and because of some fundamental results in computer science, “I know for a fact that he’ll never be able to get it exactly right.” But, Weimer says, “The vast majority of infinite loops he will be able to figure out.”
Does this we can't get above a certain percentage of correct predictions, or does it mean that more and more work will converge towards 100% (99.5, 99.9%, 99.999% etc. etc.)
Re:3D - and Resolution Maxed-Out?
on
Beyond HDTV
·
· Score: 1
Fields per second? Is that frames per second?
We need more fps anyway for lots of reasons (less/no flicker, smoother motion, less/no blurring), so let's just go with that.
Again though, the speaker itself should remain neutral and the composer can still make the original waveform to have more bass if they wish that's how it should be heard, and if the listener is STILL not happy, he can still adjust the bass control to their liking if they don't like the composer's intentions. Heck, the listener can have the bass knob permanently switched to 100% if they really like bass.
My point is that the default position should give a standard sound output across all speakers. Otherwise, it's just a loudness/bassness war game all over again.
By the way, regarding the fluorescent/LED light, have you ever considered it's not the whiteness of the LED lamp they're moaning about, but rather the abnormal spectral spikes (particularly green) in them which can distort colors? How often do you hear someone moan about sunlight? That's much whiter than incandescents.
Yes. If a tune requires more bass, then that should be the job of the original signal, not the speakers. All speakers should ideally sound the same in this case (one can always alter the treble/bass with knobs on the side). There should be something like an SRGB standard for audio.
A speaker than hasn't any bias towards low or high end will emulate anything perfectly. However, a speaker that biases towards low/high end is like a monitor always having a red tint (even for what's meant to blue or green colours), and people saying "I like the warmer rich red tint of this monitor". It's completely ridiculous.
It reminds of the compression/loudness/over-clipped wars for music actually.
I remember being almost 'very sure' about something to do with whether I could spot a little more latency in game with a certain option. However, I tried testing myself even though I thought it was almost a waste of time. I was surprised to see that the semi-blind test confounded me. Definitely can be a case of bias if one's not careful.
Hi I'm someone else. Have you tried blind listening tests on yourself?
If what you say is true, it would be nice to have something even better than applause or heavy percussion to REALLY make the imperfecctions shine in an MP3. Are you saying the same happens with an 88khz sampling frequency too?
What you actually need to do is link to a sample - an uncompressed sound that when compressed shows MP3/OGG etc. really at its weakest. A sound that one can universally use to show the artifacts that frequency domain techniques exhibit without the need for even a blind test. A sound so obviously bad in MP3, that applause or heavy percussion sounds wonderful in comparison.
I'm sure synthesizing the sound and using GA/GP evolving (with human raters to measure the badness) would result in something wonderfully terrible. We can then use that sound as an acid test (or at least one of the acid tests) for future reference and cut through the subjectivity and misinformation which is so rife in this subject.
It's embarrassing because it's almost *too* efficient. The term first came about when CPU manufacturers and the industry in general were embarrassed at how much faster the GPU was for certain algorithms (i.e. the embarrassingly parallel ones). Programmers also were generally embarrassed at not using the technology sooner, and they often spent YEARS writing efficient code for the CPU, only to have a 5 minute knock-up code job on the GPU beat it when they finally experimented with the GPU.
The definition was then further reinforced by programmers who were expected to write long convoluted code to show their managers. They were then embarrassed because the stuff was good yet very quick to write (parallel algorithms are by their nature more short and elegant), so it looked like they were being lazy.
For a 'normal' answer, maybe try: We can better see the proportion of water in the universe, and therefore indirectly deduce other stuff, perhaps say, a more accurate origin of the universe (which in turn will help us unify physics and where the universe is heading, which in turn will give us better technology for other things, which in turn will make us use said technology to stimulate our 5 senses, which in turn will make us happier, and that's where we reach the end of the line, as happiness *really is* the ultimate goal).
Also, thanks to this, perhaps we will now know where to look, or how to look more easily for water somewhat closer to us (where in the future, we may have a chance of reaching).
Yes I know what you mean, but often they don't just adjust the 'gamma' of the sound like you describe, but also truncate the floor and ceiling to make a more square-wave like sound. That's what I meant when I said 'compression' (I meant clipping), and many CDs do just that. They also do what you said of course.
