At any rate, I don't know about you, I certainly wouldn't want to have sex with somebody who [...] has three foot long razor edged spider legs growing out of her back.
It's called tentacle porn, and has a pretty large cult following both in Japan and North America.
Aribeth (Neverwinter Nights)-- a female rising to supreme commander of a military force, plus she has an actual backstory as opposed to being a one-dimensional "commanding officer" stereotype.
Also a nice change (especially if you're used to playing Japanese dating sims alot) is that if you hit on her, she actually turns you down.
How many of you assumed Samus was a man before you found out the truth? Why did you assume that?
I found out from word-of-mouth that Samus was female before I had ever played the game, but I think I would have assumed she was male had I not known. Why? Probabilities based on past experience: Pretty much every other lead character in every other video game I've played was male, and so if I use this as a model to predict what the gender of this character would be, assuming that she was male was the only logical choice.
Of course, you could argue "Another logical choice would be to make no assumption at all", but I think we, as humans, find it more natural to make assumptions and remember that we've made assumptions (i.e. "I assume Samus is male, but I have no evidence.") than to make no assumptions at all.
Whenever I re-install some Linux distro (Fedora, Debian, etc.) I have to sit through a few hours of updates too. This isn't a Microsoft specific problem.
This actually isn't so much an OS issue as it is an e-mail client issue. You could imagine someone writing an e-mail client in Linux that would execute attachments in e-mails it received.
all the other software you need to buy for Windows to make it actually do something (antivirus, mail server, more antivirus,
Most anti virus programs I know of strongly recommend that you don't install two anti virus programs onto the same OS, as one might mistake the other's viral definition files as actual viruses.
My four year old can already configure a printer, setup a network and install drivers for her new wireless card in Windows since it's so easy.
Actually, she probably could. Plug the USB cable from the printer into the computer (she can just try every port until she finds one that fits, 4 year olds can solve those "fit the block into the hole with the same shape" type problems right?), and WindowsXP says "Your printer is installed and ready to use." When I tried installing a printer on Linux, it involved selecting the printer's manufacturer and model from a drop down list, which would require reading, which a 4 year old might not be capable of.
Plug in the network cable, and Window's default behaviour is to assign you with a default workgroup name and to "magically" obtain an IP automatically. Last time I checked, most linux distribution's default action is to ask you what your local host name and domain names are, without even telling you if they should be the same across all computers on the network, or they shouldn't be the same across all computers on the network. When you try to tell Linux to get you an IP automatically, it'll ask you for the IP of a DHCP server. Damned if my 4 year old knows what a DHCP server is and what its IP address migth be.
For wireless cards, you've got me, 'cause I haven't tried installing a wireless card in Windows yet, so I don't know what aids it might provide, but I'd imagine it'd feel just as "automatic" as the previous two cases.
Will it transparently sync with itunes or will it be syncable the hard way only ?
Straight from the site's FAQ: Probably. At this point the software support is still very immature. I haven't tried any of the software available for accessing the iTunes database under Linux.
If someone donates me an iPod, I'll gladly try it out and tell ya if it works.;)
How do you use a command line interface on an ipod?
If you read the documentation, it says when the iPod boots the Linux kernel it will automatically configure its local ethernet device (ethernet over firewire/IEEE1394) and then starts inetd so that network connections via telnet may be made.
The default configuration is for the iPod to use 192.10.1.2 and to allow telnet connections. The default address may be changed by editing the/etc/rc script.
According to the site's FAQ, there's an OGG player that plays at "80% realtime", whatever that means. They expect a running ogg player on there "soon".
Is it allowed to google for a reasonable human response to your statements? [...] in this way (with a large enough db like the web), a chatbot could appear to be human, but we probably wouldn't consider this AI.
I'd categorize it as the synthesis of two previously mostly distinct fields: Distributed AIs.
At the least, I'd like to see something that fails miserably, but attempts to "learn."
