Anytime, anywere... as long as the game is played using the traditional time rules. Nobody even tries to play against the top computers in 5-10 minutes per player games. Even the top players get smashed.
"Classic" genetic algorithms are little more than a pretty efficient search algorithms. They are pretty useful only if it is easy to recongize the level of "quality" of all individuals. When "growing" programs finding such a fitness function is hard. Really hard.
The simplest approach is to come up with test cases, and rate the programs based on how many test cases they fulfill. There is a problem with this approach though: if all requirements are hard to fulfill (let's say, only 1 in 10 million random program fulfills at least one of them), there is no way that approach to program fitness is ever going to work. The complexity of any modern programming language is so high that using genetic algorithms to search for "a program in C that calculates fibonacci numbers" seems completely outlandish. The cost in electricity to find such a program would be higher than just giving the same specifications to a programmer.
Having worked on applying GAs to multi-objective optimization, I don't belive that this technique can be used effectively to optimize most programs.
The main issue is to compare the individuals generated by the genetic algorithm. To do so, we need to both compile the program under the specified settings, and then to be able to benchmark its performance. In my current job, a full build of our main product takes over 12 hours on an 8 CPU machine. Using a pretty conservative estimate, 100 generations with 100 individuals each, we'd be talking about more than a year of CPU time for a single run of the algorithm!
Even if we could ignore the amount of time required for compilation, we still have a second, more important flaw: Most programs out there are not really that easy to benchmark automatically. Database applications might need to go back to a known DB state to be able to run the benchmark faily. Also, server apps need to have the exact same load for every test if we want to be able to compare two executables fairly. This problem is increased when many compiler options will just create a 1% performance improvement or so. A poorly run system could lead the comparison function to just pick the wrong executable if the two executables didn't run in the exact same conditions.
I see how using GE for this task has a high coolness factor, and how it might even be usable for applications that are by nature easier to benchmark, but don't expect this technique to be applicable to enterprise-sized server applications, or even most GUI based apps any time soon.
That's a good amount of effort. The projectionist should also get a "big nuts" award, since that kind of thing can get you fired.
In my local theater the ad just got booed by over 50 percent of the audience. The more I think about the campaign, the more it seems to me that the MPAA is just making sure that they are seen by the public as a bunch of greedy, lying SOBs. They could at least they tried to make the stories beliveable... The starving stuntman does not cut it.
It seems to me it's all about deciding what makes a game good. I belive that trying to do something that was never done before(deer hunter, B&W), having impressive graphics (Myst, FFVII) or even just using a franchise succesfully (Enter the Matrix) makes a game better. A game doesn't have to do everything right to become a hit. Plenty of times it's enough to do one thing really, really well.
A "well educated" gamer has played more than enough good games that he can become way less forgiving with a game's flaws. Repetitive gameplay, clipping problems, or just not being an innovative game become major flaws. Does FFVII seem that impressive after playing Chrono Trigger? What about the repetitive sections of Halo? To many people, flawed games seem great, just because they've never seen what a quality, highly polished game is all about. As long as the game does one thing really well, the casual gamer will buy the game, and recommend it to all of his friends, and that's what makes a game a hit.
So yes, I belive that Enter the Matrix had some really interesting ingredients that separated it from the crowd. So did Black and White, Halo, Tomb Raider and many others. Don't ask me to play though any of them again though: I'd rather play Metroid.
I really don't think so. This game seems to have nothing to do with a standard miniature game like warhammer. In fact, just the price difference makes the games completely different. A $30-$50 game is very different from spending a few hundred dollars per player, + the time investment of painting the miniatures, which appeals to some people, but is a no-no to many others.
Warcraft does take a lot of artwork and race design from Warhammer, but the game mechanics have nothing to do with each other. Thus, a Warcraft boardgame will have very little to do with Warhammer. In fact, it doesn't even have miniatures!
Programmer productivity? let's think about TRULY importants stuff. Is there any better reason to develop a brain-machine interface than to allow me to have l33t aiming ski11z in Quake 3? Who wouldn't go through surgety to become the ultimate railgun master?
Rare is known to use Duke Nukem Forever-like development schedules. It's the main reason it got them in trouble with Nintendo in the first place. It's been a year since they've been working on the XBox, and they've already pushed Kameo back to 2004. Are they ever going to release anything worthwhile?
I don't know about you guys, but I'm not sure MS made a good deal buying Rare. With their best developers gone, are we going to see anything good from them? I sure hope so, but I'm less and less confident every month.
