But it's a good name given the work he put into the theory of geostationary satellites. However, I would like to make a suggestion: that where a marker, be it physical or electronic, is clearly designed with the intent of indicating the presence of alien life in the galaxy, the information type be measured in a ten-point scale of Clarkes. (A basic signal is one Clarke, up to a sufficiently advanced signal that makes it indistinguishable from magic, which would be ten Clarkes. In honour of Arthur C. Clarke's series on the supernatural, complex numbers are permitted.)
It's usually seen as something people hand junk to, although the ideal is that it's where you hand stuff that may be useful to others. Chaitable work is not useless work, it's work that can be reused by others. Charity also has a connotation of sacrifice, that you lose something. Never quite understood that. If a charity collects books for a public library, are you then going to be denied membership? If a charity turns desolate, polluted wasteland into a park, are you going to be denied access?
The answer, to me, is that F/L/OSS is charity, a charity that produces information the same way the above charity donating to a library produces information, and is a charity that turns a bunch of metals and chemicals into a finely-honed computing tool, the same as the above charity created a park. What we do is indeed charitable, not because we deprive ourselves, but because we enrich others. The cost to ourselves is zero, because we would have scratched our itches anyway. You can't rationally add as a cost of sharing the cost of pleasing ourselves.
Charity obviously allows for return on investment, it just means that others also get a return on your investment. But it doesn't require that others give any kind of feedback at all. If you make a public park and only you visit, it's still public, it was still an act of charity, but it's an act of charity you get exclusive benefit from.
Microsoft's statement, then, is a dark one indeed. No charity, of any kind? It says that they gain no pleasure in the results of their labour, that they suffer with every release, that every enhancement and refinement is a source of pain. Quality must be endless torment (which would explain some things). It is a bleak future when everything is misery and there is an apparent determination to spread that misery.
If they wanted to spread even just contentment, through their freely-donated hot-fixes, patches and service packs, freely-donated Microsoft Research products and freely-donated e-mail service and instant messenger, they'd be guilty of charity. Since they have denounced the charitable and all their works, these things cannot be given for the use of others. But, if they are not usable, even in theory, what are they? Microsoft's comments deride and slander all who would offer service to others, so the only conclusion is that these things are intended to cause suffering and misery, which - to judge by Vista service pack 1 - is indeed what they cause.
You don't sound elitist at all. If anything, a little generous. When people multithread and those threads use their own data and own logic, you have a MIMD solution. If they're trivial threads, you may be ok. Otherwise, the number of people who can understand all of the synchronisation and timing issues involved in MIMD... well, pick a number, divide it by itself, subtract 1.
(This is why most people stick with SIMD for local computing and MISD for grids.)
Those using MPI for parallelization are probably (hopefully) aware of PAPI, TAU, KOJAK and DAKOTA. The first two supply accurate timings to profiling software, the other two are profiling packages for MPI environments.
Unfortunately, it is limited to MPI. If you use PVM or BSP, I know of nothing similar. Likewise, if you do a remote memory-to-memory copy (skipping the CPU), profiling is of limited use. And as most MPIs use sequential operations to do collective calls, any idiot could replace an MPI collective send with a reliable multicast. Infiniband and iWarp support it. I doubt the profilers do.
This means you can't rely on high-schol techniques. You've got to plan what happens, when, and how that is to be communicated.
Ah, yes, communication. A simple cluster will work fine with a simple star topology. The bandwidth requirements rise expontially and you are likely to be forced to go with one of the Big Three: Fat Tree (masses of hotspots, prone to failure but cheap), Butterfly (medium risk of failure if the software is imabalanced, and almost any MPI program isn't just unbalanced but stark raving, Hypercube (low risk of failure, very popular with researchers, very stable, very long latencies and very high price).
You also need to weigh network performance and what price that performace is at:
Now most poor research centers can't aford a hypercube topology with 12 GB/s links onto machines capable of digesting that much data. So, instead, they learn to program efficiently, much as the old assembly programmers (like myself) had to do with sparse resources. If you want to get ahead, keep the poverty mindset but exploit the raw computing power before you.
They did most of the early work on which almost all modern parallelism is based. America has tended to follow the notion of faster sequential processing, as the cost savings were not a major factor for universities who could afford big, bright labels. And that's all it's been - playing at which kid gets the toy with the biggest label. Not all, by any means, but a lot. Beowulf clusters, PVM, OpenMPI, Unified Parallel C, Charm++, Cilk, Erlang, Myrinet, Infiniband, RDMA, iWARP... All dveloped in America and all ideally suited to clusters rather than sequential or "big iron" processing. So even though they have never held the limelight, they have achieved a lot. In Europe, Occam was - and is - the major rival. Pi-Occam provides individual programs the kind of power you'd normally use Mobile IP, MOSIX and Globus to achieve (albeit a lot slower than you'd get with Occam).