The fact that you've put learn in quotation marks hints at the problem of this reasoning. It's that it's very difficult (perhaps impossible) define "learn" (or "intelligence", for that matter). Let's say I write a chatbot that simply reads in the text that you type to it, and stores it in a log. Is that "learning"? Let's say it constructs Markov chains based on the way you type to just regurgitate exactly what you said but with enough variance so that it isn't obvious that it's simply repeating you. Is that learning? I know back in highschool sometimes I'd have to regurgitate something into an essay to get a passing grade from closed-minded teachers.
The point of having these tests is that you can actually say how successful your AI program is in a quantitive manner (e.g. % of humans who were convinced they were speaking to other humans and not chatbots).
I fail to see how fooling humans into thinking that they are having a conversation with another human, when it is really a chatbot, will do anything to produce artificial intelligence. It's an illusion, using technology, nothing more.
Not to put words in your mouth, but I'm assuming the argument you have in mind is something similar to "The computer is not really thinking, it's CPU is just sending signals around to produce output that merely LOOKS intelligent. It's an illusion".
Note though that one could argue that humans are not really thinking, their neurons are just sending signals around to rpdouce output that merely looks intelligent. It's an illusion.
Most people determine the intelligence of things by observing their behaviour. We assume that rocks are not very intelligent because their behaviour doesn't imply some sort of thinking process occuring internally. We assume that chimpanzees have some degree of intelligence because their behaviour DOES imply some sort of thinking process.
1. Get a Nintendo/PS2/XBOX/whatever I have these AND a windows machine, 'cause sometimes they release a game only for Windows.
2. Good software support. Hah. Most Windows programs are monolithic, clunky, closed systems (i.e. you can't extend them, script them, etc.). They may or may not conform to a UI model, and they may or may not even get along with each other. If you think what you're using is good, try OSX or KDE 3.1. You'll be amazed. I like eMule, Office, WinAmp, CloneCD, ICQ, MSN Messenger, Paint Shop Pro, L&H Japanese Translator, Visual Studio, Encarta, DiscJuggler, Reason, RPGMaker and Tag&Rename just ot name a few. I've never really encountered a situaiton where I'd want to script or extend them. You can name me the alternatives (lMule, OpenOffice, XMMS, etc.); I've tried them, and I don't like them as much as the originals. I may have had one or two instances of software not working together on a Windows machine (say Adobe doing some funky script thing in Office), but they tend to be merely annoyances; certainly nothing that would keep me from doing any work. But try installing an mp3 tagger in Debian which depends on something which depends on something which depends on something which depends on a version of libstdc++5 from the unstable branch, and you've pretty much killed a dozen or two of your apps.
3. I've never had Windows not just work (i.e. always, Windows just works), and I've installed it on 5 home systems, 2 systems for friends, and 10 systems in a LAN at work. Networking always worked, and video always worked, so for the few instances that the more obscure hardware didn't, I could just hop onto Windowsupdate.com and get the drivers. With every system I've installed Linux on but one, I found out the hardware was not supported. Things as basic as video, sound and networking would not work. Downloading a driver from a windows machine, splitting it into 1.4 meg chunks using WinRar, floppy-disking it onto the linux box, and downloading a linux unrar program, trying to get the driver to compile using the 20+ command line arguments the README file tells me to input, and being told that a certain option has been deprecated and not being told what the option replaces it is not fun.
Linux is a fine OS, but there are still plenty of reasons to use Windows. I run both.
The first games you play tend to be the ones you remember as being the best because you were just starting out in computer games and everything is still new and novel.
While I can see the logic in that argument, I don't think it's nescessarily a "universally true" statement. First FPS I played, I think, was Wolf3D, then Doom 1/2, ROTT, Quake 1/2, etc. Your standard path. While I thought ROTT was pretty funny, none of them really appealed to me all that much. The first FPS I really liked was Tribes, and the second, I think, was Call of Duty. In between I've played Quake 3, Unreal, Unreal Tournament, Unreal 2, Unreal Tournament 2003, Return to Castle Wolfeinstein, Half Life, and probably a whole bunch more I can't even remember.
