Padding makes clicking on the buttons faster, as explained by Fitts's law. I don't want my usability compromised because some people are using impractically small screens.
If the physics and rendering rate are higher than the refresh rate, you'll get even lower latency in exchange for minor temporal distortion on vertical eye movement and tearing artifacts. Yes, this is very wasteful, but I'm not aware of any game engine that splits the screen into horizontal slices (ideally single lines) and renders them just before the raster hits them, so the wasteful method is the only option.
Some game engines with fixed physics rate allow camera rotation synced with the rendering, and physics rate can often be increased.
Motion blur is not an acceptable solution, because real motion blur is modified by eye movement. Artificial motion blur will not look realistic because it will be present even when tracking motion with eye movement. It will produce an artifact very similar to the sample and hold blurring present seen on LCDs. Even measuring eye position can't solve this problem -- there is no getting round the fact that fast motion requires a fast frame rate to avoid artifacts.
100fps physics with 30fps rendering will be more playable than 30fps physics with 30fps rendering, but nowhere near as good as 100fps physics with 100fps rendering. Find a good CRT and test it for yourself. 30fps is nowhere near enough to adequately represent fast motion.
In the worst case, the bottom of the screen could be nearly a whole frame delayed compared to no vsync.
Vsync is also problematic because if the frame drawing time ever goes even slightly above the screen refresh time, frame rate drops by half. Ideally this shouldn't happen, but in practice it's not worth the extra wasted CPU/GPU time to guarantee it never will.
Vsync does add latency, because you have to wait for a whole frame to be drawn before you can show it. Without vsync you can display partial frames (tearing). This means the image is inconsistent as a snapshot of a single moment, but because your eyes are only focused on a small part of the screen at a time this doesn't matter, and you can learn to ignore the tearing.
Note that to test high frame rates fairly you'll need a monitor that can be synchronized refresh rate to frame rate (such as a good CRT) or you could be noticing the differences in tearing or judder rather than the differences in frame rate.
You can easily tell the difference between 100fps and 10000fps by looking at high contrast fast motion. Human eyes don't see in frames, but the point where increasing framerate won't cause any perceptible difference is probably in the thousands of fps.
Whatever temporal sampling rate you choose, it's unlikely to be fast enough
There is no practical frame rate high enough to properly portray all the motion typically encountered. It is necessary to pick a sensible rate that is slow enough to allow the video signal to be stored, routed around, and of course broadcast.
The number of single direction swipe gestures you can keep distinct is limited. I use only 4 (cardinal directions) for perfect reliability without having to pay attention to mousing form. And while fast, these gestures are still slower than clicking.
It doesn't take long to move the mouse to the navigation buttons, but it's still an additional task getting in the way of what you want to do. With side buttons you can instantly go forward or back without any disruption to your train of thought.
X-Com is much less fun once you've memorized all enemy routes and spawn points and know how to exploit the AI. Multiplayer is always better, and multiplayer is more fun when you don't have to wait for the other player to finish his turn.
Addiction is about more than just the drug. Addiction causes real harm, but so does prohibition. Without prohibition we are free to address the underlying causes of drug addiction.
Swappiness 0 is not the same as no swap. It will still swap if it really needs to, and you won't get annoying delays when things were swapped out unnecessarily.
It might save 60 seconds, but it's saving the wrong 60 seconds. I'm not going to notice everything being very slightly faster, but I'll notice Firefox being swapped back from disc. I only care how long something takes if I have to wait for it.
Kernel developers seem to mostly care about benchmarks, and interactive latency is hard to benchmark. This leads to crazy things like Andrew Morton claiming to run swappiness 100 (swappiness 0 is the only acceptable value IMO if you need swap at all). I don't use swap, and with 4GB ram I never need it.
