Seems to me that Wii sports wouldn't really shine without a 2nd player. They ought to ship 2 controllers in the box. Then they could even make games that are meant to be played with one controller in each hand.
Yeah, well Joystiq still isn't cagegorized as "Games" by Websense, so it's one of the few decent game news sites I can still visit at work. I wonder how much extortion money Weblogs, Inc. pays Websense every month...
That's a good list. Once someone has used Vista for a while, they will not want to go back. Even though many of the improvements there are pretty small by themselves, the small things really add up. When I use Windows 2000, it's the small things that really bug me, like the fact that Paint doesn't support PNG, or the lack of the "align to grid" option for the desktop icons (Auto Arrange is *not* the same thing).
The improved Explorer and file operations by themselves will be great (though long overdue).
No, I did not. You misunderstand. Autoscrolling scrolls automatically when you are not moving the mouse and it sucks. I am asking for direct control over the scroll position mapped to the mouse Y axis, at a speed which is carefully selected and consistent for every application. If you aren't moving the mouse, it shouldn't be scrolling.
Also I am saying that the mouse wheel should not click, but should scroll continuously (not automatically!). Like the new Mac Mighty Mouse scroll ball, which people rave about after they try it for a while.
Scrollbars should die. They are a terrible interface for scrolling, and they should be replaced with something better. The arrow buttons on the top and bottom are useless, the bars are way too thin as mouse targets on today's high resolution screens, yet they take up two whole sides of whatever you're scrolling, the size of the scroller part is often too small, the changing size of the scroller is often inconsistent, the behavior of releasing the drag when the mouse moves too far away is braindead (only Windows does this thankfully), the pageup/pagedown functionality when clicking on an empty area of the scrollbar is unreliable because it stops working when the scroller reaches the mouse pointer, they are *not* at all intuitive (though people have learned to use them), I could go on and on.
The mouse scroll wheel was a step in the right direction but the implementation sucks there too. It should scroll continuously instead of clicking by lines, it should control the thing under the mouse instead of having its own weird focus rules (again Windows stands alone here), and pushing the wheel button should give scrolling control directly to the mouse Y-axis instead of that weirdo autoscroll thing.
I was never a fan of the trackpoint until I got my X60s and replaced the default convex nub with a larger concave one. It is incredibly obvious to me now that video game control pads should come with trackpoints intead of analog sticks. The fact that the trackpoint hardly moves even when you press it to the maximum gives it a huge advantage over sticks for agile control. It would be especially ideal as a replacement for the PSP's analog thingy, but it would also go great on the Wii controller. Nintendo should jump on this, with their focus on innovative controls.
Of course polygons can be put into KD-trees. But that doesn't allow you to do efficient occlusion culling for a rasterizer. The only way to perform the occlusion culling would be to cast rays from the camera into the scene and only send what they hit to the rasterizer, but then you've just implemented raytracing inefficiently.
Raytracing makes instancing geometry far less expensive because there's no per-polygon cost, making forests and fields of plants much cheaper to render. Consider a tree instanced in the foreground and also far away where it covers five pixels: A rasterizer must render every polygon in the tree in both cases, perhaps thousands, while for the distant tree the raytracer only concerns itself with the five polygons which are actually hit by rays (plus some small overhead for traversing the KD-tree to throw out the rest). You're doing efficient occlusion culling at the *polygon* level, throwing out polygons which are between pixels! The polygons thrown out need never even be loaded into cache. No fancy data structure can perform that for a rasterizer. Oblivion could push its LOD distances out much farther if it was using raytracing for this reason alone.
No official numbers have been released, but speculation is that Google provides a large part of the funds that run the Mozilla Corporation, and therefore pays the salaries of many Mozilla developers. Mozilla Corporation employees work closely with Google; their offices are within walking distance in Mountain View. Some Google employees also work on Mozilla directly.
For simplified problems like 2D terrain or Quake-like levels made of hallways, you can certainly come up with fancy visibility culling algorithms that work well in practice. But I'm talking about fully general triangle soups here. With a raytracer, you just feed the entire scene in (could be paged), and visibility is computed on the fly, no matter what the scene is. For a rasterizer you must write your own domain-specific visibility culling; if you send the entire scene straight to the rasterizer it will choke. That's all I'm trying to say.
In many cases it's not obvious how to write a good visibility culling algorithm, and games shy away from that sort of situation, or end up with incredibly obvious nonoptimal LOD/culling (GTA comes to mind). A raytracer would work better, automatically. It's the same situation as with shadows and reflections; with rasterizers low-quality approximations are a lot of work and with raytracing they work better, automatically. With game companies screaming about how next-gen games are getting harder and harder to make, this sort of improvement is exactly what the industry needs.