By no means were the "first" FPSes I played my favorite, and a lot of my friends agree with me on which FPSes were great and which were just okay. When I think back to the first video games in general I played, I don't think any of those were my favorite either. On the NES, you have Duck Hunt, Super Mario Brothers, Bomberman, Milo's secret castle (or something like that), 8 eyes, etc. On the Sega Master System, you've got Outrun, Kung Fu Hero, Alexx Kidd in Wonderland, etc.
My favorites are like... Dance Dance Revolution, Chrono Trigger, Colin Mcrae Rally 3, Soul Calibur, etc.
So if there's a relationship between the order in which you play games, and the which ones you like the best, I don't think "earlier = better" is it.
3. Multilinear gameplay. THIS is where the "in the movies" feel comes from - where YOU, the GAMER, are picking what the story is.
This may be just a matter of opinion, but pretty much in every game where the burden of writing the story was put on me, the player, rather than on the designers had a pretty crappy story.
One of the best stories in a game, in my opinion, is the one from Chrono Trigger, which, while it has "multiple endings", is pretty linear otherwise.
From a technology point of view, I agree that RealMedia is pretty good; unfortunately, I think the format just got a bad rep because of the spyware loaded player.
I think this is more an issue of programs than codecs. While the article does state what programs it uses for the encoding (VirtualDub in the DivX case, Microsoft Windows Media Encoder in the WMF case, etc.), its main focus was on which of the codecs is the best, not which of the encoding programs is the best.
It cannot get any easier to setup a printer than it is in Mandrake. Simply click about 4 buttons and you are done.
Actually, I've had an easier experience on Windows: I plugged the cable into the USB port, and Windows asks me if I want to print a test page. I say yes, and it does, and the page looks fine, and so I was done. That's one button click.
Of course, you could argue that if Microsoft's Plug n Play for some reason failed, the "manual" installation process might take more than 4 clicks, but then again if the Mandrake "easy installation" process failed, the manual version would also probably take more than 4 clicks.
My experience indicates that it's extremely rare for a UNIX tool to lie in its return code value.
I don't think programmers intentionally make their command line programs lie via return codes; rather, it's not always clear what value you should return. The general guideline is that 0 means some sort of success occured, and anything else means some sort of failure.
If the program expects command line arguments, and you provide it none, the tradition is to display a short blurb about the program's usage. Is this considered a successful run, or a failure? You could argue that it failed in that it didn't do what the program was intended to do (e.g. encode mp3s), but you could also argue that the expected behaviour of the program, when provided zero parameters, is to display usage info, and it successfully did that.
What if, using whatever the proper syntax of the shell is, the user requests something like "encode all wave files in this directory" and there are zero wave files, and so the program encodes zero mp3s. Was that a success or a failure? It's a success in that nothing in the program broke, but it seems like perhaps something happened that the user isn't aware of, and perhaps the user should be informed.
What about if the user passes a "-h" parameter, asking for help, and the program displays the help. Was this a successful run of the program or a failure? It did exactly what the user asked, but it didn't fufill its main purpose, encoding mp3s in this example.
What if the user thinks "-h" means "highquality", but in fact the program interprets it as "display help"? The help blurb may be displayed, a success returned, and the files deleted.
I'm not advocating GUIs vs CLI or vice versa (I alternate between the two, depending on what I'm trying to get done), but I have a distate for Unix's return code system.
As a game designer, you wouldn't want to use floating point numbers for hit points, because float point arithmatic suffer from rounding errors.
As an oversimplification, let's say I store 200 billion in my floating point, but because of the precision I use, I only keep track of the 5 most significant digits, i.e. 2.0000 * 10 ^ 11. If someone where to deal one million points of damage, i.e. 0.00001 * 10 ^ 11, the substraction would yield 1.99999 * 10 ^ 11, which then needs to be rounded so that 5 digits are accurate. This number rounds to 2.0000 * 10 ^ 11, so with 5 digits of accuracy, dealing on million points of damage is equivalent as dealing 0 points of damage.
