Very interesting; is there a technical book (or chapter) or paper with a good overview of this comparative aspect of fly neurons?
I was just starting to look around to see what's available on comparative neuroscience in general, based on an interest in the most salient functional differences from human neurons, so anything related to that more general topic would also be welcome.
In theory, if we can capture coherent pictures in the visible spectrum from many billions of light years away, we should be able to do the same with RF.
It's actually very easy to see why the opposite is true: stars famously broadcast a truly vast amount of power in the visible spectrum, which is what makes solar energy and photosynthesis effective.
Humans clearly do not have the power resources of the entire sun to use to power RF broadcasts. The total amount of power we have at our disposal from all sources is a tiny, tiny fraction of what the sun broadcasts.
And most of our power does not go into RF in the first place, it goes into transportation, manufacturing, etc.
So it's quite straightforward that there is no comparison between the brightness of stars in the visible spectrum versus the Earth in RF. Stars win hands down.
It isn't the technology, it's just the hardware.
Unfortunately, it is very much both. It's true that we can do better by building better listening arrays, and SETI has been continually doing that for many years, but there is also a problem of signal to noise ratio that gives a hard limit on sensitivity due to noise from terrestrial sources and from thermal and quantum noise in the receiving electronics.
Part of that could be improved by putting radio telescopes e.g. on the far side of the moon. The electronics issue simply needs better technology.
Yet the sky is not saturated with their communications. So therefor those civilizations must be using some other technology.
That seems logical, but that turns out not to be the case. A SETI scientist said in a talk (and I've seen this in articles since) that our deployed SETI listening technology is still nowhere near sensitive enough to pick up signals even from as close as the nearest star (Proxima Centauri, 4 light years away), if a planet there was broadcasting RF at current Earth levels.
(That doesn't mean SETI to date is pointless, because there's always a chance of a highly directional signal beamed our way, or of just something unexpected, like signals far far brighter than Earth's.)
So no, we have no idea whether the sky is saturated with radio waves or not.
"a bit much" is an idiomatic way in English of saying "possibly you shouldn't have been modded down so much for that statement" (because the first part of what you said is arguable, not an obvious troll, not obviously flamebait, not an obviously incorrect statement of fact).
I think that you being modded down to -1 is a bit much, but there is a problem here. What you said is potentially well and good for contexts that are purely utilitarian to the degree that anything but pure pragmatic functionality is to be viewed as an active negative, such as industrial control, power plants, etc.
But for most people's desktops, people expect both functionality *and* some degree of modern aesthetics, and there is an extremely common rejection of interfaces that look 15 years old, even if they were considered close to ideally functional and aesthetic 15 years ago.
Since that is demonstrably what the market generally wants, that is therefore the general trend over time: "flashy graphics" are sometimes overdone, but the word "flashy" is in the eye of the beholder, and most improvements to GUIs over the decades have been about modernization to meet the moving target of whatever "modern" means in each era, with actual breakthroughs in usability being far less common.
Furthermore, the people who design and implement GUIs are (with the exception of 1-person development teams) rarely the same team members who would address software vulnerabilities, so maybe that's where your -1 came from.
There *is* such a thing as (write, execute), and this is what it is currently using, and people are complaining quite bitterly about it doing that, because it's a security weakness that is exploitable via buffer overflows etc.
Basically the fact that it allows for dynamic JIT compilation is exactly what people don't want the bad guys to be doing; potentially that turns a small bug into a complete take-over-the-machine exploit.
I'm afraid I was taking all that for granted in my original comment, which instead was about potential reasons why they might not have done something different as people are suggesting in hindsight.
I did mean that about time pressure, but I don't understand what you are suggesting.
The scenario is to suppose that they have some audio algorithm that is
based on creating and executing a block of code a bunch of times, but
for a very small fraction of a second, and then, critical assumption, it
needs to create and execute a slightly different block of code, and so on,
over and over.
Long ago I saw a demo of an Amiga video capture program that had precisely
that sort of thing going on, where they could only get the throughput they
needed by using self-modifying code, changing one instruction in an inner
loop on a regular basis, and their approach held up under scrutiny of 100
developers critiquing them; it really was needed.
