Did you read the article? The author isn't proposing that you have the same gameplay on every platform. Rather, he just says that you have the same gaming universe available everywhere, and the gameplay that's available on each platform will be tailored to that device. Big difference.
Cross-platform developers who actually make an effort to port to GameCube don't run into some of the problems you mention. For instance, taking advantage of GameCube's flexible vertex specification lets you compress the geometry, usually enough to make up for the difference between PS2 and GameCube main memory space. Using S3TC-compressed textures also makes a big difference. If you aren't doing these kinds of things, then you just aren't trying.
Regarding the graphics hardware capabilities, I think it breaks down like this:
-PS2: fastest (but simplest) pixels, decent polygon rate -GameCube: fast simple pixels, fastest complex pixels, slowest polygon rate (but not necessarily by much) -Xbox: slowest pixel rate, fastest polygon rate (assuming you optimize for it)
Of course these are generalizations. You can probably do some polygon computations on the GameCube faster than the other two, assuming what you want to do uniquely maps to the GameCube hardware. The same can be said for either of the other two systems.
I wonder if anyone has investigated if there's a latency involved in using an analog connection to a LCD display versus using a digital connection. The analog connection must go through a A/D step before being displayed.
Of course, depending upon the architecture of the display, there may be no difference. The display may have an internal buffer (ie, a source of latency) that is unavoidable in either case.
But anyway, the point is, there may be reasons other than pixel rise/fall times as to why an LCD may have more latency than a CRT monitor.
I believe that the blue component of OLED panels has only about 1/10th the life of the other color components. Of course, this applies only to RGB raster displays, not necessarily to white-emitting panels. (In the latter case, it doesn't matter if you can't display blue by itself.)
In this context, you need to clarify what you mean by "lasts". Most lightbulbs don't have any serviceable parts, whereas a car has lots of serviceable parts. In fact, cars generally require service to their parts to keep them going. With how many cars can you "just put in gas" and expect them to keep going 100K miles?
It's not entirely "doofey". In a convential LCD panel, the backlight is the most power-hungry part. On the other hand, a practical OLED display is still a ways off, mostly due to the lifetime of the blue-emitting component. However, if, in the mean time, they can make a practical white-emitting panel (that is more efficient than current backlights), then it may serve as a reasonable intermediate step until the OLED display itself is made practical.
Don't you understand that component TV is something completely different than digital HDTV? This card is designed to do the latter. If you look at ALL the HDTV cards for PCs, you'll find that this is true.
If you want component inputs, then you should be shopping for a monitor, or else a transcoder (component to VGA).
If what you want is to record your satellite box's HDTV signal, then you need a firewire interface, along with a satellite box that has a properly-working firewire port. That way, you get the DIGITAL HDTV signal, not the analog crap.
If you truly want a component TV analog-to-digital converter card, then you can find them if you shop around enough. They will not be cheap, and they will require a RAID array in order to store a 1080i analog-converted datastream.
Actually, not quite so. By varying the reflective properties of the spinning screen, you can control the amount of light that is reflected in the "wrong" directions. In the limit, the spinning screen is simply a mirror, and the light being reflected only goes out in exactly the reflection direction, with no angular dispersion at all.
For this particular display, the angular dispersion should be confined to 15 degrees or so. Of course, it is difficult to build a screen with such tightly-controlled reflective properties.
That's not true. One camera that moves around can give you all the data that two cameras can. You can compute the 3D positions of objects from the video by establishing corresponding points between frames. Once you figure out the camera position, you can build a 3D map of the scene.
What's the point? Well, for one you can change the camera position or angle once you've got the scene mapped out. Of course, this may lead to "holes" in the image for places where you have no data.
Another use is simply the 3D capture of the scene in order to make a model to use for another rendering.
Actually, I think that hi-def porn could work out well, once directors adjust to the medium. If they just use the same camera angles as NTSC calls for, then it will look bad. But HDTV allows for wider shots, showing more action in the scene without losing details.
Perhaps for shots that do need close-ups, digital post-production tools would be called in for some smoothing and touch-ups. I'm sure there will be a big market for the "age-cream" and "acne-remover" Adobe plug-ins.