I hate the compression ramifications as much as anyone. However, it does make the sound more like a square wave, which is how the C64 SID chip sounds (I'm also a bit of a chip fan). You get particular harmonic overtones which help give richness to a sound. So maybe people are really responding to *that*.
Obviously the producers are going about it all wrong though, since they get the 'square wave' style, but also lose detail with most of the instruments.
I understand Apple's philosophy of only allowing the best apps in their istore. However, what they don't realise is that exactly the same can be achieved with decent rating system if it's done relatively well (Slashdot, or download.com though not perfect, spring to mind). That way, people can choose the best, and we don't have a mild form of censorship. It also cuts out the maintenance.
Good, the more ridiculous the patents get, the quicker something will be done to fix the mess. Personally, I'd like to see this patent granted, and dozens of companies ordered to pay lots of damages to the angelic company that is Kootol.....if only to see the backlash from a thousand juggernauts against the current patent system;)
Wonderful post. I love your 2nd last semi-parseable sentence with 'every', every other word:)
I would only add that developers should pay our saviour, yes, but also be locked in so that if they develop for Windows, they can't develop that software for other OSs, ever. After all, what would be the need?
Or how about we all move to a cross based screw tip (open at ends) which, providing a cross-shaped screwdriver is used, has the following advantages over the standard flat/slot-head or the phillips:
1: It's cross compatible with the old flathead. 2: With a cross-shaped screwdriver, it won't slip out from either side, unlike the flathead. 3: A large degree of tolerance is allowed for the size of the cross-screwdriver, more so than the phillips. 4: It's stronger, and will therefore distort the metal (both of the screw and driver) less than a standard flat-head. 5: It's easier to fit the cross in than the flathead and almost as easy as the phillips.
So shock, it's not a zero-sum game, since we have the almost best of both worlds. The ease of use, and non-slippability of the phillips, and the strength and the longevity of the flat head. This hardline relativism sucks - some tools really ARE better. And that's coming from somebody who's never really thought about using these type of screws before, let alone having them.
Granted, it won't be the perfect screw for all occasions, but it'll suit most, and can make many other types redundant. I'm afraid that means not each tool is a special snowflake which is immune to the fate of obsolescence.
I see the pattern now. In another decade, Google will become passée, but the Chrome browser will have become dominant by that point. From here, the new giant corporate of that future era will build an OS *inside* the Chrome browser (itself running on Linux or Windows). A decade after that, someone will then build a browser inside of THAT.
In a couple of centuries, we will be stuck with a giant crumbling 20 layer behemoth with the top layer inheriting all of the bugs of the previous generations. "Hello World" will therefore take no less than 4000 lines of code to work around the bugs (as long as you include the 20 necessary semi-compatible 100 MB libraries), and will require numerous other kludges to implement correctly.
I like Google, maybe more than most, but let's just stop the insanity, cutting the numerous bloated layers of mess, and make the OS (which shock, doesn't need a browser to access the internet!) the base from which to build all else upon. Unnecessary layers are kludges; always have been, always will be.
I wish they'd make a 100 watt version of that. And by 100, I mean REALLY 100 so as to imitate a 500 watt incandescent. It'd be compatible with all light sockets and the light output would be incredible.
Solution: all bulbs should go through an external testing process which sees how long the CFLs last on average. This time would then be stated on the packaging, and it would be illegal to give a false figure.
These are good times for software patents. And by that I mean that all these companies are at each other's necks accusing each other of using their patented tech. Because of this, the whole thing will implode out of sheer exhaustion, and eventually the patent office won't be so quick to dish out patents to generic, uninspired ideas which sometimes a kid could think of.
Fair point, though apparently the initial efficiency of iter will be quite a lot less than the TWR afaik. Still, fusion would be awesome agreed! They should pump 100s of billions into nuclear/fusion research.
Slashdot Mirror
One thing I noticed in the article was this:
and because of some fundamental results in computer science, “I know for a fact that he’ll never be able to get it exactly right.” But, Weimer says, “The vast majority of infinite loops he will be able to figure out.”
Does this we can't get above a certain percentage of correct predictions, or does it mean that more and more work will converge towards 100% (99.5, 99.9%, 99.999% etc. etc.)