So I am sure that such things do arise in the real world; whether the current
codebase truly has such a need, I don't know.
So *if* they have this need, then every time they create or modify a code
block, they need it to be writable, but when they execute what they wrote,
security issues insist that it it should go from (write,no-execute) to
(no-write,execute), every single time.
Making that permissions change requires a system call every time, unavoidably.
Any scheme that allows avoiding that system call is inherently making different
assumptions than I did above. Without knowledge of their algorithms, I don't
see how we can be sure that assumptions like that are either right or wrong;
I'm just pointing out that *if* they are correct, they explain why the codebase
does what it does.
An alternative explanation that I tried to give a nod to is that they may have
simply done a premature optimization that was not actually needed, but
again, motivated as outlined above.
Whether it is noticeably expensive depends on endless details of their code, but if they did extremely fast JIT code generation on a very frequent basis (possibly just a single instruction change), then the reflagging syscall would be equally frequent, and the context switch overhead could dominate.
The scenario under which it truly hurts is possibly unlikely, I'll grant, but at the time of implementation it probably seemed like the simplest and fastest approach, and clearly they were thinking about speed, not security and Return Oriented Programming and such.
You have been misinformed. Doctors are *not* "supposed to be the smartest people on the planet", not even close.
The average citizen in the street may think so, but that's not saying anything.
As for physics, it doesn't make people smart to study physics, it just tends to attract some of the smartest students. Having pre-meds major in physics wouldn't make them any smarter.
"OpenSimulator lacks support for many of the game-specific features of Second Life (on purpose), while pursuing innovative directions towards becoming the bare bones, but extensible, server of the 3D Web."
And to think I waited a few days before registering an account here.
Tell me about it; my thinking was the same "another site to register at?". But you still managed a 3 digit ID; I delayed a couple weeks I think, when IDs were introduced, and ended up at 4 digits.
Overall a good summary, but it was IBM that inspired "FUD", not Microsoft.
IBM is (AFAIK) not especially evil anymore, but they were deeply evil and hated in the 60s and 70s. Microsoft took over from them like a runner taking the baton in a relay race.
Of course it does. That's like saying that digital electronics theory works for you, no need to bother with analog electronics theory.
True, yet digital is nonetheless based on analog.
I'm by no means the biggest advocate of string theory, but obviously it is intended (regardless of its current incompleteness) to be a deeper level of description of the universe than is the standard model.
Even if string theory (or loop quantum gravity or spin foams) eventually succeeds at being a good theory of everything, it is quite possible, even likely, that it will continue to be more practical for most purposes to work in terms of a modified standard model.
We still make heavy use of Newtonian physics, after all.
But that doesn't mean that a theory of everything is pointless; its purpose is not identical to that of simplified working models.
Years ago I knew why some premium sound cards were worth the money, but I've long since lost track of the marketplace (aside from reading various things in the last year about old versus current versus upcoming Linux sound support).
So I'm curious, what do you get on today's systems that makes it worthwhile to pay $200 for a sound card?
I had the feeling it was no longer for wavetable, nor for number of bits of d-to-a conversion. Is it 7.1 surround, or what?
Since you posted as AC and I'm answering 10 days later, you probably won't see this, but for the hell of it: I maxed out on karma years ago, back when they still displayed it numerically.
That means that it's likely that my karma will never change to speak of (since I'm not planning to ruin it with a career of trolling), so it's questionable what "karma whoring" even means, under those circumstances.
I was just peeved at the incorrect claims endlessly repeated, so I made multiple corrections rather than just one.
I did get modded down for redundancy, understandably, and that seems the more apt term. FWIW.
Sorry for the delay in answering. You raise some interesting issues, not all of which are answerable given the state of the art, as I understand it at least.
On the subject of noise, there are very strong theoretical reasons (e.g. Shannon's theorem) to think that noise in a channel in the same mode as the information modulation is an absolute limit to the rate of data transmission in that channel using that mode of modulation
The bolded caveats are important. Noise present only in one channel doesn't affect information transmission in a different channel (e.g. one channel using radio, a second using optical). Approximately white noise in the amplitude domain has limited impact on frequency domain modulation (to a first loose approximation), which is why phase locked loops manage the apparent magic of being able to track FM signals below the noise floor.
