The game of Go has proven to be incredibly hard to program, and is a much better indication of where artificial intelligence is today than the game of Chess.
This article gives an introduction to the problems involved in getting computers to play Go:
He thinks 300 page game designs stifles creativity, but how does he suggest you get 20+ people to work together to produce a coherent result without a detailed design? Idiot.
Furthermore he thinks that it stifles creativity to make a game on a budget and with a timeschedule... Well, of course it does but unless you have an operating system monopoly paying your bills this situation is not going to change anytime soon. Idiot.
And his suggestion to use focus groups in game development -- I've tried that, but the average joe is not a game designer and does not produce valuable feedback apart from "I like it" or "I don't like it". One example of why this fails is a game I worked on that had a bug in the collision response in one of the focus group builds -- it was slated by them, but _none_ of them mentioned the collision response being bad. Idiot.
And I don't know why anyone bothers listening to him. I've been making games for 10 years now and I have seen quite a examples of how not to do development...
Saying that a 300 page game design stifles creativity is completely wrong, unless for some reason your publisher is requiring you to stick to the letter of it instead of being flexible. How do you think you get a team of 20+ people to produce a coherent game? Normally you can't see which parts of the game was made by which artist and so on, why? BECAUSE THERE IS A GAME DESIGN DETAILING HOW THINGS SHOULD LOOK AND WORK. Of course if something is not fun, you come up with a new design for that part and update the game design document accordingly.
He also seems to think that everyone can do business like Microsoft where it does not matter how much money you lose because there is always the operating system monopoly there to feed you... Saying that developers make bad games because they have to make them on a budget and a timeschedule is of course true, but not very interesting as this is likely to continue to be the case for the foreseable future...
... And on the matter of Focus-group driven design, well I've tried it and I can tell you that there is a reason you don't hire 13 year olds to write your game design. You can see if they like the game or not, but if they don't like it they will not be able to put their finger on _why_ they don't like it -- there's always something to critizise, some dodgy texture or some little glitch. One game I worked on had a bug in the collision response code used for one of the focus groups and they came back with lots of critisisms, but all of them missed the real reason the game was crap. When this was fixed, they came back with positive comments instead...
What do _you_ think makes a game interesting? The packaging? The scantily clad women? The fact that you get to be james bond? Please...
Games are about user interaction, that is taking the controller and using it to move something around on the screen. There are far too many game developers who forget the fundamentals. Actually it is a mark of a true professional that they focus on the fundamentals, this is why Shigeru Miyamoto has developed an unprecedented string of hit titles and is respected by almost everyone in the industry...
There are two interesting points in the reasoning for the verdict, the first is that he was accused of being an accessory to theft (piracy) by making DeCSS available. The court applied a law that says you cannot be an accessory to a crime if you sell or give away something which can both be used for legal or illegal purposes (like a knife). The court deemed that DeCSS was a useful tool that could be used legally for people who wanted to play their DVDs under Linux, so Jon could not be convicted of being an accessory to theft.
The second is whether reverse-enginering is considered breaking copy-protection. The court ruled that translating source code into binary form did not in itself count as protecting it, because there are other good reasons for distributing binaries (although I'm sure some OSS advocates think differently;-) besides protecting the source code. Thus reverse engineering the source code from the object code could not be considered breaking any protection.
I have rarely seen Slashdotters agree so violently on anything, it really pleases me to see so many adult gamers who speak up. I just want to add that
there are games that have 'grown up', a prime example being Rez, a game which is as much a work of art as a traditional videogame. Some designers are pushing the envelope, making games that would not appeal to children, and that's a good thing.
Just for the record, everyone in the games industry (including myself) should remember that they owe their living to Miyamoto-sensei. I'd love to see Samus Blackly strapped to a chair forced to play BMX-XXX for eternity...
You're missing the point, there are no free players out there that offers the performance of Bink, so using something else means upping the system requirements for the game and reducing the amount of possible buyers. This is why games companies pay good money for Bink, it lets you play fullscreen movies with no hickups on a low-end system, thus giving you more customers and in the end it makes you money. You don't spend money on such a system for fun, or to be locked in to one specific vendor / platform or to give your money to some greedy company who sells solutions that are inferior to free alternatives.
The only good alternative is to port Bink to Linux, and I hope that is what they end up doing.
