EM waves have frequency and polarization and phase. Their "orbital angular momentum" is some combination of these parameters so you can't increase bandwidth over what can be done using some combination of these.
Actually, I don't think their OAM is a combination of those parameters. It's about the spatial distribution of the phase around the axis of transmission.
It's using the spatial variation of the signal. In cylindrical coordinates (r,theta,z) aligned with the axis of transmission z, it uses different phases at different thetas. In particular, a bunch of superposed signals each with phase varying around the z axis as cos(i*theta) for i=0,1,2,... should stay conveniently distinct from transmitter to receiver.
I think the axis of transmission is baked into the idea pretty deeply - it's inherently unidirectional. I also think it's not robust to superposition: another such bunch of signals passing obliquely across the receiver would mess everything up.
The Prius Lane Keep Assist feature does steer a bit, gently - the wheel tends to drift toward it's best guess of the center of the lane. It won't drive for you, though: if you take your hands off the wheel it notices (I think it notices the absence of any applied torque over some reasonably short interval), sounds an alarm and turns off the feature.
Heart rate monitors can and have been used in PE classes to grade students on their exertion rather than their capabilities. In stereotypes: the idea is to reward the nerd busting his ass rather than the jock breezing through, even though the jock might jog faster than the nerd can run.
An effect tending to limit the total number of laws is a benefit, not a drawback. Better fewer, simpler, wiser laws that apply correctly to ever-broadening human experience than an ever-broadening mass of special-case rules for the each hot topic over the course of history.
Not that it's easy. All communication seems to be harder to get right tersely than verbosely. Surely legal code must be one of the hardest cases of this problem.
Such speed-of-light latency would exist, as you say; and would be irrelevant due to other factors, as you say. What you seem to be missing is that that latency is present for _any_ kind of link. In fact, the speed of light in fibre (just to take as an example one of the more likely convential signalling media) is slightly less than the speed of light in a vacuum you used in your calculation. Ain't no way bits'll get from point A to point B faster'n light.
Thanks to an ingenious suggestion from my fiancee, I store my various computer cables in the many pockets of a hang-on-the-back-of-the-closet-door style shoe organizer. It works beautifully.
On line 15 of page 9 of the nearly illegible scanned documents in the link given, is the interesting requirement of the (to be determined) standard that:
any software portion of such standards is based on open source code.
On of my (many) concerns with this legislation has been that an adopted standard would be unimplementable in an open-source OS. This seems to address that. Hmm.
The solution to this is logically trivial, though
the business end may be troublesome:
Net access (even cable modems or DSL) should be sold per unit bandwidth, not per user. The whole point of _inter_net is the sort of interconnection we're talking about here. When selling a connection, restricting the end user not to have other connections is, in the end, counter productive. However, if many users try to use one connection, their bandwidth needs will increase. A better pricing model is all that's needed. If, at some point, consumers want to do something for which a different pricing model is better for everyone, something will bend.
Don't get me wrong, this sound's pretty cool, but the article makes false claims:
There are around 20 non-tethered undersea exploratory robots in the world, but they are of limited utility as they all run on expensive silver-zinc power cells that can be recharged no more than 50 times or so before they become useless.
I worked on a building a brain upgrade for ABE. The original system runs in FORTH and C on an agglomeration of hand-coded microcontrollers and Transputers. The new system (still under development) is a PC104 stack... running Linux.
A watch factory is an augmentation of a watchmaker. Someone programmed the robots in the factory, after all. If it's possible for a small group of people to make many watches by building a factory, should a society really waste perfectly good people on the task of moving small gears into specific arrangements? Much better that those people should be able to get on with tasks that aren't so, well, mechanical.
If a persons job could be eliminated with a machine, how rewarding could the job have been?
Now, if you want to argue that we don't allocate sufficent resources to education, so that a replaced worker can't go back to school and learn something else more interesting, I'll agree with you there. But there's too much fundamental work to do in the world -- getting food to hungry people strikes me as a good example -- for there to be anything wrong with the elimination of jobs. The problem isn't a shortage of work, it's a failure of organization.
