Funny you should mention that. We communicate between the PC104 flight computer and the PIC microcontrollers using a CAN bus. This is the same technology used to communicate between the microcontrollers in modern automobiles. There is a Linux driver for this bus, but we are finding that it is of dubious quality (can't use it O_RDWR, for instance).
> I mean, what happened? Which teams did what? Was anything interesting learned from this? Give us an overall ranking of all entries.
The detailed results are now posted.
I have a problem with the judges' scoring methodology. They simply totalled points across all the trials. However, different maps had different maximum amounts of points available, so that the results on a few maps dominated the results on the rest of the maps.
It would have been more fair to use, say, percentages of the maximum score acheived on a particular map, so that the maps were weighted equally.
> Such is life and the imperative languages have still yet to see a victory in ICFP.
Well, in the second year of the contest, the task was an abstract game player on parallel hardware. The winner was written in CILK, a C language with built-in parallelism primitives. The team had previously written a state-of-the-art parallel chess program. This demonstrated that domain expertise and language familiarity trumps any particular language paradigm.
Mod this up. This is the one out of a hundred posts that has the facts right.
Although it was quite an achievement in Deep Blue to implement a tuned, parallel chess evaluation function and 4-ply search in silicon, it would be an even greater achievement to understand Kasparov well enough to construct an opponent model of him for use in a chess program. It is notoriously hard to tune a chess program for a particular style, much less to play well against the style of a particular opponent. Much of the observable style of a chess program is emergent behaviour based on subtle differences in hundreds of coefficients down in the evaluation function.
I do believe IBM was unfair to the chess community in mothballing the system after its final match. IBM only saw Deep Blue as a successful marketing promotion, which could only be devalued by risking losses in further competition. Quite a far cry from all preceding chess program research, in which the researchers were eager to test their programs' mettle. As far as I know, Deep Blue has only played one private match with another computer (a version of Genius), although there are many annual computer chess competitions.
Ambrosia won't make EV multiplayer, but there are some smaller scale alternatives:
Matt Burch wrote a networked 3D space dogfight engine called Exobattle, and released it as open source!
Tycho is writing a two player networked arena game called Helian Wars.
This topic comes up frequently in the EV Developer's web board on the Ambrosia web site.
Funny, but true! Why do you think Deep Blue was dismantled after beating Kasparov? Because to IBM, there was nothing more to gain, and everything to risk. The same goes for most publishers of commercial computer chess programs. They only compete if they have a better-than-average chance to get that magic sticker on their shrink-wrapped box.
In this particular instance (CCT4), the tournament doesn't yet have the prestige of the existing computer tournament cycles. Many of the commercials aren't bothering to compete for that reason (nothing to gain, risk to lose).
Ian
Now that computers are starting to plateau off in strength, more game results are coming down to the quality of the opening book (just like in real Grandmaster tournaments). Here the commercial programs have an edge, with a person paid to tailor an opening book which is 1) suited to the style of the program, and 2) lays traps for all the other chess programs!
Yes, they do tailor books for computer competitions, and there have been numerous upsets due to a stronger program falling into the opening trap of a weaker program. For instance, many of the programs which play on the internet chess servers (e.g. crafty) normally use a very wide book to keep their games entertaining. They switch to narrower less risky books for tournaments.
There are a number of efforts at automating the process of creating an opening book, but the results have been mediocre.
Also, getting out of book is not the weapon it used to be! Programs are strong enough now to improvise well even in the opening. (This was one of the things which killed Kasparov in his last match with Deep Blue.)
However, outside of their books full of rote opening moves, most programs don't have opening specific knowledge, such as knowing about certain thematic sacrifices and plans in particular openings. For example, a computer out of its book in the Sicilian Dragon will not make the thematic queenside RxN sacrifice.
Another problem with opening preparation against computers, compared with humans, is that they tend to play a wider range of openings. You might prepare a trap in a particular opening line, yet only have a 0.1% chance of getting to play it.