So to discover that other nations have just walked by and left the EU and US in the dust is a little annoying, but let's face it. Their resources were kept minimal by the west, so it's no surprise they learned the golden rule of boolean logic. waste not, want nots. C'mon, don't pretend we didn't expect this. OpenMP wasn't developed for the fun of it. Parallel has been on the way in for 25+ years. 25 years of Moore's Law applying to their work. You can catch up if you like. It will mean less visible glory, but it would mean doing something real.
Fortunately, there are plenty of software mechanisms already around to solve part of the problem. Unfortunately, very few have been tested outside of small labs or notebooks. We have no practical means of knowing what the different QoS strategies would mean in a real-world network. The sooner Linux and the *BSDs can include those not already provided, the better. We can then - and only then - get an idea of what it is that needs fixing. (Linux has a multitude of TCP congestion control algorithms, plus WEB100 for automatic tuning, so it follows that if there's a rea problem, then it's not really there.)
I know that only a handful of these have been implemented for Linux or *BSD, even fewer for both. Instead of Summer of Code producing stuff nobody ever sees, how about one of the big players invest in students producing some of these meaty chunks of code?
Schemes for reducing packet loss by active queue management: REM, RED, GRED, WRED, SRED, Adaptive RED, RED-Worcester, Self-Configuring RED, Exponential RED, BLUE, SFB, GREEN, BLACK, PURPLE, WHITE
Schemes for adjusting packet queues: CBQ, Enhanced CBQ, HFSC, CSFQ, CSPFQ, PrFQ, Local Flow Separation,
Schemes for scheduling traffic: Gaussian, Least Attained Service, ABE, CSDPS
Schemes for shaping traffic flows: DSS, Constant bit Rate
Schemes for bandwidth allocation: RSVP, YESSIR, M-YESSIR
Schemes for active flow control: ECN, Mark Front ECN
Schemes for managing queues: Adaptive Virtual Queue, PRIO
In a twist, it turns out at least one search was performed by a contractor paid by an Obama advisor. It also appears that the records were accessed multiple times, not just the once (with quick reaction) initially stated. Now, I personally think that passport information is personal information and that personal information deserves a very high level of protection. I totally agree with the EU and the UK on that, although I think both have been too willing to compromise on principles in order to get anywhere with the US where there is no meaningful privacy at all.
(I find it sad that in America, private property is often guarded with deadly force, but private property is replaceable, whereas privacy has no protection at all and privacy can never be replaced. Once privacy is lost, it is lost forever.)
Exactly. Although I now see I could do the same with a prism that split out the red, green and blue, using three different cameras. However, this would seem to be a more expensive option and the prism must absorb some of the light energy. Nonetheless, the prism method seems to be the popular method for very high-end photography, to judge from the companies selling the components. Thoughts on how the two methods would compare in practice would be appreciated.
16bpp is for a single pre-processed image, so only gives you one colour plane, as you are filtering out unwanted planes. This means you actually have 48bpp for the post-processed image, 16bpp is merely the capability of typical device,so we work round that by only capturing part of the range at a time.
If, in 1915, you could take superb photos of the natural world with clunky colour filters, then a pinwheel on a stepper motor should be vastly superior. The device needs only to respond 3 times faster than you need to reduce blur to acceptable limits. (If you were using film, this would mean using 400 film in situations where 100 would be fine.)
Engineering-wise, one click of the button takes three equal-duration photographs. The stepper motor is triggered such that each photograph is through a different filter. Thus, you have three photographs, not one, and must use photographic or electronic techniques to recombine the images.
You could use a Dichroic prism to achieve a similar result, provided the output from each surface goes to a CCD of high enough resolution. This is how many high-end CCD devices work. It achieves the same result (recording each colour plane as a seperate and independent channel) and doesn't suffer the speed problem but does lose some light intensity from absorbtion and also requires manufacture to finer tolerences.
The CCD wikipedia page states: There are some high-end still cameras that use a rotating color filter to achieve both color-fidelity and high-resolution. These multi-shot cameras are rare and can only photograph objects that are not moving.