He's just talking about two sega games, Jet Set Radio and Jet Set Radio Future, that happened to sell way less than expected in the US. MS ended up blundling JSRF with the XBox.
The "Megaman Battle Network" series for the GBA has been an RPG for a long time, and they are actually pretty good. Making another game with similar mechanics in a home console seemed like a pretty logical step to me. Besides, haven't all megaman shooters been the same game with minor tweaks and a graphical overhaul? Do we really need yet another one?
The issue is not one of the lucky few that were able to make a fortune based on their skill and/or effort. The problem is that the money that gets transferred to those few people doesn't really go back to the general population. When, let's say, Jeff Bezos dies at the venerable age of 85, where does the money go to? to his children, who just were born to the right parents. A 8 or 10 billion dollar fortune means at least 4 or 5 generations of people that really don't have to do much more than hiring an accountant to earn a living that most kids born in the inner city can't ever accomplish: a new nobility of sorts. Having most people be able to become rich is a good thing for society, despite very few people being able to actually make it. People that don't ever have to work to earn a living aren't really helping anybody but themselves. The most important thing a government must do to have a healthy society is to try to avoid having people doomed to poverty, and people that are insanely wealthy w/o having worked for it.
IMNSHO, the main problem is that on many games the enemy AI is coded way too close to the end of developement. If the internal data structures in a strategy game are not designed with the AI in mind, developing a challenging and fair AI is a daunting task. Thus, many strategy game developers end up having to create the enemy AI by reusing code. Yes, the same sub-par code that they created 2 or three years ago for the previous version of the game, or maybe just another game that used similar mechanics.After a week or two making sure that the computer uses some of the latest game features that the previous version didn't have, the AI is done.
Of course, some game makers like Lionhead are making AI in strategy games more of an issue, but many developers are still cutting corners in the enemy AI due to the publisher's pressure.
If you buy a PC strategy game this fall, and you get the feeling the enemy AI is pretty similar to the one of the previous version, who knows? you might be facing the exact same enemy you played 4 years ago.
Unfortunately, lower prices have little to do with game costs, or with how many people buy it. The way games are priced nowadays is really simple: "Whatever the market can bear". If going to the movies costed $50, nobody would go watch one, since the average moviegoer desn't believe that price is right, and would switch to other forms of entertainment. Thus, the studios/theatres keep the prices lower.
The videogame publishers believe that if they released new games at $20-30 a pop from week one their profits would not increase at all. The people that wouldn't buy the game for $50 will just buy it later, when it is discounted in the "platinum-series" or "Nintendo Classics" or whatever. Didn't you notice how games that are still selling well, like Halo, haven't been discounted on most major stores? Besides, game makers like Nintendo and Sony have been saying that the only thing they feared in "the console wars" was not cheaper consoles, but cheaper games, since if a company starts releasing their latest-and-greatest games for $30 to outsell the competition, everyone else will have to do the same, and all you'd get is lowered profits for everyone.
All of this, and the always increasing cost of making flashy and complicated games, leads to the only possible conclusion: Since the publishers believe that cheaper games aren't good for them, we won't see cheap games any time soon.
In the red/blue to gold/silver transition a beefed up pokemon from red/blue woudn't help much until you have the badge that allows you to "tame" high level pokemon. They'd just refuse to obey your orders at all. After you got the badge though, the game was pretty much over.
You seem to be confusing freecraft, a RTS pretty similar to Warcraft II, to bnetd, an open-source battle.net server.
It is pretty likely thta bnetd, despite having legit uses, was in some way hurting Vivendi's bottom line. However, I don't see how freecraft can hurt their bottom line at all. What do they gain by killing a Linux freeware game?
I meant, of course, the GCN version. The Kirby license did get removed from the game, even though it would use there Kriby's Tilt and Tumble cartdridge's motion sensor. I've not found an article about the game in about a year, but Nintendo still has a product page for it on their website. It's definetly not going to be a high profile game, but I doubt that Nintendo has canned it altogether.
Kirby's Tilt 'n Tumble just lost it's Kirby license (They didn't want to dilute the Kirby franchise, IIRC), and is now called Roll-O-Rama. The game has no release date in Japan, and hasn't been announced as an US or EU bound title AFAIK.
Cord too short? buy a Wavebird, Nintendo's wireless controller.
For me, it's precisely how good the controller is that makes me rather play multiplatform games in the cube rather than in the other two consoles, even though the terrible jaggies on most multiplatform PS2 games is a major factor too.
It's likely some hardware incompatibility issue. Unacceptable, if you ask me. However, don't think that Windows doesn't have this problems too.