The Miles sound system is implemented in a very portable way, it is available for all major Consoles, Windows and Mac. We use it for our game "IGI2: Covert Strike", and I have to say it is very good (Much better than using DirectSound which is the main alternative on the Windows platform). Since I know that the Miles license includes access to the source code (I built Miles myself in search of a bug, and submitted a patch for it), so why not port it to Linux? I don't think it would be very hard at all (and in contrast to 99.9% of the other posters, I have seen the source), and it would also make bringing any other game using Miles to Linux that much easier. Miles is used in hundreds of games, as it is the fastest, most reliable, tried and tested sound system out there... AFAIK, all EA's games use it. Bringing it to Linux would be the right way to go, and would bring Linux gaming one step closer to a reality.
RAD Game Tools, the company who sells Miles is the same company that sells Bink, but since I have never used Bink, I don't know how easy it would be to port it to Linux. But as a general rule, it would be better to use effort to port APIs that many developers use rather than porting specific applications like NNN... Just my.02 Euros...
All the programmers at my workplace have been turned on to sushi lately -- finally, food you can order that doesn't exasperate that ulcer that any programmer who has gone through a few product cycles in the games industry is bound to develop...
Since only jokes have been posted so far, I'll try something else -- a real recipe that anyone can cook, in less than 15 minutes with very good results... Pasta with fresh tomatoes:
You need per person:
200g Spagetti or other kind of pasta
2-3 fresh ripe red tomatoes
Start boiling water (as this is the the task that takes longest to complete), put some salt in the water (a tablespoon).
Split the tomatoes open, get rid of all the watery goo and seeds so only the firm flesh of the tomatoes remain. Chop them as finely as you can be bothered to.
Put some oil in a frying pan, put the pan on the heat and finely chop the garlic. Put the garlic in the pan, and then before it turns brown (that is after 10 seconds) add the tomatoes.
When the water boils, add the pasta. If you don't turn down the heat very much, so the water continues to boil violently you don't have to stir very much;-)
Keep stirring the tomatoes around until they turn soft and start looking a bit like sauce. Check if the pasta is finished by getting a strand out of the boiling water once in a while and biting it. If it doesn't have a hard core, it's finished.
Get rid of the water for the pasta. Add tomatoes to pasta, add some extra virgin olive oil and stir. Put on plates, sprinkle with grated chese and some leaves of basil if available...
There you go, one of my favourite easy dishes...
If all you have is a hammer, every problem looks..
on
A New Kind of Science
·
· Score: 1
I've had the chance to look at this book, and I think it is very much a product of our time. Like the mathematical/mechanical worldview of the 19th and 20th century, this book sees the world through the eyes of a computer scientist. You could say it is introducing an algorithmic worldview.
However, I think it's a real problem that he is not presenting any results. At the end of the day, any scientific method has to be judged by the results it produces, and all Wolfram does is to re-cast familiar results in terms of cellular automa.
So to sum it up -- if you're dying to learn about cellular automa, read the book. If you're looking for a revolution, I'm afraid you'll have to start it yourself...
Why is no1 talking about the real reason for 9/11?
on
Globalism Post 9/11
·
· Score: 1
The terrorists attacked the US because they are funding Israel's attacks on the arabic minority. Is that so hard to see? You're so fucking myopic thinking that you were attacked because the rest of the world is so envious of your high standard of living or whatever...
The US is giving Israel billions of dollars worth of arms every year that they use to oppress arabs, all the time pretending to try to make peace in the middle east. How can you be a peacebroker when you're supplying the arms to one side??
I didn't say 'programmers make bad managers' -- In fact the title was 'Programmers _can_ be managers from hell'. My opinion is that being a good manager has very little to do with programming and everything to do with management skills -- making each team member perform to their optimum, and getting the information you need to make good decisions and so on.
Some people have a natural talent for this, and they can become good managers without education -- but in most cases I would prefer to get someone with some credentials (other than being a former programmer)...
Programmers can be managers from hell.
on
Do You Like Your Job?
·
· Score: 5, Interesting
My previous manager was once a programmer (a very bad programmer, I guess). This lead to him
being totally lacking in people-skills.
Thinking he could do every piece of code better than me.
When I disagreed with him on point 2, he would call meetings with the other programmers to 'teach me how to do things'. These meetings tended to be three programmers spending three hours convincing him he was wrong to start with.
He would check out my code when I wasn't looking and 'optimize' it -- that is making it run slower and introduce subtle bugs that I would spend days tracking down.