Axiom: When the ability to copy is ubiquitous, and when the incremental cost of copying is effectively zero,
the effective value of any given copy -- including the "original" copy -- is zero. (I state this as axiomatic, but
I'm willing to discuss its merits. And please note that this assertion says nothing about the effort/resources
required to create the original in the first place.)
Your claim about the original copy being worthless
isn't really true, as you yourself go on to prove.
Once information has been copied and is freely available, it's value does drop precipitously. But previously unreleased information has value.
Imagine you're going to a party in a hypothetical net-friendly world. There's not much point in bringing along a popular mp3 -- your host or anyone else could just download it if they like it. Bringing (or uploading, whatever) a good mp3 no-one's heard yet, however, would be a different matter entirely. Presupposing your friends share your taste, new music is a definite positive contribution to the party.
The value in information shifts from ownership of ideas to the distribution of new ideas. I would therefore argue that an original copy of information -- before it is copied -- is very valuable. I think some of your models for revenue generation support this claim.
The situation in which open source shines is that in which the code is widely understandable. While an idea is too new to pass that test, open source may not be the best choice. Once the idea is well known -- so that it's well known what needs doing and all that remains is to do it and do it right -- then it's time for open source.
Consider whether or not you can split your driver into two pieces
one closed source module which implements a well documented interface, hiding whatever secrets you wish, which you keep as small as possible
one open source module which uses the aforementioned well documented interface to implement the driver
If, as time passes, your secret becomes less of a big deal, you can open the closed module, or merge the two.
Open Source and secrecy do not, of course, work too well together. Fortunately, they're not as much competitors as one might think. Secrecy is advantageous when an idea is new -- not just an implementation but what-to-implement is needed. Open Source is advantageous when an idea is old (of course old is very much a relative term in this business) -- a good method is known and an implementation is all that's needed. If you can separate out what's truly innovative in your product and keep only that secret, I'd judge you're making a reasonable compromise.
With kernel code, one should pay extra special attention to the interface between the two pieces. A bug in your closed source code can affect the whole system, not just an application using your product. For that reason, you should not only be keeping your closed code as short and simple as possible, but document every guarantee you can and can't make. That way, if (heaven forbid) there is a problem with your code, someone working with just the open source part can diagnose your closed module as failing to obey its own constraints.
This scenario isn't as rare as one might think -- it's just that the open/closed boundary is usually (for open source OSs) in the same place as the software/hardware boundary.
In my experience, these "Recovery CD"s make setting up multiboot systems very difficult. When my friends come to me with their new computers, wanting to try Linux, they usually want a multiboot system -- they don't want to give up what they know before they've tried the new option. This new distribution method makes it very hard to give them that option.
With the "Recovery CD"s, the only option for reinstalling Windows forces a repartition and reformat of the entire drive -- it will wipe any other operating system off the machine.
It may still be possible to use a dynamic repartitioning tool to squash a Windows partition down and make room for another OS without using the "Recovery CD". Still, if anything ever goes wrong with the Windows installation (which is, empirically, likely) and Joe User is forced to use the CD, then all of a sudden the other OS will be gone.
a mix of good sense and strong bias
on
Interface Zen
·
· Score: 1
Disclaimer: I am an emacs user.
I believe the arguments against chording are specious. I often achieve this zen state while programming with emacs, which uses chorded combinations extensively. In optimizing my environments for myself, I tend to choose chords and sequences in roughly equal proportion. Heck, I've got my X modmap configured to let me use all five buckybits. I don't claim everyone should, just that I'm a counterexample.