Ebay is definitely the place to go for inexpensive arcade machine parts. Just take note of the categories where such things are found, such as "collectibles/coin-op".
You can take most any trackball from an arcade machine and interface it to a PC using the parts mentioned in other posts. If you're really cheap, you can just hack an old mouse (you just need to wire in the LEDs and optical sensors).
MS isn't getting just an "ATI GPU". They're getting ATI to do a custom GPU for them. It could include features such as an "Nvidia-mode" which understands Nvidia commands and formats directly.
Or, if necessary to skirt legal issues, they could have a mode that understands commands and formats which are very similar to Nvidia's, but only slightly different, such that it would take very little processing to convert them at run time.
It would be easiest to launch from the equator, right? Obviously, I'm not the first person to have thought of this. In fact, there's a company set up to do this: http://www.sea-launch.com
Any modern hard drive that uses fluid bearings runs very quietly. This includes Seagate Barracuda (IV and up) as well as some Maxtors (and probably others). As for the seek noise, I believe some drives have a "quiet seek mode" that can be activated with appropriate software. It trades off a bit of seek speed, of course.
I think that the best solution for quieting your drives is to just get new quiet drives.
I think the "facts" that they were referring to are their corporate power and influence and ability to buy their own decisions, given enough time and a large enough infusion of cash in the right places (in the US, at least).
But it's not the people who live there who decided to build the dam. It's the bureaucrats who decided. So it's still a case of assholes telling other people how to live.
Slashdot Mirror
> could USDTV keep prices low and still support local content since they have no cable network to maintain, and no satellites to launch?
But if they want to go national, they need to build, buy, or lease radio transmitter stations across the country. That won't be cheap.
Did you read the article? The author isn't proposing that you have the same gameplay on every platform. Rather, he just says that you have the same gaming universe available everywhere, and the gameplay that's available on each platform will be tailored to that device. Big difference.
In this in reference to that piece on TechTV (Wired for Sex, I think) about a guy who got the hots over his Mac?
Cross-platform developers who actually make an effort to port to GameCube don't run into some of the problems you mention. For instance, taking advantage of GameCube's flexible vertex specification lets you compress the geometry, usually enough to make up for the difference between PS2 and GameCube main memory space. Using S3TC-compressed textures also makes a big difference. If you aren't doing these kinds of things, then you just aren't trying.
Regarding the graphics hardware capabilities, I think it breaks down like this:
-PS2: fastest (but simplest) pixels, decent polygon rate
-GameCube: fast simple pixels, fastest complex pixels, slowest polygon rate (but not necessarily by much)
-Xbox: slowest pixel rate, fastest polygon rate (assuming you optimize for it)
Of course these are generalizations. You can probably do some polygon computations on the GameCube faster than the other two, assuming what you want to do uniquely maps to the GameCube hardware. The same can be said for either of the other two systems.
I wonder if anyone has investigated if there's a latency involved in using an analog connection to a LCD display versus using a digital connection. The analog connection must go through a A/D step before being displayed.
Of course, depending upon the architecture of the display, there may be no difference. The display may have an internal buffer (ie, a source of latency) that is unavoidable in either case.
But anyway, the point is, there may be reasons other than pixel rise/fall times as to why an LCD may have more latency than a CRT monitor.
I believe that the blue component of OLED panels has only about 1/10th the life of the other color components. Of course, this applies only to RGB raster displays, not necessarily to white-emitting panels. (In the latter case, it doesn't matter if you can't display blue by itself.)
In this context, you need to clarify what you mean by "lasts". Most lightbulbs don't have any serviceable parts, whereas a car has lots of serviceable parts. In fact, cars generally require service to their parts to keep them going. With how many cars can you "just put in gas" and expect them to keep going 100K miles?
It's not entirely "doofey". In a convential LCD panel, the backlight is the most power-hungry part. On the other hand, a practical OLED display is still a ways off, mostly due to the lifetime of the blue-emitting component. However, if, in the mean time, they can make a practical white-emitting panel (that is more efficient than current backlights), then it may serve as a reasonable intermediate step until the OLED display itself is made practical.
What's all this yapping about component inputs?