In real life, IEEE double precision gives you about 15 digits of accuracy, which is more digits than there are in "100 billion", so this might work, but I think most game designers will still avoid using FPNs because of these potential rounding errors.
It's not exactly an RPG, though it has RPG-ish element (i.e. you get experience for killing badguys which you can spend on upgrading your skills). I guess the closest equivalent is FF Tactics.
I haven't beaten the game yet, but it looks like it's possible to play the whole game without ever going through a single random encounter.
The way it works is you have a map that shows all the locations of interest, and there's an icon representing your location. You click where you want to go, and your icon pretty much travels in a straight line to where ever you clicked.
Every now and then, red dots appear randomly around the map. If you come into contact with one, then a random battle starts.
However, these dots, while not rare in the temporal sense, are pretty sparse in space. What I mean by that is that there'll always be one or two dots, but the dots are so small compared to the size of the map that if you don't want to get into random battles, it's trivial to walk around them and onto the "plot battles".
The benefit of them being common, temporal wise, is that if you get stuck on a mission, you can get into random battles pretty much at will, and level up your characters. Or maybe just play them for fun (since Silent Storm is very combat oriented, being a FF Tactics type game.)
So far I've voluntarily went through one random battle, and have never gotten into a random battle involuntarily. A friend of mine says he's beaten the game without ever having gotten into a random battle.
I think this is really the "optimal" solution: Let the players who want to have random battles play their random battles, and let players who don't like random battles not go through random battles.
Any potential RPG designers might want to give Silent Storm a try and see if they can adapt it to their designs.
Slashdot Mirror
How did it take them three years to figure that out? Wasn't the data right there in their hands?
The way I understand it, he crashed 3 years ago, they knew about it. 3 years later (i.e. today), he crashed again, and they know about it again.
At any rate, I don't know about you, I certainly wouldn't want to have sex with somebody who [...] has three foot long razor edged spider legs growing out of her back.
It's called tentacle porn, and has a pretty large cult following both in Japan and North America.
Aribeth (Neverwinter Nights)-- a female rising to supreme commander of a military force, plus she has an actual backstory as opposed to being a one-dimensional "commanding officer" stereotype.
Also a nice change (especially if you're used to playing Japanese dating sims alot) is that if you hit on her, she actually turns you down.
How many of you assumed Samus was a man before you found out the truth? Why did you assume that?
I found out from word-of-mouth that Samus was female before I had ever played the game, but I think I would have assumed she was male had I not known. Why? Probabilities based on past experience: Pretty much every other lead character in every other video game I've played was male, and so if I use this as a model to predict what the gender of this character would be, assuming that she was male was the only logical choice.
Of course, you could argue "Another logical choice would be to make no assumption at all", but I think we, as humans, find it more natural to make assumptions and remember that we've made assumptions (i.e. "I assume Samus is male, but I have no evidence.") than to make no assumptions at all.
Whenever I re-install some Linux distro (Fedora, Debian, etc.) I have to sit through a few hours of updates too. This isn't a Microsoft specific problem.
This actually isn't so much an OS issue as it is an e-mail client issue. You could imagine someone writing an e-mail client in Linux that would execute attachments in e-mails it received.
all the other software you need to buy for Windows to make it actually do something (antivirus, mail server, more antivirus,
Most anti virus programs I know of strongly recommend that you don't install two anti virus programs onto the same OS, as one might mistake the other's viral definition files as actual viruses.
My four year old can already configure a printer, setup a network and install drivers for her new wireless card in Windows since it's so easy.
Actually, she probably could. Plug the USB cable from the printer into the computer (she can just try every port until she finds one that fits, 4 year olds can solve those "fit the block into the hole with the same shape" type problems right?), and WindowsXP says "Your printer is installed and ready to use." When I tried installing a printer on Linux, it involved selecting the printer's manufacturer and model from a drop down list, which would require reading, which a 4 year old might not be capable of.