Fine, show the license, but don't require us to manually select "I Agree" in order to use the software. Forcing the user to legally agree to pages of tiny text in a non-resizable box is just as asinine when it's the GPL as when it's a Microsoft EULA.
I have also written a raytracer. (hasn't everyone?) I suppose the confusion is sort of a matter of definitions. If the rasterizer is sent an infinite list of triangles to rasterize, it will take an infinitely long time to render them, while if a raytracer is given an infinite number of triangles in a KD-tree, it can finish raytracing because it can skip nodes of the tree.
Of course it would take an infinitely long time to build the KD-tree, this is just a silly example. But even if the rasterizer was given the prebuilt KD-tree, it wouldn't know what to do with it. You have to determine visibility in order to choose which nodes of the KD-tree to send to the rasterizer. Visibility is expensive to compute by itself (Quake engines do it as a preprocessing step during level design), but the raytracer does it automatically as part of the rendering process, so it has an advantage.
The "context-aware" phone described in the article would be a nightmare to use. It's basically got a bunch of different modes (contexts, same thing) that it selects automatically based on what it thinks you're doing. I guarantee it will be wrong all the time, and then you will have to figure out how to manually switch to a different mode to get the option you want. The best user interfaces *reduce* the number of possible modes of the system and strive for predictability. Automatic switching between hundreds of modes is the opposite of a good user interface.
If the number of triangles is infinite a rasterizer will never finish, because it has to consider each triangle in the scene at least once. Theoretically, a raytracer running on an infinitely large scene could finish. Raytracers use fancy data structures to avoid even needing to look at huge parts of the scene if they don't intersect any rays, which provides a sort of automatic fine-grained culling that gives them an advantage, not just in infinite scenes, but in real-life scenes which games are fast approaching in complexity.
Ray tracing also tends to do specular well, but isn't as good at diffuse
It's certainly not any worse than rasterization; rasterization hardly does diffuse lighting at all. You have to resort to hacks like ambient lights, lightmaps, etc, which all can be implemented in a raytracer in exactly the same way. But with raytracing you have the ability to use more advanced techniques which can be a lot better.
Why would you do that when CPU cores are just about to become plentiful? I think multi-core computing should be the death of dedicated acceleration cards. When you have all these cores just lying around you don't need a dedicated sound DSP, network accelerator, or even (eventually) a video accelerator. On, say, a 64-core system, you could do real-time raytracing without any specialized hardware at all, and still dedicate a core each to network and sound if you wanted.
This kind of general-purpose architecture is far more flexible and in the end faster than dedicated accelerators.
I think he should edit the English Wikipedia. If he makes some minor grammar mistakes, but increases the useful information content of an article, I think that's a win for Wikipedia. The number of people who can correct his grammar is likely larger than the number of people who could have added the information he knew.
It's been a long time since I was stuck in a menu system and unable to talk to a person. Just press zero, or simply wait. Many even tell you this. For Cingular's, you can even say out loud "I want to talk to a person" and the computer responds "Okay!" and connects you.
Sometimes, though, the automated system is just what you need. UPS has a great system; when you miss a delivery they leave you a note with a confirmation number you can read to their system and get the status of your package, and even schedule when to pick it up. I particularly like the "beep beep boop" computery sounds it plays while it is "thinking," just to let you know it is still on the line.
Wow, it's obvious you didn't read the article very closely. The 50-layer thing has nothing to do with virtualization; that's Microsoft's *current* (failing) effort to rescue the Windows codebase from collapse under its own weight, without resorting to virtualization.
What Gartner is saying (and they didn't make this up but are parroting smarter people who suggested it first) is that Microsoft should ditch the backwards compatibility that is hobbling them and start over from scratch with a new OS. (I'd prefer something based on Singularity but that's a long shot.) Virtualization then provides the means for backwards compatibility; simply virtualize Vista in the background to run all your legacy apps and drivers. But this virtualization doesn't have to work like VMWare or Virtual PC today; Microsoft can produce a version of Vista that integrates seamlessly with the host OS when run under virtualization. This is what they're talking about with "integrating data across partitions" and "creating a consistent user experience".
Microsoft has resisted this path, claiming that it's impossible. But Apple has already proved it can work, with Mac OS Classic in OS X. That's what we're talking about here: Microsoft needs to pull an OS X and rewrite their OS from the ground up, then produce "Vista classic" for backward compatibility.
The assault rifle was more fun than either the SMG or the battle rifle. It looked and sounded cooler. The battle rifle is annoying; it feels like it takes forever to do any damage with it.