Fields per second? Is that frames per second?
We need more fps anyway for lots of reasons (less/no flicker, smoother motion, less/no blurring), so let's just go with that.
Again though, the speaker itself should remain neutral and the composer can still make the original waveform to have more bass if they wish that's how it should be heard, and if the listener is STILL not happy, he can still adjust the bass control to their liking if they don't like the composer's intentions. Heck, the listener can have the bass knob permanently switched to 100% if they really like bass.
My point is that the default position should give a standard sound output across all speakers. Otherwise, it's just a loudness/bassness war game all over again.
By the way, regarding the fluorescent/LED light, have you ever considered it's not the whiteness of the LED lamp they're moaning about, but rather the abnormal spectral spikes (particularly green) in them which can distort colors? How often do you hear someone moan about sunlight? That's much whiter than incandescents.
Yes. If a tune requires more bass, then that should be the job of the original signal, not the speakers. All speakers should ideally sound the same in this case (one can always alter the treble/bass with knobs on the side). There should be something like an SRGB standard for audio.
A speaker than hasn't any bias towards low or high end will emulate anything perfectly. However, a speaker that biases towards low/high end is like a monitor always having a red tint (even for what's meant to blue or green colours), and people saying "I like the warmer rich red tint of this monitor". It's completely ridiculous.
It reminds of the compression/loudness/over-clipped wars for music actually.
Exactly. No matter how bad of Apple this, Anon hit the target. There are various shades of grey, even when it comes to monopolies.
I remember being almost 'very sure' about something to do with whether I could spot a little more latency in game with a certain option. However, I tried testing myself even though I thought it was almost a waste of time. I was surprised to see that the semi-blind test confounded me. Definitely can be a case of bias if one's not careful.
Hi I'm someone else.
Have you tried blind listening tests on yourself?
If what you say is true, it would be nice to have something even better than applause or heavy percussion to REALLY make the imperfecctions shine in an MP3. Are you saying the same happens with an 88khz sampling frequency too?
What you actually need to do is link to a sample - an uncompressed sound that when compressed shows MP3/OGG etc. really at its weakest. A sound that one can universally use to show the artifacts that frequency domain techniques exhibit without the need for even a blind test. A sound so obviously bad in MP3, that applause or heavy percussion sounds wonderful in comparison.
I'm sure synthesizing the sound and using GA/GP evolving (with human raters to measure the badness) would result in something wonderfully terrible. We can then use that sound as an acid test (or at least one of the acid tests) for future reference and cut through the subjectivity and misinformation which is so rife in this subject.
It's embarrassing because it's almost *too* efficient. The term first came about when CPU manufacturers and the industry in general were embarrassed at how much faster the GPU was for certain algorithms (i.e. the embarrassingly parallel ones). Programmers also were generally embarrassed at not using the technology sooner, and they often spent YEARS writing efficient code for the CPU, only to have a 5 minute knock-up code job on the GPU beat it when they finally experimented with the GPU.
The definition was then further reinforced by programmers who were expected to write long convoluted code to show their managers. They were then embarrassed because the stuff was good yet very quick to write (parallel algorithms are by their nature more short and elegant), so it looked like they were being lazy.
For a 'normal' answer, maybe try:
We can better see the proportion of water in the universe, and therefore indirectly deduce other stuff, perhaps say, a more accurate origin of the universe (which in turn will help us unify physics and where the universe is heading, which in turn will give us better technology for other things, which in turn will make us use said technology to stimulate our 5 senses, which in turn will make us happier, and that's where we reach the end of the line, as happiness *really is* the ultimate goal).
Also, thanks to this, perhaps we will now know where to look, or how to look more easily for water somewhat closer to us (where in the future, we may have a chance of reaching).
Yes I know what you mean, but often they don't just adjust the 'gamma' of the sound like you describe, but also truncate the floor and ceiling to make a more square-wave like sound. That's what I meant when I said 'compression' (I meant clipping), and many CDs do just that. They also do what you said of course.
I hate the compression ramifications as much as anyone. However, it does make the sound more like a square wave, which is how the C64 SID chip sounds (I'm also a bit of a chip fan). You get particular harmonic overtones which help give richness to a sound. So maybe people are really responding to *that*.