Plug in the network cable, and Window's default behaviour is to assign you with a default workgroup name and to "magically" obtain an IP automatically. Last time I checked, most linux distribution's default action is to ask you what your local host name and domain names are, without even telling you if they should be the same across all computers on the network, or they shouldn't be the same across all computers on the network. When you try to tell Linux to get you an IP automatically, it'll ask you for the IP of a DHCP server. Damned if my 4 year old knows what a DHCP server is and what its IP address migth be.
For wireless cards, you've got me, 'cause I haven't tried installing a wireless card in Windows yet, so I don't know what aids it might provide, but I'd imagine it'd feel just as "automatic" as the previous two cases.
Will it transparently sync with itunes or will it be syncable the hard way only ?
Straight from the site's FAQ: Probably. At this point the software support is still very immature. I haven't tried any of the software available for accessing the iTunes database under Linux.
If someone donates me an iPod, I'll gladly try it out and tell ya if it works. ;)
How do you use a command line interface on an ipod?
If you read the documentation, it says when the iPod boots the Linux kernel it will automatically configure its local ethernet device (ethernet over firewire/IEEE1394) and then starts inetd so that network connections via telnet may be made.
The default configuration is for the iPod to use 192.10.1.2 and to allow telnet connections. The default address may be changed by editing the /etc/rc script.
Why would one run linux on an ipod?
So that you can put an OGG player on there and not have to convert your OGGs whenever you wanna listen to them on the go.
According to the site's FAQ, there's an OGG player that plays at "80% realtime", whatever that means. They expect a running ogg player on there "soon".
Is it allowed to google for a reasonable human response to your statements? [...] in this way (with a large enough db like the web), a chatbot could appear to be human, but we probably wouldn't consider this AI.
I'd categorize it as the synthesis of two previously mostly distinct fields: Distributed AIs.
At the least, I'd like to see something that fails miserably, but attempts to "learn."
The fact that you've put learn in quotation marks hints at the problem of this reasoning. It's that it's very difficult (perhaps impossible) define "learn" (or "intelligence", for that matter). Let's say I write a chatbot that simply reads in the text that you type to it, and stores it in a log. Is that "learning"? Let's say it constructs Markov chains based on the way you type to just regurgitate exactly what you said but with enough variance so that it isn't obvious that it's simply repeating you. Is that learning? I know back in highschool sometimes I'd have to regurgitate something into an essay to get a passing grade from closed-minded teachers.
The point of having these tests is that you can actually say how successful your AI program is in a quantitive manner (e.g. % of humans who were convinced they were speaking to other humans and not chatbots).
I fail to see how fooling humans into thinking that they are having a conversation with another human, when it is really a chatbot, will do anything to produce artificial intelligence. It's an illusion, using technology, nothing more.
Not to put words in your mouth, but I'm assuming the argument you have in mind is something similar to "The computer is not really thinking, it's CPU is just sending signals around to produce output that merely LOOKS intelligent. It's an illusion".
Note though that one could argue that humans are not really thinking, their neurons are just sending signals around to rpdouce output that merely looks intelligent. It's an illusion.
Most people determine the intelligence of things by observing their behaviour. We assume that rocks are not very intelligent because their behaviour doesn't imply some sort of thinking process occuring internally. We assume that chimpanzees have some degree of intelligence because their behaviour DOES imply some sort of thinking process.
1. Get a Nintendo/PS2/XBOX/whatever I have these AND a windows machine, 'cause sometimes they release a game only for Windows.