They're not malicious, just incompetent. In my case when I called them they happily sent out a check for the prorated amount. In the meantime I got several bills from their computers requesting payment of $-47.55, but it all worked out in the end.
The only thing about the 360's blade interface is that you have to wait for the fancy animations to finish before you can use the menus. It's only a second or so of delay but it adds up every time you do something and gets annoying fast. Construct the menus first, *then* do the fancy animations. Also, the Live Marketplace menus leave a lot to be desired. They're pretty minimal (where are the screenshots?), and the snail's-pace scrolling of the description text in its needlessly tiny box is inexcusable.
I'm not sure about the Sony menus. For $600 I'd expect a little more pizazz in my UI. But I'll take usability over pizazz, if they can deliver it.
Labelling was a huge setback for food irradiation, though it can help prevent food poisoning. Maybe labelling is not a bad idea, but it puts the food industry at the whim of those knee-jerk fearmongering articles the media loves to spew on the most specious evidence. And as others have pointed out, labelling every single thing done to a product would require a book, which, aside from being impractical, would *discourage* reading of labels.
We must pick and choose which things are important enough to require labelling, and which are simply not relevant. And we should choose based on scientific evidence, not by the latest media fad scare.
Did you even read what I said? Of course MS would be finished if they misrepresented the contents of a critical update that blatantly. Even without lawsuits, the backlash from their customers would be incredible, and nowhere do I dispute this. But that doesn't change the fact that the power exists, and it could be used. And the reality is that if you use Microsoft products, you give Microsoft this power (by installing their opaque binary updates), or your computer is hacked tomorrow. Not much of a "choice". Microsoft is certainly not unique in this regard, they are simply the widest-deployed.
It's not FUD, it's a fact. You can choose not to install critical updates, but if you don't your security is forfeit. You may have declined WGA, but did you decline the patches for the latest remote code execution exploits?
Furthermore, whether or not you choose to install critical updates, there are enough people who do to give Microsoft control over a significant fraction of the world economy at the push of a button. I'm not insinuating that Microsoft would ever be stupid enough to use that power themselves, as it would of course be instant corporate suicide. But the power exists.
Then make sure to stop using your cable box, your cell phone, your game console connected to an online service, and your PC running Windows.
To me the scariest of those is Windows. Microsoft has total control at a moment's notice of the large percentage of machines worldwide with automatic updates enabled, and the rest could be compromised with a trojan in a manually-installable critical update. Can you imagine the chaos if world's Windows machines erased their hard drives tomorrow? Not that Microsoft would ever intentionally do that, but still, that is a lot of power held up there in Redmond. It could be used militarily if it were seized by the government.
Slashdot Mirror
Seems to me that Wii sports wouldn't really shine without a 2nd player. They ought to ship 2 controllers in the box. Then they could even make games that are meant to be played with one controller in each hand.
Yeah, well Joystiq still isn't cagegorized as "Games" by Websense, so it's one of the few decent game news sites I can still visit at work. I wonder how much extortion money Weblogs, Inc. pays Websense every month...
That's a good list. Once someone has used Vista for a while, they will not want to go back. Even though many of the improvements there are pretty small by themselves, the small things really add up. When I use Windows 2000, it's the small things that really bug me, like the fact that Paint doesn't support PNG, or the lack of the "align to grid" option for the desktop icons (Auto Arrange is *not* the same thing).
The improved Explorer and file operations by themselves will be great (though long overdue).
Also I am saying that the mouse wheel should not click, but should scroll continuously (not automatically!). Like the new Mac Mighty Mouse scroll ball, which people rave about after they try it for a while.
Scrollbars should die. They are a terrible interface for scrolling, and they should be replaced with something better. The arrow buttons on the top and bottom are useless, the bars are way too thin as mouse targets on today's high resolution screens, yet they take up two whole sides of whatever you're scrolling, the size of the scroller part is often too small, the changing size of the scroller is often inconsistent, the behavior of releasing the drag when the mouse moves too far away is braindead (only Windows does this thankfully), the pageup/pagedown functionality when clicking on an empty area of the scrollbar is unreliable because it stops working when the scroller reaches the mouse pointer, they are *not* at all intuitive (though people have learned to use them), I could go on and on.
The mouse scroll wheel was a step in the right direction but the implementation sucks there too. It should scroll continuously instead of clicking by lines, it should control the thing under the mouse instead of having its own weird focus rules (again Windows stands alone here), and pushing the wheel button should give scrolling control directly to the mouse Y-axis instead of that weirdo autoscroll thing.