Little is known about sources of noise in this new OAM channel. Certainly nature must produce *some* noise there (due to various fundamental principles), but it is likely IMHO to decrease with increasing OAM energy (see the history of physics topic "ultraviolet catastrophe" to see why). But we have much to learn; we shall see.
As to single photon RF systems and exotic technology, well, we're certainly talking about a very exotic technology here, so I wouldn't worry about exotic-ness.:-)
Finally, as to photon counting versus group ensemble production/detection in other regards, that's something interesting to ponder. I have a dozen half-baked thoughts on the subject, but mostly they are about various kinds of tradeoffs and circumstances, rather than about a rule of thumb concerning maximizing data rate. Instead of trying to to verbalize all that, I'll just say that I don't know.
Usually I and others will just make a correction once, and shrug off repetition of nonsense (although I must admit a lot of it was unusually erudite nonsense), but I got peeved this time.
It's been a pet topic, ever since I heard a little tidbit about it years ago I've been waiting for the other shoe to drop, and it finally did (a bit, anyway), but what I thought was some of the coolest news of the century was swamped by noise. Argh!
"I looked at it. It looks like they're using an interferometer to gauge degrees of polarization."
Well, you're contradicting them. Look at the left side of the page; there's an animation of polarized light.
Look at the right side of the page; there's an animation of light with orbital angular momentum.
Look at the text in between the right and the left, explaining how polarization is different than OAM, and that what they are doing is the latter, not the former.
Now pretend they're lying, is that your plan? Their page is unambiguous, so I don't know what you think you'll accomplish by stubbornly misinterpreting it.
I agree that Shannon steps in with random noise; no one said differently.
That's not the case. The Nyquist limit using past technology for a 2Mhz-wide carrier is at most 1 megabit per second, no matter how you modulate, no matter how you fiddle with sidebands etc.
This new technology allows further data transmission on top of that, for a total of *more* than 1 megabit per second (how much more isn't clear yet, since there's no obvious cap on the orbital angular momentum).
If it were what you and the OP are claiming, then the absolute limit would still be 1 mbps.
This isn't some sort of disproof of Shannon/Nyquist, of course; it's just the appearance of a new channel in addition to the previously known channels, so presumably it, too, will follow the traditional theorems (although I personally don't know the correct way to *apply* them to this new domain).
Slashdot Mirror
Maybe I missed something.
Very interesting; is there a technical book (or chapter) or paper with a good overview of this comparative aspect of fly neurons?
I was just starting to look around to see what's available on comparative neuroscience in general, based on an interest in the most salient functional differences from human neurons, so anything related to that more general topic would also be welcome.
It's actually very easy to see why the opposite is true: stars famously broadcast a truly vast amount of power in the visible spectrum, which is what makes solar energy and photosynthesis effective.
Humans clearly do not have the power resources of the entire sun to use to power RF broadcasts. The total amount of power we have at our disposal from all sources is a tiny, tiny fraction of what the sun broadcasts.
And most of our power does not go into RF in the first place, it goes into transportation, manufacturing, etc.
So it's quite straightforward that there is no comparison between the brightness of stars in the visible spectrum versus the Earth in RF. Stars win hands down.
Unfortunately, it is very much both. It's true that we can do better by building better listening arrays, and SETI has been continually doing that for many years, but there is also a problem of signal to noise ratio that gives a hard limit on sensitivity due to noise from terrestrial sources and from thermal and quantum noise in the receiving electronics.
Part of that could be improved by putting radio telescopes e.g. on the far side of the moon. The electronics issue simply needs better technology.
That seems logical, but that turns out not to be the case. A SETI scientist said in a talk (and I've seen this in articles since) that our deployed SETI listening technology is still nowhere near sensitive enough to pick up signals even from as close as the nearest star (Proxima Centauri, 4 light years away), if a planet there was broadcasting RF at current Earth levels.