Because people build hardware that's "good enough" for the majority, and anything better becomes ridiculously expensive or not not available at all. This is why all LCD monitors are still locked at 60Hz, even though humans are capable of seeing the difference up to at least 200Hz. When I played FPSs I'd use 800x600 at 160Hz, which made a big difference to tracking very fast motion, giving me a competitive advantage. For normal desktop use I run 1600x1200@100Hz, and the difference in mouse movement smoothness compared to an LCD is easily noticeable. The distance the mouse cursor jumps between frames is much shorter, so it's easier for the eye to track it. This is especially important with high mouse sensitivity/acceleration. I can also play 720p video at 120Hz, for zero pulldown judder with all common framerates (in practice there is a small amount, because I don't know of any Linux video player that will use the graphics card as the timing source, so I can't get the video perfectly synchronized).
All these advantages are important enough to me that I still use the CRT, but they would not be immediately obvious to the untrained viewer.
Less important people are *more* likely to be genuinely harassed in this way, as people are much less likely to believe them. Research MK-ULTRA. I see no reason to believe that governments no longer do such things.
As somebody who's experienced this effect on mushrooms, I can say that even very high dose cannabis will not take you to the same place. There are some similarities, as cannabis is a mild psychedelic, but the cannabis experience is always grounded in physical reality. The "overview effect" is not a "nice place", but not a bad place either. It's a place where such concepts are meaningless. There's no good way to explain it to somebody who has not experienced it.
Lesser known actors will work for cheap/free, cameras are getting cheaper all the time, better cameras will work with natural lighting, CPU time for rendering is dropping in price even faster than the cameras, less time is needed on CGI modeling when you can freely reuse existing models, better 3d tools are improving productivity. Real props aren't getting any cheaper, but they're usually a minor portion of the budget, and rapid fabrication systems could cut costs here in the future.
I've once scanned a YUDEN000T02 DVD+R with *zero* parity inner failures, so the outer layer of error correction isn't used at all. Best media I've ever seen.
The low mass is an advantage for accurate sound reproduction, but these speakers are impractical in most cases. I expect the nanotube speaker will have similar characteristics.
Slashdot Mirror
Padding makes clicking on the buttons faster, as explained by Fitts's law. I don't want my usability compromised because some people are using impractically small screens.
If the physics and rendering rate are higher than the refresh rate, you'll get even lower latency in exchange for minor temporal distortion on vertical eye movement and tearing artifacts. Yes, this is very wasteful, but I'm not aware of any game engine that splits the screen into horizontal slices (ideally single lines) and renders them just before the raster hits them, so the wasteful method is the only option.
Some game engines with fixed physics rate allow camera rotation synced with the rendering, and physics rate can often be increased.
Motion blur is not an acceptable solution, because real motion blur is modified by eye movement. Artificial motion blur will not look realistic because it will be present even when tracking motion with eye movement. It will produce an artifact very similar to the sample and hold blurring present seen on LCDs. Even measuring eye position can't solve this problem -- there is no getting round the fact that fast motion requires a fast frame rate to avoid artifacts.
100fps physics with 30fps rendering will be more playable than 30fps physics with 30fps rendering, but nowhere near as good as 100fps physics with 100fps rendering. Find a good CRT and test it for yourself. 30fps is nowhere near enough to adequately represent fast motion.
In the worst case, the bottom of the screen could be nearly a whole frame delayed compared to no vsync.
Vsync is also problematic because if the frame drawing time ever goes even slightly above the screen refresh time, frame rate drops by half. Ideally this shouldn't happen, but in practice it's not worth the extra wasted CPU/GPU time to guarantee it never will.
I use a CRT, at 120Hz.
Vsync does add latency, because you have to wait for a whole frame to be drawn before you can show it. Without vsync you can display partial frames (tearing). This means the image is inconsistent as a snapshot of a single moment, but because your eyes are only focused on a small part of the screen at a time this doesn't matter, and you can learn to ignore the tearing.
Note that to test high frame rates fairly you'll need a monitor that can be synchronized refresh rate to frame rate (such as a good CRT) or you could be noticing the differences in tearing or judder rather than the differences in frame rate.
Virtually all LCDs are locked to 60Hz refresh rate, and will drop or blend frames if you try to input higher frame rates.
It depends on how much tearing bothers you. I'd rather waste CPU time and tolerate the tearing to get the minimum latency possible.