I question this, as CCD is much faster than film from 1916, never mind the filter change times, and yet the library of congress photos from 1916 include riverbank scenes in which moving water is only marginally blurred. It may be limited to slow scenes, but stationary seems excessive.
Alternatively, you could place an intensifier in front of the CCD, reducing its response time to a couple of hundred picoseconds. In turn, this reduces the time for the rotating filter method to a massive 600 picoseconds. Short enough even for Indicar.
"True" (or at least a damn-good facsimilie) of high-resolution HDR photography is clearly possible with some engineering skill and a good camera. Out of curiosity, I wondered just how far you could go, if you had the money. Atmel sell a CCD (AT71201M) that supports a resolution of 4096x4096. No indication on dynamic range or speed. For a landscape, you wouldn't necessarily care. 16 megapixels isn't as much as the gigapixel project but is still damn nice. Their TH7888A offers a very nice dynamic range (10,000dB) and a superb operating temperature range (not space-worthy, but close), and a superb data rate. The resolution is a crappy 1024x1024 (ie: 1 megapixel) though. Atmel has since sold these designs to e2v, but e2v's website is a mess and I can't find their new names or how they compare with e2v's other products. I did find one e2v product of interest - CCD48-20 - which operates in a range of -120 to +50 celcius. It's only a megapixel resolution, though. The spetral range (200nm - 1100nm) is extremely nice.
Fairchild's CCD486 is a full 16 megapixel device, has a low-end operating temperature suitable for space or Alaska, and a very respectable dynamic range. Hamamatsu produce two interesting CCDs - the S7030-1008 and the S7031-1008. The resolution is roughly half a megapixel, but they see clearly into both the ultraviolet and infrared. To give you an idea, the minimum wavelength they respond to is 200 nm, placing it right in the middle of the UV range. An maximum of 1200 nm is the majority of IR-A. It would be hard - and expensive - to build up the resolutions of these devices,
If you use a high-res 16bpp b/w digital camera, you can produce "true" HDR images by using the same technique as an early Russian photographer - simply rotate between red, green and blue filters. You now have a 48bpp colour image. If you now apply the 3D techniques, you would get a far more realistic 3D image (as you have far better data to work with).
Remember that device resources have to be scheduled. Make sure that your Relational Operating System isn't suffering from interrupt exhaustion. You should probably verify that your signal handlers are compatible, or the handshake won't go through correctly. Some devices appear to have a shared buffer, so input to them is rejected if they have not completed a communications transaction. Others seem to operate over transactional TCP, rejecting any session that is not complete to specifiction. So long asthe I/O is RFC-compliant, you shouldn't have any issues.
I thought everyone knew that hotnwetware operates strictly on a point-to-point protocol, and that broadscast protocols were inefficient on bandwidth and were less stable.
Electoral body generates a pool of public/private key pairs.
As each key pair is generated, a voter not yet allocated a key is associated with the key pair.
The public key is attached in electronic form to the voter registration card for that voter.
Some other Department is sent an electronic request for a cryptographic hash of some repeatably testable biometric data for that voter.
The public key is used to digitally sign the biometric hash.
The private key, plus hash and signature, is stored in a tamper-proof data repository where the electoral body has write-only access.
The voter is sent a device for measuring the repeatable biometric data, or can obtain one from an independent group.
The voter records their biometric data. The values are cryptographically hashed and are not retained.
The voter inserts the public key store into the device, which copies the hash onto it.
On voting (whether on an issued PDA or at a voting station), the voter inserts their public key store and records their vote.
The voting machine encrypts the vote using the public key and digitally signs it with the hash.
The electronic ballot is electronically delivered completely intact to the counting centre, and optionally to a neutral monitoring facility, immediately after being digitally signed.
On receipt, the vote+signature is itself cryptographically hashed and a table of these hash values is publicly exposed.
Also on receipt, the total number of received votes is incremented.
Also on receipt, since the signature is unsalted, duplicate votes by the same voter will show as identical signatures and can be dropped. The total number of duplicate votes is then incremented.
The voter can, at any time, use a computer that knows the encryption algorithm and can read the public key and hash:
Regenerate the ballot they cast
Encrypt and sign that ballot
Cryptographically hash the ballot
Search the list of published hashes to determine if any listed vote matches the regenerated ballot
Both the official organization and the monitoring organization can, at the time of counting:
Use the signing algorithm with the biometric hashes to determine which hash belongs to which vote (first vote the hash belongs to "wins")
Decrypt that vote using the corresponding private key and mark the vote with a tag to say which key was used
Discard both private key and biometric hash
Repeat until no remaining keys decrypt any remaining votes
Validate that votes decrypted + keys remaining = total number of registered voters
Validate that votes decrypted + votes remaining = votes cast
Count votes from decrypted ballots
Validate that the totals for each recipient are the same
If they are NOT the same:
Run through the discarded keys on each of the remaining votes
If a vote decrypts, then the voter has voted twice. This vote's tag will match a different vote elsewhere. The tag of both votes will be marked as suspicious.