(That doesn't mean SETI to date is pointless, because there's always a chance of a highly directional signal beamed our way, or of just something unexpected, like signals far far brighter than Earth's.)
So no, we have no idea whether the sky is saturated with radio waves or not.
"a bit much" is an idiomatic way in English of saying "possibly you shouldn't have been modded down so much for that statement" (because the first part of what you said is arguable, not an obvious troll, not obviously flamebait, not an obviously incorrect statement of fact).
I think that you being modded down to -1 is a bit much, but there is a problem here. What you said is potentially well and good for contexts that are purely utilitarian to the degree that anything but pure pragmatic functionality is to be viewed as an active negative, such as industrial control, power plants, etc.
But for most people's desktops, people expect both functionality *and* some degree of modern aesthetics, and there is an extremely common rejection of interfaces that look 15 years old, even if they were considered close to ideally functional and aesthetic 15 years ago.
Since that is demonstrably what the market generally wants, that is therefore the general trend over time: "flashy graphics" are sometimes overdone, but the word "flashy" is in the eye of the beholder, and most improvements to GUIs over the decades have been about modernization to meet the moving target of whatever "modern" means in each era, with actual breakthroughs in usability being far less common.
Furthermore, the people who design and implement GUIs are (with the exception of 1-person development teams) rarely the same team members who would address software vulnerabilities, so maybe that's where your -1 came from.
There *is* such a thing as (write, execute), and this is what it is currently using, and people are complaining quite bitterly about it doing that, because it's a security weakness that is exploitable via buffer overflows etc.
Basically the fact that it allows for dynamic JIT compilation is exactly what people don't want the bad guys to be doing; potentially that turns a small bug into a complete take-over-the-machine exploit.
I'm afraid I was taking all that for granted in my original comment, which instead was about potential reasons why they might not have done something different as people are suggesting in hindsight.
The scenario is to suppose that they have some audio algorithm that is based on creating and executing a block of code a bunch of times, but for a very small fraction of a second, and then, critical assumption, it needs to create and execute a slightly different block of code, and so on, over and over.
Long ago I saw a demo of an Amiga video capture program that had precisely that sort of thing going on, where they could only get the throughput they needed by using self-modifying code, changing one instruction in an inner loop on a regular basis, and their approach held up under scrutiny of 100 developers critiquing them; it really was needed.
So I am sure that such things do arise in the real world; whether the current codebase truly has such a need, I don't know.
So *if* they have this need, then every time they create or modify a code block, they need it to be writable, but when they execute what they wrote, security issues insist that it it should go from (write,no-execute) to (no-write,execute), every single time.
Making that permissions change requires a system call every time, unavoidably.
Any scheme that allows avoiding that system call is inherently making different assumptions than I did above. Without knowledge of their algorithms, I don't see how we can be sure that assumptions like that are either right or wrong; I'm just pointing out that *if* they are correct, they explain why the codebase does what it does.
An alternative explanation that I tried to give a nod to is that they may have simply done a premature optimization that was not actually needed, but again, motivated as outlined above.
The scenario under which it truly hurts is possibly unlikely, I'll grant, but at the time of implementation it probably seemed like the simplest and fastest approach, and clearly they were thinking about speed, not security and Return Oriented Programming and such.
Yes, but the re-flagging costs a system call, which is potentially very non-trivial overhead in anything that is using JIT for speed.
The average citizen in the street may think so, but that's not saying anything.
As for physics, it doesn't make people smart to study physics, it just tends to attract some of the smartest students. Having pre-meds major in physics wouldn't make them any smarter.
https://en.wikipedia.org/wiki/Windows-1250
Sounds cool, but not for the purpose at hand.
(and: PDP-11 FTW -- I was part of one of the efforts mentioned in this Strange Birth article)
Indeed. I'm with you, brother.
And to think I waited a few days before registering an account here.
Tell me about it; my thinking was the same "another site to register at?". But you still managed a 3 digit ID; I delayed a couple weeks I think, when IDs were introduced, and ended up at 4 digits.