Finally, when I pointed out his inadequacies as a manager, he got all vengeful and removed all resources from my project, hoping to kill it and get me fired.
When the project succeeded anyway, he took all credit for it.
Now tell me again that programmers make good managers, and I will laugh in your general direction. The best managers I've had knew nothing about programming, but they knew how to ask the right questions (when will it be done, what do you need to do it faster, how can I help you achieve your goals) and leave the programming to the experts.
Why are americans so hell-bent on imposing their value systems on everyone else? What the chinese do should be their business. If american companies want to do business with China, it will of course be done on the Chinese's terms -- critizising the companies for this is counter-productive.
Now, I value my right to free speach as much as you, but I don't feel I have the right to critisize the Chinese way of doing things. After all, China is the oldest state in the world -- they must have been doing something right.
Charles Poynton maintains an excellent FAQ for color and gamma related questions, that is infrequenctly posted to comp.graphics.algorithms. You can get it online here:
http://www.inforamp.net/~poynton/ColorFAQ.html
Operator overloading would be a good idea if it was implemented correctly. C++'s implementation of operator overloading is not very good because:
It generates temporary objects that are created and deleted behind the scenes. This means something as simple as A = B + C will turn into a creation of a temporary to hold the result of B+C and then a call to the copy constructor or operator= for the assignment, and then deleting the temporary.
operators cannot be made virtual, so the static linking nature of C++ sometimes give you other results than you expect. For example, overloading << for ofstream, you can write other classes to ofstream than before -- BUT if you write a string to the ofstream first, it returns a reference to an ostream (even if it's the same ofstream as before) and your fancy output is not called because it's not defined for ostream..
Trying to use as many language features as possible is a common pitfall when programming in modern day languages. If you try to use multiple inheritance, operator overloading etc. just because you can, you will end up with an unreadable, unmaintainable, undebuggable mess.
You should first analyze the requirements of your project before deciding which language is most suitable for the job. Just going by 'as many features as possible' will get you in trouble.
That being said, I think Python fulfills almost all the requirements your boss had. (I am not a Python programmer myself, but I'm interested in language design, so I have read some articles about Python...) On the other hand, if you really have a very complicated problem domain (like AI), Common LISP might be what you're looking for. If you want rapid prototyping above all, maybe Smalltalk is your best bet - If you want performance, write in C or C++, if you want portability you can do a lot worse than picking Java -- You get the picture...
Please read the article before posting next time! He's talking about PlayStation development, not Mac, not PC!
I'm working in the games development industry, both console development and PC -- the article is talking about a platform for _development_, not for running the final product. In a typical game development environment, you spend half your time writing tools to create content -- if you can write tools faster on OS X with Cocoa, you'll save time and money. I found this article very interesting and I've already invested in an iBook to do some homework and try out Cocoa...
> The game of go is even worse for comps. Top
> programs still can't touch the human masters.
I don't think the writer appreciates just how different the situation is for go. The top programs can't beat a relative beginner, while a professional player can pass 25 times in a row and still beat them (Janice Kim vs. Handtalk).
The problem with go is that searching doesn't work for two reasons:
Writing an evaluation function for middle game positions is very very hard
The branching factor (the number of legal moves in a position) is much much higher than in chess
Add to that the fact that there are no easy way of applying opening books to go (the professional players say you can successfully start anywhere on the 19x19 board as long as you stay away from the first and second lines from the edges), and you have a situation where the brute-force approach breaks down completely.
The current top Go programs all try to mimic the human way of playing Go, by applying pattern matches and having expert-system like rules for what to do in each pattern. Of course this fails miserably because you need a whole-board understanding to advance beyond the beginner level.
If you write a Go program that can beat a reasonably strong amateur (1 dan level), you will be a millionare for sure, so get started, go look at "the interactive way to Go" and learn the rules, write a computer player and collect the cash;-)
In fact, this (more) technical explanation of what they are doing sounds a lot more sensible. Using multidimensional analysis tools in an entropy coder is in fact a novel idea, and if you look up Kolmogorov Complexity on MathWorld you will see that this is an entropy measure for multidimensional data just as Shannon defined entropy for single dimensional data.
The further you get from this (i.e. their press release and then the Reuters article) the more the facts are distorted and the claims inflated.