The machine I'm typing this message on is an Athlon 1ghz that was supposed to dual-boot windows 2000 and Suse 8. I decided that the best way of doing this was just to install 2000 on the new, unformatted disk, and then let the Suse installer handle the partition resizing/dual booting magic. I was surprised when the win2k installer would just crash every single time. The same installer disk worked fine on all of my other machines, so I was starting to think that the hardware was to blame. I tried to install Suse, just to confirm my hypothesis, and Io! Suse installed flawlessly using the normal installer. Suse 8 has been running with no problems ever since. I thought I'd never see the day a Linux install would be easier than a Windows' one, but 2k's crashing installer ain't really that hard to beat.
I started using computers in the early 80s and I make my living at it. If Windows refuses to install while Linux works fine, Bill might have something to worry about.
Not quite. It seems pretty easy to me to think that those people just believe is a completely different set of IP laws. In a country where IP law claimed just that you can't modify/take ownership of somebody else's work, but you can make as many copies of it as you want, the GPL would work quite nicely, while allowing filesharing. There are a number of other possible interpretations of IP that would match this scenario.
This doesn't mean that every single filesharer that believes in the GPL isn't an hypocrite though: I'm sure you'll still find plenty of those.
BTW, I don't hold the beliefs I gave as an example, I just know people that do.
Remember good old Doom II deathmatch? On many, many maps, the game was more of a "cat and mouse" game, with only 4 players max and some really huge maps once you got past level 10 or so. I don't think that kind of multiplayer would be very popular with most people though.
Our ability to have robot wrkers for anything but completely automated assemly lines is limited by the pathetic status of any advanced AI research. The time where researchers dreamed of human-like computer brains are mostly gone. It seems that computer science has hit a roadblock: We have no idea of how to even start building an AI that had any real skill at communicating in a human language. If a robot cannot handle advanced communication, it can't really do most human work.
I'd rather put my money on our understanding of genetics changing the way we live. Super humans seem more likely than human-like robots to me, as long as the current limitations on human cloning/research get lifted.
A circular grid around the current unit would make things a little more manageable, but squares and hexes are the best way to make calculating movement and firing ranges easier.
Slashdot Mirror
Anytime, anywere... as long as the game is played using the traditional time rules. Nobody even tries to play against the top computers in 5-10 minutes per player games. Even the top players get smashed.
"Classic" genetic algorithms are little more than a pretty efficient search algorithms. They are pretty useful only if it is easy to recongize the level of "quality" of all individuals. When "growing" programs finding such a fitness function is hard. Really hard.
The simplest approach is to come up with test cases, and rate the programs based on how many test cases they fulfill. There is a problem with this approach though: if all requirements are hard to fulfill (let's say, only 1 in 10 million random program fulfills at least one of them), there is no way that approach to program fitness is ever going to work. The complexity of any modern programming language is so high that using genetic algorithms to search for "a program in C that calculates fibonacci numbers" seems completely outlandish. The cost in electricity to find such a program would be higher than just giving the same specifications to a programmer.
Having worked on applying GAs to multi-objective optimization, I don't belive that this technique can be used effectively to optimize most programs.
The main issue is to compare the individuals generated by the genetic algorithm. To do so, we need to both compile the program under the specified settings, and then to be able to benchmark its performance. In my current job, a full build of our main product takes over 12 hours on an 8 CPU machine. Using a pretty conservative estimate, 100 generations with 100 individuals each, we'd be talking about more than a year of CPU time for a single run of the algorithm!
Even if we could ignore the amount of time required for compilation, we still have a second, more important flaw: Most programs out there are not really that easy to benchmark automatically. Database applications might need to go back to a known DB state to be able to run the benchmark faily. Also, server apps need to have the exact same load for every test if we want to be able to compare two executables fairly. This problem is increased when many compiler options will just create a 1% performance improvement or so. A poorly run system could lead the comparison function to just pick the wrong executable if the two executables didn't run in the exact same conditions.I see how using GE for this task has a high coolness factor, and how it might even be usable for applications that are by nature easier to benchmark, but don't expect this technique to be applicable to enterprise-sized server applications, or even most GUI based apps any time soon.
That's a good amount of effort. The projectionist should also get a "big nuts" award, since that kind of thing can get you fired.
In my local theater the ad just got booed by over 50 percent of the audience. The more I think about the campaign, the more it seems to me that the MPAA is just making sure that they are seen by the public as a bunch of greedy, lying SOBs. They could at least they tried to make the stories beliveable... The starving stuntman does not cut it.