IBM is (AFAIK) not especially evil anymore, but they were deeply evil and hated in the 60s and 70s. Microsoft took over from them like a runner taking the baton in a relay race.
http://en.wikipedia.org/wiki/Fear,_uncertainty_and_doubt#Definition
A modified standard model works for me
Of course it does. That's like saying that digital electronics theory works for you, no need to bother with analog electronics theory.
True, yet digital is nonetheless based on analog.
I'm by no means the biggest advocate of string theory, but obviously it is intended (regardless of its current incompleteness) to be a deeper level of description of the universe than is the standard model.
Even if string theory (or loop quantum gravity or spin foams) eventually succeeds at being a good theory of everything, it is quite possible, even likely, that it will continue to be more practical for most purposes to work in terms of a modified standard model.
We still make heavy use of Newtonian physics, after all.
But that doesn't mean that a theory of everything is pointless; its purpose is not identical to that of simplified working models.
So I'm curious, what do you get on today's systems that makes it worthwhile to pay $200 for a sound card?
I had the feeling it was no longer for wavetable, nor for number of bits of d-to-a conversion. Is it 7.1 surround, or what?
That means that it's likely that my karma will never change to speak of (since I'm not planning to ruin it with a career of trolling), so it's questionable what "karma whoring" even means, under those circumstances.
I was just peeved at the incorrect claims endlessly repeated, so I made multiple corrections rather than just one.
I did get modded down for redundancy, understandably, and that seems the more apt term. FWIW.
On the subject of noise, there are very strong theoretical reasons (e.g. Shannon's theorem) to think that noise in a channel in the same mode as the information modulation is an absolute limit to the rate of data transmission in that channel using that mode of modulation
The bolded caveats are important. Noise present only in one channel doesn't affect information transmission in a different channel (e.g. one channel using radio, a second using optical). Approximately white noise in the amplitude domain has limited impact on frequency domain modulation (to a first loose approximation), which is why phase locked loops manage the apparent magic of being able to track FM signals below the noise floor.
Little is known about sources of noise in this new OAM channel. Certainly nature must produce *some* noise there (due to various fundamental principles), but it is likely IMHO to decrease with increasing OAM energy (see the history of physics topic "ultraviolet catastrophe" to see why). But we have much to learn; we shall see.
As to single photon RF systems and exotic technology, well, we're certainly talking about a very exotic technology here, so I wouldn't worry about exotic-ness. :-)
Finally, as to photon counting versus group ensemble production/detection in other regards, that's something interesting to ponder. I have a dozen half-baked thoughts on the subject, but mostly they are about various kinds of tradeoffs and circumstances, rather than about a rule of thumb concerning maximizing data rate. Instead of trying to to verbalize all that, I'll just say that I don't know.
Apparently, so is the art of RTFA. And reading even the other comments.
Usually I and others will just make a correction once, and shrug off repetition of nonsense (although I must admit a lot of it was unusually erudite nonsense), but I got peeved this time.
It's been a pet topic, ever since I heard a little tidbit about it years ago I've been waiting for the other shoe to drop, and it finally did (a bit, anyway), but what I thought was some of the coolest news of the century was swamped by noise. Argh!
Well, you're contradicting them. Look at the left side of the page; there's an animation of polarized light.
Look at the right side of the page; there's an animation of light with orbital angular momentum.
Look at the text in between the right and the left, explaining how polarization is different than OAM, and that what they are doing is the latter, not the former.
Now pretend they're lying, is that your plan? Their page is unambiguous, so I don't know what you think you'll accomplish by stubbornly misinterpreting it.
I agree that Shannon steps in with random noise; no one said differently.
This new technology allows further data transmission on top of that, for a total of *more* than 1 megabit per second (how much more isn't clear yet, since there's no obvious cap on the orbital angular momentum).
If it were what you and the OP are claiming, then the absolute limit would still be 1 mbps.
This isn't some sort of disproof of Shannon/Nyquist, of course; it's just the appearance of a new channel in addition to the previously known channels, so presumably it, too, will follow the traditional theorems (although I personally don't know the correct way to *apply* them to this new domain).