This reeks of cheap ploy to trick investors - 1) A company that has never developed a working product but made great claims to how they are going to revolutionize websearches buys a small company no-one has heard of. 2) An announcement is made to the financial world that this small company has a revolutionary technology that combined with the revolutionary technology of the original company will make everyone's dreams come true. 3) The stock price of the first company goes up. 4) Everyone who knows anything sells their shares and leave, leaving the investors high and dry...
I thought the point of writing up an interview was to isolate the interesting information from all the [sigh] and [mumbles] and "well..." and " I think [uhm] maybe - well, if you look at it like this, it might be.."
Why was this terrible writeup ever posted? I couldn't bear to read more than half of it...
Why didn't they pick the elegant solution?
on
Broadband Crackdown
·
· Score: 1
They should just create a strain of Code Red that goes in and patches the hole - shouldn't be too difficult... They could call it Code Green or something -- There, I even did the hard part for them, coming up with the name;)
Slashdot Mirror
This article gives an introduction to the problems involved in getting computers to play Go:
http://www.ishipress.com/times-go.htm
and I don't know why anyone listens to him.
He thinks 300 page game designs stifles creativity, but how does he suggest you get 20+ people to work together to produce a coherent result without a detailed design? Idiot.
Furthermore he thinks that it stifles creativity to make a game on a budget and with a timeschedule... Well, of course it does but unless you have an operating system monopoly paying your bills this situation is not going to change anytime soon. Idiot.
And his suggestion to use focus groups in game development -- I've tried that, but the average joe is not a game designer and does not produce valuable feedback apart from "I like it" or "I don't like it". One example of why this fails is a game I worked on that had a bug in the collision response in one of the focus group builds -- it was slated by them, but _none_ of them mentioned the collision response being bad. Idiot.
Saying that a 300 page game design stifles creativity is completely wrong, unless for some reason your publisher is requiring you to stick to the letter of it instead of being flexible. How do you think you get a team of 20+ people to produce a coherent game? Normally you can't see which parts of the game was made by which artist and so on, why? BECAUSE THERE IS A GAME DESIGN DETAILING HOW THINGS SHOULD LOOK AND WORK. Of course if something is not fun, you come up with a new design for that part and update the game design document accordingly.
He also seems to think that everyone can do business like Microsoft where it does not matter how much money you lose because there is always the operating system monopoly there to feed you... Saying that developers make bad games because they have to make them on a budget and a timeschedule is of course true, but not very interesting as this is likely to continue to be the case for the foreseable future...
Games are about user interaction, that is taking the controller and using it to move something around on the screen. There are far too many game developers who forget the fundamentals. Actually it is a mark of a true professional that they focus on the fundamentals, this is why Shigeru Miyamoto has developed an unprecedented string of hit titles and is respected by almost everyone in the industry...
The second is whether reverse-enginering is considered breaking copy-protection. The court ruled that translating source code into binary form did not in itself count as protecting it, because there are other good reasons for distributing binaries (although I'm sure some OSS advocates think differently ;-) besides protecting the source code. Thus reverse engineering the source code from the object code could not be considered breaking any protection.
Just for the record, everyone in the games industry (including myself) should remember that they owe their living to Miyamoto-sensei. I'd love to see Samus Blackly strapped to a chair forced to play BMX-XXX for eternity...
The only good alternative is to port Bink to Linux, and I hope that is what they end up doing.
RAD Game Tools, the company who sells Miles is the same company that sells Bink, but since I have never used Bink, I don't know how easy it would be to port it to Linux. But as a general rule, it would be better to use effort to port APIs that many developers use rather than porting specific applications like NNN... Just my .02 Euros...
All the programmers at my workplace have been turned on to sushi lately -- finally, food you can order that doesn't exasperate that ulcer that any programmer who has gone through a few product cycles in the games industry is bound to develop...
You need per person:
Start boiling water (as this is the the task that takes longest to complete), put some salt in the water (a tablespoon).
Split the tomatoes open, get rid of all the watery goo and seeds so only the firm flesh of the tomatoes remain. Chop them as finely as you can be bothered to.
Put some oil in a frying pan, put the pan on the heat and finely chop the garlic. Put the garlic in the pan, and then before it turns brown (that is after 10 seconds) add the tomatoes.
When the water boils, add the pasta. If you don't turn down the heat very much, so the water continues to boil violently you don't have to stir very much ;-)
Keep stirring the tomatoes around until they turn soft and start looking a bit like sauce. Check if the pasta is finished by getting a strand out of the boiling water once in a while and biting it. If it doesn't have a hard core, it's finished.