Slashdot Mirror
Funny you should mention that. We communicate between the PC104 flight computer and the PIC microcontrollers using a CAN bus. This is the same technology used to communicate between the microcontrollers in modern automobiles. There is a Linux driver for this bus, but we are finding that it is of dubious quality (can't use it O_RDWR, for instance).
> I mean, what happened? Which teams did what? Was anything interesting learned from this? Give us an overall ranking of all entries.
The detailed results are now posted.
I have a problem with the judges' scoring methodology. They simply totalled points across all the trials. However, different maps had different maximum amounts of points available, so that the results on a few maps dominated the results on the rest of the maps.
It would have been more fair to use, say, percentages of the maximum score acheived on a particular map, so that the maps were weighted equally.
Ian
> Such is life and the imperative languages have still yet to see a victory in ICFP.
Well, in the second year of the contest, the task was an abstract game player on parallel hardware. The winner was written in CILK, a C language with built-in parallelism primitives. The team had previously written a state-of-the-art parallel chess program. This demonstrated that domain expertise and language familiarity trumps any particular language paradigm.
Ian
Mod this up. This is the one out of a hundred posts that has the facts right.
Although it was quite an achievement in Deep Blue to implement a tuned, parallel chess evaluation function and 4-ply search in silicon, it would be an even greater achievement to understand Kasparov well enough to construct an opponent model of him for use in a chess program. It is notoriously hard to tune a chess program for a particular style, much less to play well against the style of a particular opponent. Much of the observable style of a chess program is emergent behaviour based on subtle differences in hundreds of coefficients down in the evaluation function.
I do believe IBM was unfair to the chess community in mothballing the system after its final match. IBM only saw Deep Blue as a successful marketing promotion, which could only be devalued by risking losses in further competition. Quite a far cry from all preceding chess program research, in which the researchers were eager to test their programs' mettle. As far as I know, Deep Blue has only played one private match with another computer (a version of Genius), although there are many annual computer chess competitions.
Ambrosia won't make EV multiplayer, but there are some smaller scale alternatives:
Matt Burch wrote a networked 3D space dogfight engine called Exobattle, and released it as open source!
Tycho is writing a two player networked arena game called Helian Wars.
This topic comes up frequently in the EV Developer's web board on the Ambrosia web site.
Comic artist Scott McCloud, author of Reinventing Comics, has voiced his own opinions on this issue in a series of online comics.
Part 1
Part 2
Funny, but true! Why do you think Deep Blue was dismantled after beating Kasparov? Because to IBM, there was nothing more to gain, and everything to risk. The same goes for most publishers of commercial computer chess programs. They only compete if they have a better-than-average chance to get that magic sticker on their shrink-wrapped box. In this particular instance (CCT4), the tournament doesn't yet have the prestige of the existing computer tournament cycles. Many of the commercials aren't bothering to compete for that reason (nothing to gain, risk to lose). Ian
Good question.
Now that computers are starting to plateau off in strength, more game results are coming down to the quality of the opening book (just like in real Grandmaster tournaments). Here the commercial programs have an edge, with a person paid to tailor an opening book which is 1) suited to the style of the program, and 2) lays traps for all the other chess programs!
Yes, they do tailor books for computer competitions, and there have been numerous upsets due to a stronger program falling into the opening trap of a weaker program. For instance, many of the programs which play on the internet chess servers (e.g. crafty) normally use a very wide book to keep their games entertaining. They switch to narrower less risky books for tournaments.
There are a number of efforts at automating the process of creating an opening book, but the results have been mediocre.
Also, getting out of book is not the weapon it used to be! Programs are strong enough now to improvise well even in the opening. (This was one of the things which killed Kasparov in his last match with Deep Blue.)
However, outside of their books full of rote opening moves, most programs don't have opening specific knowledge, such as knowing about certain thematic sacrifices and plans in particular openings. For example, a computer out of its book in the Sicilian Dragon will not make the thematic queenside RxN sacrifice.
Another problem with opening preparation against computers, compared with humans, is that they tend to play a wider range of openings. You might prepare a trap in a particular opening line, yet only have a 0.1% chance of getting to play it.
Ian