2. Good software support. Hah. Most Windows programs are monolithic, clunky, closed systems (i.e. you can't extend them, script them, etc.). They may or may not conform to a UI model, and they may or may not even get along with each other. If you think what you're using is good, try OSX or KDE 3.1. You'll be amazed. I like eMule, Office, WinAmp, CloneCD, ICQ, MSN Messenger, Paint Shop Pro, L&H Japanese Translator, Visual Studio, Encarta, DiscJuggler, Reason, RPGMaker and Tag&Rename just ot name a few. I've never really encountered a situaiton where I'd want to script or extend them. You can name me the alternatives (lMule, OpenOffice, XMMS, etc.); I've tried them, and I don't like them as much as the originals. I may have had one or two instances of software not working together on a Windows machine (say Adobe doing some funky script thing in Office), but they tend to be merely annoyances; certainly nothing that would keep me from doing any work. But try installing an mp3 tagger in Debian which depends on something which depends on something which depends on something which depends on a version of libstdc++5 from the unstable branch, and you've pretty much killed a dozen or two of your apps.
3. I've never had Windows not just work (i.e. always, Windows just works), and I've installed it on 5 home systems, 2 systems for friends, and 10 systems in a LAN at work. Networking always worked, and video always worked, so for the few instances that the more obscure hardware didn't, I could just hop onto Windowsupdate.com and get the drivers. With every system I've installed Linux on but one, I found out the hardware was not supported. Things as basic as video, sound and networking would not work. Downloading a driver from a windows machine, splitting it into 1.4 meg chunks using WinRar, floppy-disking it onto the linux box, and downloading a linux unrar program, trying to get the driver to compile using the 20+ command line arguments the README file tells me to input, and being told that a certain option has been deprecated and not being told what the option replaces it is not fun.
Linux is a fine OS, but there are still plenty of reasons to use Windows. I run both.
The first games you play tend to be the ones you remember as being the best because you were just starting out in computer games and everything is still new and novel.
While I can see the logic in that argument, I don't think it's nescessarily a "universally true" statement. First FPS I played, I think, was Wolf3D, then Doom 1/2, ROTT, Quake 1/2, etc. Your standard path. While I thought ROTT was pretty funny, none of them really appealed to me all that much. The first FPS I really liked was Tribes, and the second, I think, was Call of Duty. In between I've played Quake 3, Unreal, Unreal Tournament, Unreal 2, Unreal Tournament 2003, Return to Castle Wolfeinstein, Half Life, and probably a whole bunch more I can't even remember.
By no means were the "first" FPSes I played my favorite, and a lot of my friends agree with me on which FPSes were great and which were just okay. When I think back to the first video games in general I played, I don't think any of those were my favorite either. On the NES, you have Duck Hunt, Super Mario Brothers, Bomberman, Milo's secret castle (or something like that), 8 eyes, etc. On the Sega Master System, you've got Outrun, Kung Fu Hero, Alexx Kidd in Wonderland, etc.
My favorites are like... Dance Dance Revolution, Chrono Trigger, Colin Mcrae Rally 3, Soul Calibur, etc.
So if there's a relationship between the order in which you play games, and the which ones you like the best, I don't think "earlier = better" is it.
3. Multilinear gameplay. THIS is where the "in the movies" feel comes from - where YOU, the GAMER, are picking what the story is.
This may be just a matter of opinion, but pretty much in every game where the burden of writing the story was put on me, the player, rather than on the designers had a pretty crappy story.
One of the best stories in a game, in my opinion, is the one from Chrono Trigger, which, while it has "multiple endings", is pretty linear otherwise.
From a technology point of view, I agree that RealMedia is pretty good; unfortunately, I think the format just got a bad rep because of the spyware loaded player.
I think this is more an issue of programs than codecs. While the article does state what programs it uses for the encoding (VirtualDub in the DivX case, Microsoft Windows Media Encoder in the WMF case, etc.), its main focus was on which of the codecs is the best, not which of the encoding programs is the best.
PC users usually don't toss around Apple marketing terms. They usually go for Intel/AMD marketing terms.
It cannot get any easier to setup a printer than it is in Mandrake. Simply click about 4 buttons and you are done.