You can easily tell the difference between 100fps and 10000fps by looking at high contrast fast motion. Human eyes don't see in frames, but the point where increasing framerate won't cause any perceptible difference is probably in the thousands of fps.
Here's a good explanation of the issues of motion reproduction:
http://www.microsoft.com/whdc/archive/TempRate.mspx
The number of single direction swipe gestures you can keep distinct is limited. I use only 4 (cardinal directions) for perfect reliability without having to pay attention to mousing form. And while fast, these gestures are still slower than clicking.
It doesn't take long to move the mouse to the navigation buttons, but it's still an additional task getting in the way of what you want to do. With side buttons you can instantly go forward or back without any disruption to your train of thought.
Under $10 - no good mice exist Under $20 - MS Intellimouse Optical 1.1 Under $30 - MS Intellimouse Optical 1.1
X-Com is much less fun once you've memorized all enemy routes and spawn points and know how to exploit the AI. Multiplayer is always better, and multiplayer is more fun when you don't have to wait for the other player to finish his turn.
http://en.wikipedia.org/wiki/Rat_park
Addiction is about more than just the drug. Addiction causes real harm, but so does prohibition. Without prohibition we are free to address the underlying causes of drug addiction.
Swappiness 0 is not the same as no swap. It will still swap if it really needs to, and you won't get annoying delays when things were swapped out unnecessarily.
It might save 60 seconds, but it's saving the wrong 60 seconds. I'm not going to notice everything being very slightly faster, but I'll notice Firefox being swapped back from disc. I only care how long something takes if I have to wait for it.
Kernel developers seem to mostly care about benchmarks, and interactive latency is hard to benchmark. This leads to crazy things like Andrew Morton claiming to run swappiness 100 (swappiness 0 is the only acceptable value IMO if you need swap at all). I don't use swap, and with 4GB ram I never need it.
Because people build hardware that's "good enough" for the majority, and anything better becomes ridiculously expensive or not not available at all. This is why all LCD monitors are still locked at 60Hz, even though humans are capable of seeing the difference up to at least 200Hz. When I played FPSs I'd use 800x600 at 160Hz, which made a big difference to tracking very fast motion, giving me a competitive advantage. For normal desktop use I run 1600x1200@100Hz, and the difference in mouse movement smoothness compared to an LCD is easily noticeable. The distance the mouse cursor jumps between frames is much shorter, so it's easier for the eye to track it. This is especially important with high mouse sensitivity/acceleration. I can also play 720p video at 120Hz, for zero pulldown judder with all common framerates (in practice there is a small amount, because I don't know of any Linux video player that will use the graphics card as the timing source, so I can't get the video perfectly synchronized). All these advantages are important enough to me that I still use the CRT, but they would not be immediately obvious to the untrained viewer.
Less important people are *more* likely to be genuinely harassed in this way, as people are much less likely to believe them. Research MK-ULTRA. I see no reason to believe that governments no longer do such things.
As somebody who's experienced this effect on mushrooms, I can say that even very high dose cannabis will not take you to the same place. There are some similarities, as cannabis is a mild psychedelic, but the cannabis experience is always grounded in physical reality. The "overview effect" is not a "nice place", but not a bad place either. It's a place where such concepts are meaningless. There's no good way to explain it to somebody who has not experienced it.
The optics are the limiting factor here. Increasing pixel count wouldn't add any more detail.
Lesser known actors will work for cheap/free, cameras are getting cheaper all the time, better cameras will work with natural lighting, CPU time for rendering is dropping in price even faster than the cameras, less time is needed on CGI modeling when you can freely reuse existing models, better 3d tools are improving productivity. Real props aren't getting any cheaper, but they're usually a minor portion of the budget, and rapid fabrication systems could cut costs here in the future.
I've once scanned a YUDEN000T02 DVD+R with *zero* parity inner failures, so the outer layer of error correction isn't used at all. Best media I've ever seen.
Thin film speakers already exist, although the film is held between rigid electrodes:
http://en.wikipedia.org/wiki/Electrostatic_speaker
The low mass is an advantage for accurate sound reproduction, but these speakers are impractical in most cases. I expect the nanotube speaker will have similar characteristics.
How can I copy farms?