This article combines a good point -- that a user interface should keep out of a user's way -- buried in a tremendous bulk of plugs for particular opinions in age old flamewars. I was surprised I didn't find any arguments for or against the use of keyclick. Without the pro-vi baggage, the article makes a good, but simple, point. (As a side note, this article brings to my mind the pro-ed diatribe distributed in the JOKES file with emacs. I recommend it to anyone who hasn't read it).
Slashdot Mirror
EM waves have frequency and polarization and phase. Their "orbital angular momentum" is some combination of these parameters so you can't increase bandwidth over what can be done using some combination of these.
Actually, I don't think their OAM is a combination of those parameters. It's about the spatial distribution of the phase around the axis of transmission.
It's using the spatial variation of the signal. In cylindrical coordinates (r,theta,z) aligned with the axis of transmission z, it uses different phases at different thetas. In particular, a bunch of superposed signals each with phase varying around the z axis as cos(i*theta) for i=0,1,2,... should stay conveniently distinct from transmitter to receiver.
I think the axis of transmission is baked into the idea pretty deeply - it's inherently unidirectional. I also think it's not robust to superposition: another such bunch of signals passing obliquely across the receiver would mess everything up.
The Prius Lane Keep Assist feature does steer a bit, gently - the wheel tends to drift toward it's best guess of the center of the lane. It won't drive for you, though: if you take your hands off the wheel it notices (I think it notices the absence of any applied torque over some reasonably short interval), sounds an alarm and turns off the feature.
978-0517545164
"Images may not be copied or offloaded, and the images and their texts may not be published. All digital images of the Oxyrhynchus Papyri are © Imaging Papyri Project, University of Oxford. The papyri themselves are owned by the Egypt Exploration Society, London. All rights reserved."
They want help transcribing these documents, but don't want anyone to keep copies of the images? How rude.
The article says they're stopping production of the roadster to focus on the S. It says nothing about anything failing.
Heart rate monitors can and have been used in PE classes to grade students on their exertion rather than their capabilities. In stereotypes: the idea is to reward the nerd busting his ass rather than the jock breezing through, even though the jock might jog faster than the nerd can run.
An effect tending to limit the total number of laws is a benefit, not a drawback. Better fewer, simpler, wiser laws that apply correctly to ever-broadening human experience than an ever-broadening mass of special-case rules for the each hot topic over the course of history.
Not that it's easy. All communication seems to be harder to get right tersely than verbosely. Surely legal code must be one of the hardest cases of this problem.
Such speed-of-light latency would exist, as you say; and would be irrelevant due to other factors, as you say. What you seem to be missing is that that latency is present for _any_ kind of link. In fact, the speed of light in fibre (just to take as an example one of the more likely convential signalling media) is slightly less than the speed of light in a vacuum you used in your calculation. Ain't no way bits'll get from point A to point B faster'n light.
Thanks to an ingenious suggestion from my fiancee, I store my various computer cables in the many pockets of a hang-on-the-back-of-the-closet-door style shoe organizer. It works beautifully.
Good points. Thanks.
On of my (many) concerns with this legislation has been that an adopted standard would be unimplementable in an open-source OS. This seems to address that. Hmm.
The solution to this is logically trivial, though
the business end may be troublesome:
Net access (even cable modems or DSL) should be sold per unit bandwidth, not per user. The whole point of _inter_net is the sort of interconnection we're talking about here. When selling a connection, restricting the end user not to have other connections is, in the end, counter productive. However, if many users try to use one connection, their bandwidth needs will increase. A better pricing model is all that's needed. If, at some point, consumers want to do something for which a different pricing model is better for everyone, something will bend.
Don't get me wrong, this sound's pretty cool, but the article makes false claims:
The Autonomous Benthic Explorer (ABE) built at the Deep Submergence Laboratory of the Woods Hole Oceanographic Institute originally used lead-acid batteries and now uses lithium cells.
I worked on a building a brain upgrade for ABE. The original system runs in FORTH and C on an agglomeration of hand-coded microcontrollers and Transputers. The new system (still under development) is a PC104 stack... running Linux.