It seems to me it's all about deciding what makes a game good. I belive that trying to do something that was never done before(deer hunter, B&W), having impressive graphics (Myst, FFVII) or even just using a franchise succesfully (Enter the Matrix) makes a game better. A game doesn't have to do everything right to become a hit. Plenty of times it's enough to do one thing really, really well.
A "well educated" gamer has played more than enough good games that he can become way less forgiving with a game's flaws. Repetitive gameplay, clipping problems, or just not being an innovative game become major flaws. Does FFVII seem that impressive after playing Chrono Trigger? What about the repetitive sections of Halo? To many people, flawed games seem great, just because they've never seen what a quality, highly polished game is all about. As long as the game does one thing really well, the casual gamer will buy the game, and recommend it to all of his friends, and that's what makes a game a hit.
So yes, I belive that Enter the Matrix had some really interesting ingredients that separated it from the crowd. So did Black and White, Halo, Tomb Raider and many others. Don't ask me to play though any of them again though: I'd rather play Metroid.
I really don't think so. This game seems to have nothing to do with a standard miniature game like warhammer. In fact, just the price difference makes the games completely different. A $30-$50 game is very different from spending a few hundred dollars per player, + the time investment of painting the miniatures, which appeals to some people, but is a no-no to many others.
Warcraft does take a lot of artwork and race design from Warhammer, but the game mechanics have nothing to do with each other. Thus, a Warcraft boardgame will have very little to do with Warhammer. In fact, it doesn't even have miniatures!
I guess someone is using Perl a little too much lately. Fortunately, instead of a channel discussing Perl modules, we're stuck with C-SPANInstead.
Programmer productivity? let's think about TRULY importants stuff. Is there any better reason to develop a brain-machine interface than to allow me to have l33t aiming ski11z in Quake 3? Who wouldn't go through surgety to become the ultimate railgun master?
Rare is known to use Duke Nukem Forever-like development schedules. It's the main reason it got them in trouble with Nintendo in the first place. It's been a year since they've been working on the XBox, and they've already pushed Kameo back to 2004. Are they ever going to release anything worthwhile?
I don't know about you guys, but I'm not sure MS made a good deal buying Rare. With their best developers gone, are we going to see anything good from them? I sure hope so, but I'm less and less confident every month.
He's just talking about two sega games, Jet Set Radio and Jet Set Radio Future, that happened to sell way less than expected in the US. MS ended up blundling JSRF with the XBox.
The "Megaman Battle Network" series for the GBA has been an RPG for a long time, and they are actually pretty good. Making another game with similar mechanics in a home console seemed like a pretty logical step to me. Besides, haven't all megaman shooters been the same game with minor tweaks and a graphical overhaul? Do we really need yet another one?
The issue is not one of the lucky few that were able to make a fortune based on their skill and/or effort. The problem is that the money that gets transferred to those few people doesn't really go back to the general population. When, let's say, Jeff Bezos dies at the venerable age of 85, where does the money go to? to his children, who just were born to the right parents. A 8 or 10 billion dollar fortune means at least 4 or 5 generations of people that really don't have to do much more than hiring an accountant to earn a living that most kids born in the inner city can't ever accomplish: a new nobility of sorts. Having most people be able to become rich is a good thing for society, despite very few people being able to actually make it. People that don't ever have to work to earn a living aren't really helping anybody but themselves. The most important thing a government must do to have a healthy society is to try to avoid having people doomed to poverty, and people that are insanely wealthy w/o having worked for it.
IMNSHO, the main problem is that on many games the enemy AI is coded way too close to the end of developement. If the internal data structures in a strategy game are not designed with the AI in mind, developing a challenging and fair AI is a daunting task. Thus, many strategy game developers end up having to create the enemy AI by reusing code. Yes, the same sub-par code that they created 2 or three years ago for the previous version of the game, or maybe just another game that used similar mechanics.After a week or two making sure that the computer uses some of the latest game features that the previous version didn't have, the AI is done.
Of course, some game makers like Lionhead are making AI in strategy games more of an issue, but many developers are still cutting corners in the enemy AI due to the publisher's pressure.
If you buy a PC strategy game this fall, and you get the feeling the enemy AI is pretty similar to the one of the previous version, who knows? you might be facing the exact same enemy you played 4 years ago.
Unfortunately, lower prices have little to do with game costs, or with how many people buy it. The way games are priced nowadays is really simple: "Whatever the market can bear". If going to the movies costed $50, nobody would go watch one, since the average moviegoer desn't believe that price is right, and would switch to other forms of entertainment. Thus, the studios/theatres keep the prices lower.