I was never a fan of the trackpoint until I got my X60s and replaced the default convex nub with a larger concave one. It is incredibly obvious to me now that video game control pads should come with trackpoints intead of analog sticks. The fact that the trackpoint hardly moves even when you press it to the maximum gives it a huge advantage over sticks for agile control. It would be especially ideal as a replacement for the PSP's analog thingy, but it would also go great on the Wii controller. Nintendo should jump on this, with their focus on innovative controls.
Of course polygons can be put into KD-trees. But that doesn't allow you to do efficient occlusion culling for a rasterizer. The only way to perform the occlusion culling would be to cast rays from the camera into the scene and only send what they hit to the rasterizer, but then you've just implemented raytracing inefficiently.
Raytracing makes instancing geometry far less expensive because there's no per-polygon cost, making forests and fields of plants much cheaper to render. Consider a tree instanced in the foreground and also far away where it covers five pixels: A rasterizer must render every polygon in the tree in both cases, perhaps thousands, while for the distant tree the raytracer only concerns itself with the five polygons which are actually hit by rays (plus some small overhead for traversing the KD-tree to throw out the rest). You're doing efficient occlusion culling at the *polygon* level, throwing out polygons which are between pixels! The polygons thrown out need never even be loaded into cache. No fancy data structure can perform that for a rasterizer. Oblivion could push its LOD distances out much farther if it was using raytracing for this reason alone.
No official numbers have been released, but speculation is that Google provides a large part of the funds that run the Mozilla Corporation, and therefore pays the salaries of many Mozilla developers. Mozilla Corporation employees work closely with Google; their offices are within walking distance in Mountain View. Some Google employees also work on Mozilla directly.
For simplified problems like 2D terrain or Quake-like levels made of hallways, you can certainly come up with fancy visibility culling algorithms that work well in practice. But I'm talking about fully general triangle soups here. With a raytracer, you just feed the entire scene in (could be paged), and visibility is computed on the fly, no matter what the scene is. For a rasterizer you must write your own domain-specific visibility culling; if you send the entire scene straight to the rasterizer it will choke. That's all I'm trying to say.
In many cases it's not obvious how to write a good visibility culling algorithm, and games shy away from that sort of situation, or end up with incredibly obvious nonoptimal LOD/culling (GTA comes to mind). A raytracer would work better, automatically. It's the same situation as with shadows and reflections; with rasterizers low-quality approximations are a lot of work and with raytracing they work better, automatically. With game companies screaming about how next-gen games are getting harder and harder to make, this sort of improvement is exactly what the industry needs.
Fine, show the license, but don't require us to manually select "I Agree" in order to use the software. Forcing the user to legally agree to pages of tiny text in a non-resizable box is just as asinine when it's the GPL as when it's a Microsoft EULA.
I have also written a raytracer. (hasn't everyone?) I suppose the confusion is sort of a matter of definitions. If the rasterizer is sent an infinite list of triangles to rasterize, it will take an infinitely long time to render them, while if a raytracer is given an infinite number of triangles in a KD-tree, it can finish raytracing because it can skip nodes of the tree.
Of course it would take an infinitely long time to build the KD-tree, this is just a silly example. But even if the rasterizer was given the prebuilt KD-tree, it wouldn't know what to do with it. You have to determine visibility in order to choose which nodes of the KD-tree to send to the rasterizer. Visibility is expensive to compute by itself (Quake engines do it as a preprocessing step during level design), but the raytracer does it automatically as part of the rendering process, so it has an advantage.
The "context-aware" phone described in the article would be a nightmare to use. It's basically got a bunch of different modes (contexts, same thing) that it selects automatically based on what it thinks you're doing. I guarantee it will be wrong all the time, and then you will have to figure out how to manually switch to a different mode to get the option you want. The best user interfaces *reduce* the number of possible modes of the system and strive for predictability. Automatic switching between hundreds of modes is the opposite of a good user interface.
It's certainly not any worse than rasterization; rasterization hardly does diffuse lighting at all. You have to resort to hacks like ambient lights, lightmaps, etc, which all can be implemented in a raytracer in exactly the same way. But with raytracing you have the ability to use more advanced techniques which can be a lot better.
Why would you do that when CPU cores are just about to become plentiful? I think multi-core computing should be the death of dedicated acceleration cards. When you have all these cores just lying around you don't need a dedicated sound DSP, network accelerator, or even (eventually) a video accelerator. On, say, a 64-core system, you could do real-time raytracing without any specialized hardware at all, and still dedicate a core each to network and sound if you wanted.
This kind of general-purpose architecture is far more flexible and in the end faster than dedicated accelerators.