A watch factory is an augmentation of a watchmaker. Someone programmed the robots in the factory, after all. If it's possible for a small group of people to make many watches by building a factory, should a society really waste perfectly good people on the task of moving small gears into specific arrangements? Much better that those people should be able to get on with tasks that aren't so, well, mechanical. If a persons job could be eliminated with a machine, how rewarding could the job have been? Now, if you want to argue that we don't allocate sufficent resources to education, so that a replaced worker can't go back to school and learn something else more interesting, I'll agree with you there. But there's too much fundamental work to do in the world -- getting food to hungry people strikes me as a good example -- for there to be anything wrong with the elimination of jobs. The problem isn't a shortage of work, it's a failure of organization.
Imagine you're going to a party in a hypothetical net-friendly world. There's not much point in bringing along a popular mp3 -- your host or anyone else could just download it if they like it. Bringing (or uploading, whatever) a good mp3 no-one's heard yet, however, would be a different matter entirely. Presupposing your friends share your taste, new music is a definite positive contribution to the party.
The value in information shifts from ownership of ideas to the distribution of new ideas. I would therefore argue that an original copy of information -- before it is copied -- is very valuable. I think some of your models for revenue generation support this claim.
The situation in which open source shines is that in which the code is widely understandable. While an idea is too new to pass that test, open source may not be the best choice. Once the idea is well known -- so that it's well known what needs doing and all that remains is to do it and do it right -- then it's time for open source.
Consider whether or not you can split your driver into two pieces
If, as time passes, your secret becomes less of a big deal, you can open the closed module, or merge the two.
Open Source and secrecy do not, of course, work too well together. Fortunately, they're not as much competitors as one might think. Secrecy is advantageous when an idea is new -- not just an implementation but what-to-implement is needed. Open Source is advantageous when an idea is old (of course old is very much a relative term in this business) -- a good method is known and an implementation is all that's needed. If you can separate out what's truly innovative in your product and keep only that secret, I'd judge you're making a reasonable compromise.
With kernel code, one should pay extra special attention to the interface between the two pieces. A bug in your closed source code can affect the whole system, not just an application using your product. For that reason, you should not only be keeping your closed code as short and simple as possible, but document every guarantee you can and can't make. That way, if (heaven forbid) there is a problem with your code, someone working with just the open source part can diagnose your closed module as failing to obey its own constraints.
This scenario isn't as rare as one might think -- it's just that the open/closed boundary is usually (for open source OSs) in the same place as the software/hardware boundary.
In my experience, these "Recovery CD"s make setting up multiboot systems very difficult. When my friends come to me with their new computers, wanting to try Linux, they usually want a multiboot system -- they don't want to give up what they know before they've tried the new option. This new distribution method makes it very hard to give them that option.
With the "Recovery CD"s, the only option for reinstalling Windows forces a repartition and reformat of the entire drive -- it will wipe any other operating system off the machine.
It may still be possible to use a dynamic repartitioning tool to squash a Windows partition down and make room for another OS without using the "Recovery CD". Still, if anything ever goes wrong with the Windows installation (which is, empirically, likely) and Joe User is forced to use the CD, then all of a sudden the other OS will be gone.
No, the entry must be self-contained.
I believe the arguments against chording are specious. I often achieve this zen state while programming with emacs, which uses chorded combinations extensively. In optimizing my environments for myself, I tend to choose chords and sequences in roughly equal proportion. Heck, I've got my X modmap configured to let me use all five buckybits. I don't claim everyone should, just that I'm a counterexample.
This article combines a good point -- that a user interface should keep out of a user's way -- buried in a tremendous bulk of plugs for particular opinions in age old flamewars. I was surprised I didn't find any arguments for or against the use of keyclick. Without the pro-vi baggage, the article makes a good, but simple, point. (As a side note, this article brings to my mind the pro-ed diatribe distributed in the JOKES file with emacs. I recommend it to anyone who hasn't read it).