Don't you understand that component TV is something completely different than digital HDTV? This card is designed to do the latter. If you look at ALL the HDTV cards for PCs, you'll find that this is true.
If you want component inputs, then you should be shopping for a monitor, or else a transcoder (component to VGA).
If what you want is to record your satellite box's HDTV signal, then you need a firewire interface, along with a satellite box that has a properly-working firewire port. That way, you get the DIGITAL HDTV signal, not the analog crap.
If you truly want a component TV analog-to-digital converter card, then you can find them if you shop around enough. They will not be cheap, and they will require a RAID array in order to store a 1080i analog-converted datastream.
Actually, not quite so. By varying the reflective properties of the spinning screen, you can control the amount of light that is reflected in the "wrong" directions. In the limit, the spinning screen is simply a mirror, and the light being reflected only goes out in exactly the reflection direction, with no angular dispersion at all.
For this particular display, the angular dispersion should be confined to 15 degrees or so. Of course, it is difficult to build a screen with such tightly-controlled reflective properties.
That's not true. One camera that moves around can give you all the data that two cameras can. You can compute the 3D positions of objects from the video by establishing corresponding points between frames. Once you figure out the camera position, you can build a 3D map of the scene.
What's the point? Well, for one you can change the camera position or angle once you've got the scene mapped out. Of course, this may lead to "holes" in the image for places where you have no data.
Another use is simply the 3D capture of the scene in order to make a model to use for another rendering.
Actually, I think that hi-def porn could work out well, once directors adjust to the medium. If they just use the same camera angles as NTSC calls for, then it will look bad. But HDTV allows for wider shots, showing more action in the scene without losing details.
Perhaps for shots that do need close-ups, digital post-production tools would be called in for some smoothing and touch-ups. I'm sure there will be a big market for the "age-cream" and "acne-remover" Adobe plug-ins.
Ebay is definitely the place to go for inexpensive arcade machine parts. Just take note of the categories where such things are found, such as "collectibles/coin-op".
You can take most any trackball from an arcade machine and interface it to a PC using the parts mentioned in other posts. If you're really cheap, you can just hack an old mouse (you just need to wire in the LEDs and optical sensors).
MS isn't getting just an "ATI GPU". They're getting ATI to do a custom GPU for them. It could include features such as an "Nvidia-mode" which understands Nvidia commands and formats directly.
Or, if necessary to skirt legal issues, they could have a mode that understands commands and formats which are very similar to Nvidia's, but only slightly different, such that it would take very little processing to convert them at run time.
It would be easiest to launch from the equator, right?
Obviously, I'm not the first person to have thought of this.
In fact, there's a company set up to do this:
http://www.sea-launch.com
From the article:
:-)
> Wires are just one urban challenge. Bedeviled by ballooning rat populations,
> Shanghai has turned not to poison but to rodent contraceptives.
Who gets the job of fitting all the little guys with condoms?
Well, based on the images in the article, I suppose this means that the person at the keyboard isn't wearing any clothers.
I think this mode of play would be much more popular if the on-screen graphic were more accurate.
Any modern hard drive that uses fluid bearings runs very quietly. This includes Seagate Barracuda (IV and up) as well as some Maxtors (and probably others). As for the seek noise, I believe some drives have a "quiet seek mode" that can be activated with appropriate software. It trades off a bit of seek speed, of course.
I think that the best solution for quieting your drives is to just get new quiet drives.
The quietest power supply:
http://www.siliconacoustics.com/silpc.html
and then add a pneumatic hammer or flipper and some armour, plus a bit of AI, and you'll have _real_ Robot Wars.
I can't believe there's not a single reference to Mr. Fusion (Back to the Future) here (that I can find).
We'll just have to correct that...
I think the "facts" that they were referring to are their corporate power and influence and ability to buy their own decisions, given enough time and a large enough infusion of cash in the right places (in the US, at least).
Wasn't their profit for the last quarter something like $2 billion?
When you've got profit margins like that, it pretty much takes the sting out of everything.
I suppose not enough people plan for when their house is suddenly under water, and air somehow becomes scarce.
But it's not the people who live there who decided to build the dam. It's the bureaucrats who decided. So it's still a case of assholes telling other people how to live.