Get rid of the water for the pasta. Add tomatoes to pasta, add some extra virgin olive oil and stir. Put on plates, sprinkle with grated chese and some leaves of basil if available...
There you go, one of my favourite easy dishes...
However, I think it's a real problem that he is not presenting any results. At the end of the day, any scientific method has to be judged by the results it produces, and all Wolfram does is to re-cast familiar results in terms of cellular automa.
So to sum it up -- if you're dying to learn about cellular automa, read the book. If you're looking for a revolution, I'm afraid you'll have to start it yourself...
The US is giving Israel billions of dollars worth of arms every year that they use to oppress arabs, all the time pretending to try to make peace in the middle east. How can you be a peacebroker when you're supplying the arms to one side??
Some people have a natural talent for this, and they can become good managers without education -- but in most cases I would prefer to get someone with some credentials (other than being a former programmer)...
Now tell me again that programmers make good managers, and I will laugh in your general direction. The best managers I've had knew nothing about programming, but they knew how to ask the right questions (when will it be done, what do you need to do it faster, how can I help you achieve your goals) and leave the programming to the experts.
Now, I value my right to free speach as much as you, but I don't feel I have the right to critisize the Chinese way of doing things. After all, China is the oldest state in the world -- they must have been doing something right.
Charles Poynton maintains an excellent FAQ for color and gamma related questions, that is infrequenctly posted to comp.graphics.algorithms. You can get it online here: http://www.inforamp.net/~poynton/ColorFAQ.html
You should first analyze the requirements of your project before deciding which language is most suitable for the job. Just going by 'as many features as possible' will get you in trouble.
That being said, I think Python fulfills almost all the requirements your boss had. (I am not a Python programmer myself, but I'm interested in language design, so I have read some articles about Python...) On the other hand, if you really have a very complicated problem domain (like AI), Common LISP might be what you're looking for. If you want rapid prototyping above all, maybe Smalltalk is your best bet - If you want performance, write in C or C++, if you want portability you can do a lot worse than picking Java -- You get the picture...
I'm working in the games development industry, both console development and PC -- the article is talking about a platform for _development_, not for running the final product. In a typical game development environment, you spend half your time writing tools to create content -- if you can write tools faster on OS X with Cocoa, you'll save time and money. I found this article very interesting and I've already invested in an iBook to do some homework and try out Cocoa...
> programs still can't touch the human masters.
I don't think the writer appreciates just how different the situation is for go. The top programs can't beat a relative beginner, while a professional player can pass 25 times in a row and still beat them (Janice Kim vs. Handtalk).
The problem with go is that searching doesn't work for two reasons:
Writing an evaluation function for middle game positions is very very hard
The branching factor (the number of legal moves in a position) is much much higher than in chess
Add to that the fact that there are no easy way of applying opening books to go (the professional players say you can successfully start anywhere on the 19x19 board as long as you stay away from the first and second lines from the edges), and you have a situation where the brute-force approach breaks down completely.
The current top Go programs all try to mimic the human way of playing Go, by applying pattern matches and having expert-system like rules for what to do in each pattern. Of course this fails miserably because you need a whole-board understanding to advance beyond the beginner level.
If you write a Go program that can beat a reasonably strong amateur (1 dan level), you will be a millionare for sure, so get started, go look at "the interactive way to Go" and learn the rules, write a computer player and collect the cash ;-)
The further you get from this (i.e. their press release and then the Reuters article) the more the facts are distorted and the claims inflated.
This reeks of cheap ploy to trick investors - 1) A company that has never developed a working product but made great claims to how they are going to revolutionize websearches buys a small company no-one has heard of. 2) An announcement is made to the financial world that this small company has a revolutionary technology that combined with the revolutionary technology of the original company will make everyone's dreams come true. 3) The stock price of the first company goes up. 4) Everyone who knows anything sells their shares and leave, leaving the investors high and dry...
I thought the point of writing up an interview was to isolate the interesting information from all the [sigh] and [mumbles] and "well..." and " I think [uhm] maybe - well, if you look at it like this, it might be.."
Why was this terrible writeup ever posted? I couldn't bear to read more than half of it...
They should just create a strain of Code Red that goes in and patches the hole - shouldn't be too difficult... They could call it Code Green or something -- There, I even did the hard part for them, coming up with the name ;)