applied to values of the wrong type) but can usually only do so dynamically. Classical examples include Smalltalk and Scheme; more modern examples include Python, Ruby and JavaScript.

Why a seminar on latently-typed languages?

Look around. Despite the numerous (academic) successes of statically-typed languages, latently-typed languages not only exist, there seem to be more of them coming into existence all the time. Moreover, these languages have been disproportionately successful in mainstream and industrial use, e.g., for ``scripting''. Does this merely reflect the ignorance or laziness of programmers, or do these languages have some important characteristics that are worth deeper study? Instead of concluding the former by fiat, we'd like to study this question in greater depth.

What do you hope to get out of the seminar?

We are ambitious. We hope for a lot:

• Most of all, we want to introduce communities of people to each other to promote future communication and collaboration. This is especially important for the junior participants. They should forge better ties than has been done in the past!
• We want to find out what is new, exciting and deep in these different communities.
• We want to identify key research problems. Dually, we want to know about solutions one community has that another may exploit.
• We want to try to catalog insights into why latently-typed languages are successful—even situations where they are used mistakenly!
• Last but not least, we want to replace cynicism with hope. Sadly, some people still fight ancient battles or despair about missed opportunities. For sure, we do not want to paper over past mistakes, but we want only to learn from them, not dwell on them. We want to make whole new and interesting mistakes! We are attracting a vibrant, excited new cadre of graduate students; let's inspire them.

What was the seminar's genesis?

Stéphane Ducasse and Michele Lanza met Shriram Krishnamurthi for the first time at a Dagstuhl seminar co-organized by Ralph Johnson. At their very first conversation, they discovered that they had numerous common interests, and yet had not met each other before. (Since then, they have even more in common!) Shriram realized from talking with them, and with Ralph, what others have long known: that the Smalltalk and Scheme communities have long had a lot to teach other, and often admire each other from a distance, but for cultural reasons tend to publish in different venues (OOPSLA vs ICFP, etc.) and focus on different problems. At that seminar, they thought it would be neat to try to bring these two communities together. This seminar is the result of that initial contact.

What the heck is a ``Dagstuhl'', anyway?

Dagstuhl is a wonderful conference center in western Germany. It's located in a ``castle'', but that's a dangerous term to use because you might begin to imagine Neuschwanstein. Rather, think of Dagstuhl as a rather classy ``country house'', and you'll be impressed.

One of the important aspects of Dagstuhl is that it's in a rather remote location. You are expected to come and spend the week focusing primarily on your contact with other attendees, not on other worldly matters. Don't worry, though: entertainment is not a problem. Dagstuhl features a world-class computer science library, magazines and newspapers, a wine cellar, a good CD collection, musical instruments, pool, ping-pong (table tennis), bicycles, and, most of all, abundant wild fauna in the form of fellow attendees.

All meals for the week are served by the center, and their cost is included in the seminar price. (The center is vegetarian-friendly, but be sure to tell them in advance.) There are certain quirks to Dagstuhl: for instance, rooms cannot be locked from the outside, only the inside; and for lunch and dinner, seating is assigned by a mysterious algorithm. You can learn much more about the center from their pages.

When will you publish a talk schedule?

Dagstuhl seminars are not meant to be like typical computer science conferences; rather, they are an opportunity to learn and communicate without being shackled to traditional constraints. Therefore, even though your invitation letter asked you what you would like to speak about, disregard this. Ideally, a Dagstuhl event should publish a ``listen'' schedule rather than a ``talk'' schedule.

All this is by way of saying the organizers have a framework in place but are still working out the details. Don't worry: we'll contact you in due course. For now, rather than preparing slides, you should be preparing an open mind.

If we do ask you to speak, we'll be giving you some guidelines on the nature and content. Do keep this in mind: the single most boring and ineffectual thing you can do is to just repeat your latest conference talk. Your audience, almost by definition, consists of people who are ``culturally'' different from you; Dagstuhl approved of our petition to hold this event precisely to bring these cultures together. If you only speak to your own, you'll have wasted this golden opportunity of getting together with them. If, instead, you keep them in mind and give a talk that reaches out to them, you can influence their thinking. They may even start to follow your work. Heck, they might even cite you. Now do we have your attention?

Will you publish proceedings?

We don't know yet, one way or another. It will depend on participant interest. If we do, however, it will be with the goal of getting ideas in the open, not burnishing vitas, so we will only compile an informal document; Dagstuhl already has a facility for doing this, and encourages such publications. This will leave you free to publish your write-up formally later in a properly refereed venue.

Will we have an excursion?

Traditionally, the afternoon of the middle day of a Dagstuhl event (usually Wednesday, but for us Thursday) is free of talks. The organizers schedule something: anything as simple as a hike to as complex as an organized tour of a nearby city, cultural attraction, or vineyard.

We plan to not have an organized tour with a bus: this tends to be quite tiring for some attendees (since, invariably, the tour does not return until 10pm or so), and burdens graduate students (of whom we have a few) with additional costs. We are leaning towards something simple that takes us outdoors (the climate in May is usually very good for this): perhaps we can split into groups that hike and groups that bike. There is also a swimming pool in town not far away. Some people may also use this time to be anti-social, i.e., to lock themselves up in their room and accomplish some pressing task, with a view towards being social the rest of the week!

If you have suggestions or strong preferences, let us know.

Is there good network connectivity?

As Reinhard Wilhelm asks, ``Are you here for the network connectivity or for the social connectivity?'' That said, yes, there is excellent network connectivity, and there are even machines available, so you can keep up with events at home. There is also wireless connectivity though, mysteriously, the signal seems to not penetrate the main room for talks....

Travel to Dagstuhl is quite easy. For most people, it decomposes into three steps: (a) get to Frankfurt, (b) take a train, (c) take a bus. If you travel as a group, you will find (c') taking a taxi not much more expensive (and a time-saver). Simply reverse the steps to return. The Dagstuhl site provides good information on this. On weekends, the regional trains will let you travel cheaper [information in German, information in English].

While the bus and train used to synchronize years ago, the train schedule has crept by so much that they no longer do. You really do have to wait nearly an hour for the bus (or two, if you take the wrong train). So if you're impatient, you will want to take a taxi; otherwise, you may want to bring reading material (may we recommend papers by other participants at the seminar?).

Because you will be traveling to Dagstuhl on a public holiday, be aware that services will run less frequently, and be sure to check the right schedules. Effectively, you should expect travel to follow the Sunday schedule. The German rail system will present trains by date.

Suggestion for travelers from the USA: I (Shriram) prefer to arrive in Europe a day early to get over jet-lag. My Dagstuhl routine (for a traditional Dagstuhl event, which begins Sunday evening) is to fly from the US on a Friday, landing in Frankfurt on Saturday morning. The Frankfurt international airport (FRA) is roughly half-way between Frankfurt and the town of Mainz, with Mainz in the same direction as Dagstuhl and Frankfurt in the opposite direction. I therefore overnight in Mainz. Dagstuhl has a special rate arrangement with a simple hotel in Mainz right across the street from the main train station; the staff are very friendly and helpful (and fluent in English). Mainz is not a great culinary destination, unless you like traditional German fare, but it is historically and culturally interesting, and has an old town that drips with charm. It's also very easy to day-trip to Frankfurt, which has excellent restaurants, museums, bookstores and all the other attractions of a major city. On the way out, too, I again stay in Mainz, due to the easy access to trains, making it a touch easier to get to the airport.