Actually, I've had an easier experience on Windows: I plugged the cable into the USB port, and Windows asks me if I want to print a test page. I say yes, and it does, and the page looks fine, and so I was done. That's one button click.
Of course, you could argue that if Microsoft's Plug n Play for some reason failed, the "manual" installation process might take more than 4 clicks, but then again if the Mandrake "easy installation" process failed, the manual version would also probably take more than 4 clicks.
My experience indicates that it's extremely rare for a UNIX tool to lie in its return code value.
I don't think programmers intentionally make their command line programs lie via return codes; rather, it's not always clear what value you should return. The general guideline is that 0 means some sort of success occured, and anything else means some sort of failure.
If the program expects command line arguments, and you provide it none, the tradition is to display a short blurb about the program's usage. Is this considered a successful run, or a failure? You could argue that it failed in that it didn't do what the program was intended to do (e.g. encode mp3s), but you could also argue that the expected behaviour of the program, when provided zero parameters, is to display usage info, and it successfully did that.
What if, using whatever the proper syntax of the shell is, the user requests something like "encode all wave files in this directory" and there are zero wave files, and so the program encodes zero mp3s. Was that a success or a failure? It's a success in that nothing in the program broke, but it seems like perhaps something happened that the user isn't aware of, and perhaps the user should be informed.
What about if the user passes a "-h" parameter, asking for help, and the program displays the help. Was this a successful run of the program or a failure? It did exactly what the user asked, but it didn't fufill its main purpose, encoding mp3s in this example.
What if the user thinks "-h" means "highquality", but in fact the program interprets it as "display help"? The help blurb may be displayed, a success returned, and the files deleted.
I'm not advocating GUIs vs CLI or vice versa (I alternate between the two, depending on what I'm trying to get done), but I have a distate for Unix's return code system.
As a game designer, you wouldn't want to use floating point numbers for hit points, because float point arithmatic suffer from rounding errors.
As an oversimplification, let's say I store 200 billion in my floating point, but because of the precision I use, I only keep track of the 5 most significant digits, i.e. 2.0000 * 10 ^ 11. If someone where to deal one million points of damage, i.e. 0.00001 * 10 ^ 11, the substraction would yield 1.99999 * 10 ^ 11, which then needs to be rounded so that 5 digits are accurate. This number rounds to 2.0000 * 10 ^ 11, so with 5 digits of accuracy, dealing on million points of damage is equivalent as dealing 0 points of damage.
In real life, IEEE double precision gives you about 15 digits of accuracy, which is more digits than there are in "100 billion", so this might work, but I think most game designers will still avoid using FPNs because of these potential rounding errors.
It's not exactly an RPG, though it has RPG-ish element (i.e. you get experience for killing badguys which you can spend on upgrading your skills). I guess the closest equivalent is FF Tactics.
I haven't beaten the game yet, but it looks like it's possible to play the whole game without ever going through a single random encounter.
The way it works is you have a map that shows all the locations of interest, and there's an icon representing your location. You click where you want to go, and your icon pretty much travels in a straight line to where ever you clicked.
Every now and then, red dots appear randomly around the map. If you come into contact with one, then a random battle starts.
However, these dots, while not rare in the temporal sense, are pretty sparse in space. What I mean by that is that there'll always be one or two dots, but the dots are so small compared to the size of the map that if you don't want to get into random battles, it's trivial to walk around them and onto the "plot battles".
The benefit of them being common, temporal wise, is that if you get stuck on a mission, you can get into random battles pretty much at will, and level up your characters. Or maybe just play them for fun (since Silent Storm is very combat oriented, being a FF Tactics type game.)
So far I've voluntarily went through one random battle, and have never gotten into a random battle involuntarily. A friend of mine says he's beaten the game without ever having gotten into a random battle.
I think this is really the "optimal" solution: Let the players who want to have random battles play their random battles, and let players who don't like random battles not go through random battles.
Any potential RPG designers might want to give Silent Storm a try and see if they can adapt it to their designs.