I think he should edit the English Wikipedia. If he makes some minor grammar mistakes, but increases the useful information content of an article, I think that's a win for Wikipedia. The number of people who can correct his grammar is likely larger than the number of people who could have added the information he knew.
It's been a long time since I was stuck in a menu system and unable to talk to a person. Just press zero, or simply wait. Many even tell you this. For Cingular's, you can even say out loud "I want to talk to a person" and the computer responds "Okay!" and connects you.
Sometimes, though, the automated system is just what you need. UPS has a great system; when you miss a delivery they leave you a note with a confirmation number you can read to their system and get the status of your package, and even schedule when to pick it up. I particularly like the "beep beep boop" computery sounds it plays while it is "thinking," just to let you know it is still on the line.
Wow, it's obvious you didn't read the article very closely. The 50-layer thing has nothing to do with virtualization; that's Microsoft's *current* (failing) effort to rescue the Windows codebase from collapse under its own weight, without resorting to virtualization.
What Gartner is saying (and they didn't make this up but are parroting smarter people who suggested it first) is that Microsoft should ditch the backwards compatibility that is hobbling them and start over from scratch with a new OS. (I'd prefer something based on Singularity but that's a long shot.) Virtualization then provides the means for backwards compatibility; simply virtualize Vista in the background to run all your legacy apps and drivers. But this virtualization doesn't have to work like VMWare or Virtual PC today; Microsoft can produce a version of Vista that integrates seamlessly with the host OS when run under virtualization. This is what they're talking about with "integrating data across partitions" and "creating a consistent user experience".
Microsoft has resisted this path, claiming that it's impossible. But Apple has already proved it can work, with Mac OS Classic in OS X. That's what we're talking about here: Microsoft needs to pull an OS X and rewrite their OS from the ground up, then produce "Vista classic" for backward compatibility.
The assault rifle was more fun than either the SMG or the battle rifle. It looked and sounded cooler. The battle rifle is annoying; it feels like it takes forever to do any damage with it.
They're not malicious, just incompetent. In my case when I called them they happily sent out a check for the prorated amount. In the meantime I got several bills from their computers requesting payment of $-47.55, but it all worked out in the end.
The only thing about the 360's blade interface is that you have to wait for the fancy animations to finish before you can use the menus. It's only a second or so of delay but it adds up every time you do something and gets annoying fast. Construct the menus first, *then* do the fancy animations. Also, the Live Marketplace menus leave a lot to be desired. They're pretty minimal (where are the screenshots?), and the snail's-pace scrolling of the description text in its needlessly tiny box is inexcusable.
I'm not sure about the Sony menus. For $600 I'd expect a little more pizazz in my UI. But I'll take usability over pizazz, if they can deliver it.
Labelling was a huge setback for food irradiation, though it can help prevent food poisoning. Maybe labelling is not a bad idea, but it puts the food industry at the whim of those knee-jerk fearmongering articles the media loves to spew on the most specious evidence. And as others have pointed out, labelling every single thing done to a product would require a book, which, aside from being impractical, would *discourage* reading of labels.
We must pick and choose which things are important enough to require labelling, and which are simply not relevant. And we should choose based on scientific evidence, not by the latest media fad scare.
Did you even read what I said? Of course MS would be finished if they misrepresented the contents of a critical update that blatantly. Even without lawsuits, the backlash from their customers would be incredible, and nowhere do I dispute this. But that doesn't change the fact that the power exists, and it could be used. And the reality is that if you use Microsoft products, you give Microsoft this power (by installing their opaque binary updates), or your computer is hacked tomorrow. Not much of a "choice". Microsoft is certainly not unique in this regard, they are simply the widest-deployed.
It's not FUD, it's a fact. You can choose not to install critical updates, but if you don't your security is forfeit. You may have declined WGA, but did you decline the patches for the latest remote code execution exploits?
Furthermore, whether or not you choose to install critical updates, there are enough people who do to give Microsoft control over a significant fraction of the world economy at the push of a button. I'm not insinuating that Microsoft would ever be stupid enough to use that power themselves, as it would of course be instant corporate suicide. But the power exists.
Then make sure to stop using your cable box, your cell phone, your game console connected to an online service, and your PC running Windows.
To me the scariest of those is Windows. Microsoft has total control at a moment's notice of the large percentage of machines worldwide with automatic updates enabled, and the rest could be compromised with a trojan in a manually-installable critical update. Can you imagine the chaos if world's Windows machines erased their hard drives tomorrow? Not that Microsoft would ever intentionally do that, but still, that is a lot of power held up there in Redmond. It could be used militarily if it were seized by the government.
Carl Bialik from the WSJ posts stories all the time. He may never post comments, but perhaps he has a personal account for that.