Obviously the producers are going about it all wrong though, since they get the 'square wave' style, but also lose detail with most of the instruments.
You can get the best of both worlds.
I understand Apple's philosophy of only allowing the best apps in their istore. However, what they don't realise is that exactly the same can be achieved with decent rating system if it's done relatively well (Slashdot, or download.com though not perfect, spring to mind). That way, people can choose the best, and we don't have a mild form of censorship. It also cuts out the maintenance.
You get levels though. 99% secure or 99.99999% secure is a huge difference.
I'm usually the last person to speak about the benefits of more work (and will cite the broken windows fallacy at the nearest opportunity).
But on this occasion, it looks like they could do with more workers. Qualified workers at that.
Good, the more ridiculous the patents get, the quicker something will be done to fix the mess. Personally, I'd like to see this patent granted, and dozens of companies ordered to pay lots of damages to the angelic company that is Kootol. ....if only to see the backlash from a thousand juggernauts against the current patent system ;)
Wonderful post. I love your 2nd last semi-parseable sentence with 'every', every other word :)
I would only add that developers should pay our saviour, yes, but also be locked in so that if they develop for Windows, they can't develop that software for other OSs, ever. After all, what would be the need?
Or how about we all move to a cross based screw tip (open at ends) which, providing a cross-shaped screwdriver is used, has the following advantages over the standard flat/slot-head or the phillips:
1: It's cross compatible with the old flathead.
2: With a cross-shaped screwdriver, it won't slip out from either side, unlike the flathead.
3: A large degree of tolerance is allowed for the size of the cross-screwdriver, more so than the phillips.
4: It's stronger, and will therefore distort the metal (both of the screw and driver) less than a standard flat-head.
5: It's easier to fit the cross in than the flathead and almost as easy as the phillips.
So shock, it's not a zero-sum game, since we have the almost best of both worlds. The ease of use, and non-slippability of the phillips, and the strength and the longevity of the flat head. This hardline relativism sucks - some tools really ARE better. And that's coming from somebody who's never really thought about using these type of screws before, let alone having them.
Granted, it won't be the perfect screw for all occasions, but it'll suit most, and can make many other types redundant. I'm afraid that means not each tool is a special snowflake which is immune to the fate of obsolescence.
I see the pattern now. In another decade, Google will become passée, but the Chrome browser will have become dominant by that point. From here, the new giant corporate of that future era will build an OS *inside* the Chrome browser (itself running on Linux or Windows). A decade after that, someone will then build a browser inside of THAT.
In a couple of centuries, we will be stuck with a giant crumbling 20 layer behemoth with the top layer inheriting all of the bugs of the previous generations. "Hello World" will therefore take no less than 4000 lines of code to work around the bugs (as long as you include the 20 necessary semi-compatible 100 MB libraries), and will require numerous other kludges to implement correctly.
I like Google, maybe more than most, but let's just stop the insanity, cutting the numerous bloated layers of mess, and make the OS (which shock, doesn't need a browser to access the internet!) the base from which to build all else upon. Unnecessary layers are kludges; always have been, always will be.
RAM is cheap. Let's just use that (apart from cache obviously).
I wish they'd make a 100 watt version of that. And by 100, I mean REALLY 100 so as to imitate a 500 watt incandescent. It'd be compatible with all light sockets and the light output would be incredible.
Except it obviously isn't a commodity. Yet.
Solution: all bulbs should go through an external testing process which sees how long the CFLs last on average. This time would then be stated on the packaging, and it would be illegal to give a false figure.
Q: How many fanless spinning heatsinks does it take to change a lightbulb?
A: Just one. It stays put, and lets the CPU, PC (and attached world) revolve around it.
These are good times for software patents. And by that I mean that all these companies are at each other's necks accusing each other of using their patented tech. Because of this, the whole thing will implode out of sheer exhaustion, and eventually the patent office won't be so quick to dish out patents to generic, uninspired ideas which sometimes a kid could think of.
Fair point, though apparently the initial efficiency of iter will be quite a lot less than the TWR afaik.
Still, fusion would be awesome agreed! They should pump 100s of billions into nuclear/fusion research.