The videogame publishers believe that if they released new games at $20-30 a pop from week one their profits would not increase at all. The people that wouldn't buy the game for $50 will just buy it later, when it is discounted in the "platinum-series" or "Nintendo Classics" or whatever. Didn't you notice how games that are still selling well, like Halo, haven't been discounted on most major stores? Besides, game makers like Nintendo and Sony have been saying that the only thing they feared in "the console wars" was not cheaper consoles, but cheaper games, since if a company starts releasing their latest-and-greatest games for $30 to outsell the competition, everyone else will have to do the same, and all you'd get is lowered profits for everyone.
All of this, and the always increasing cost of making flashy and complicated games, leads to the only possible conclusion: Since the publishers believe that cheaper games aren't good for them, we won't see cheap games any time soon.
In the red/blue to gold/silver transition a beefed up pokemon from red/blue woudn't help much until you have the badge that allows you to "tame" high level pokemon. They'd just refuse to obey your orders at all. After you got the badge though, the game was pretty much over.
You seem to be confusing freecraft, a RTS pretty similar to Warcraft II, to bnetd, an open-source battle.net server.
It is pretty likely thta bnetd, despite having legit uses, was in some way hurting Vivendi's bottom line. However, I don't see how freecraft can hurt their bottom line at all. What do they gain by killing a Linux freeware game?
I meant, of course, the GCN version. The Kirby license did get removed from the game, even though it would use there Kriby's Tilt and Tumble cartdridge's motion sensor. I've not found an article about the game in about a year, but Nintendo still has a product page for it on their website. It's definetly not going to be a high profile game, but I doubt that Nintendo has canned it altogether.
Kirby's Tilt 'n Tumble just lost it's Kirby license (They didn't want to dilute the Kirby franchise, IIRC), and is now called Roll-O-Rama. The game has no release date in Japan, and hasn't been announced as an US or EU bound title AFAIK.
Cord too short? buy a Wavebird, Nintendo's wireless controller.
For me, it's precisely how good the controller is that makes me rather play multiplatform games in the cube rather than in the other two consoles, even though the terrible jaggies on most multiplatform PS2 games is a major factor too.
It's likely some hardware incompatibility issue. Unacceptable, if you ask me. However, don't think that Windows doesn't have this problems too.
The machine I'm typing this message on is an Athlon 1ghz that was supposed to dual-boot windows 2000 and Suse 8. I decided that the best way of doing this was just to install 2000 on the new, unformatted disk, and then let the Suse installer handle the partition resizing/dual booting magic. I was surprised when the win2k installer would just crash every single time. The same installer disk worked fine on all of my other machines, so I was starting to think that the hardware was to blame. I tried to install Suse, just to confirm my hypothesis, and Io! Suse installed flawlessly using the normal installer. Suse 8 has been running with no problems ever since. I thought I'd never see the day a Linux install would be easier than a Windows' one, but 2k's crashing installer ain't really that hard to beat.
I started using computers in the early 80s and I make my living at it. If Windows refuses to install while Linux works fine, Bill might have something to worry about.
Not quite. It seems pretty easy to me to think that those people just believe is a completely different set of IP laws. In a country where IP law claimed just that you can't modify/take ownership of somebody else's work, but you can make as many copies of it as you want, the GPL would work quite nicely, while allowing filesharing. There are a number of other possible interpretations of IP that would match this scenario.
This doesn't mean that every single filesharer that believes in the GPL isn't an hypocrite though: I'm sure you'll still find plenty of those.
BTW, I don't hold the beliefs I gave as an example, I just know people that do.
Yep, In some countries, it's illegal. I wish the US laws helped the employee for once.
Remember good old Doom II deathmatch? On many, many maps, the game was more of a "cat and mouse" game, with only 4 players max and some really huge maps once you got past level 10 or so. I don't think that kind of multiplayer would be very popular with most people though.
Our ability to have robot wrkers for anything but completely automated assemly lines is limited by the pathetic status of any advanced AI research. The time where researchers dreamed of human-like computer brains are mostly gone. It seems that computer science has hit a roadblock: We have no idea of how to even start building an AI that had any real skill at communicating in a human language. If a robot cannot handle advanced communication, it can't really do most human work.
I'd rather put my money on our understanding of genetics changing the way we live. Super humans seem more likely than human-like robots to me, as long as the current limitations on human cloning/research get lifted.
A circular grid around the current unit would make things a little more manageable, but squares and hexes are the best way to make calculating movement and firing ranges easier.