Create a max/min range for each candidate, based on the maximum and minimum they could have scored, once the extra votes are considered.
The max/min ranges established by the official body and the monitor should be identical, and both of their initial tallies should be within those ranges.
How the suspicious votes are considered should be determined by a policy agreed nationally by both parties and voters. Otherwise, all suspicious votes should be rejected and therefore the minimum score for each candidate should be the official score.
An Internet-accessible site run by the official election group should list the results, but also provide a searchable table of just the digital signatures of the biometric hashes for the votes deemed valid.
That is "for nerds" (such as four more bands releasing freely-downloadble albums, with one band intentionally uploading to the pirate bay) gets relatively little coverage.
Hell, they've got a server, all they need to do is multihome it and have the second webserver called wikinews, or something. Declare it a sister site for things that are major but not technically leaks.
Didn't he also have a Law of SciFi that said that science fiction should be true to science known, bar any one law that you must violate in a consistant and explained manner?
I have to say that geek writing is vastly superior to academic writing. Academia teaches methods that are astonishingly unclear. Sir Humphrey would be proud of how papers are written. Your writing, as is almost always the case with Slashdot posters who take the time to write serious stuff, is vastly superior in style. (This is one reason I'm surprised geeks don't often band up and do science journalism. They could do so with vastly greater accuracy than science journalists currently do, and with vastly superior presentation than academics ever manage.)
His attempts at nuclear fusion research by accelerating chairs to a high velocity have inspired countless thousands of comic strip writers, and I think it likely that he will always be remembered for this most valuable contribution to society... no matter how hard he tries to make us forget...
So, at least in those two States, we can be certain that even the State government ordering a test of voting machines is likely to be declared illegal under the laws as they stand. If I ever move to either, I'm most certainly not voting on unaudited computers. Hell, I'd prefer an honest dictatorship over what would essentially be a dishonest de-facto dictatorship. That's probably why the British who've stayed in Britain are fine with the draconian system there. However perverse it gets, it's honest and open about it. Honest, sincere depravity.
You've got me curious - what was the method you were thinking of? Particularly in noisy images, my method may produce an impractical set of sets of possible results, simply because the uncertainty is too high. In such a system, a simpler method that is less influenced by noise may be far more suitable.
What I think you're saying is that people will themselves reflect what they believe the logo reflects when they subliminally see the logo. The association becomes a suggestion. You can't control the association (and therefore cannot control the suggestion) but what can be controlled is how blurred the line is between the two. Would that be correct?
After the DotCom crash, there were a lot fewer Java and ASP programmers getting hired, although that was more of a blip in the job market. Well, to a degree. There's much more demand for J2EE programmers than pure Java developers these days, with the rise of servlet engines such as Tomcat. C# and.NET have caught on very fast, I don't see quite as many pure C positions as I used to, and I can't remember the last time I saw a Visual Basic job in the area where I live. Visual Basic was a major language 5 years ago.
In general, though, yes, there is a lot of stability in job requirements over time. Certainly enough to be sure that if you learned all that is current now, you'd know almost all that was current five, ten, fifteen years down the road. Those who only learn two or three languages have the most need to speculate, as they're at greatest risk of getting left behind, but those who've mastered the bulk of common languages today are in excellent shape for the future.
Nonetheless, I believe the advice holds true. If you keep learning, you will be the stronger for it, and if you second-guess correctly which way the market will swing in the future, you will be the more secure for it.
Exactly. There are mathematical models which allow you to determine the optimum values for those variables, but you've got to know what the variables are and what constrains/impacts them before you can model them. That's daunting enough for engineers, I seriously doubt politicians can figure out the ideal balance point. (There may, in fact, be multiple ideal balance points, where different combinations of variables produce an impact of I for some group of people P, such that IxP is the same for all those balance points.)
Common sense says that for any function, there must be some number of points (between 1 and infinity) where that function is at its minimum value. It might be that the minimum is achieved when there is no recycling (x=0), it might be at some other point or points, and if the function turns out flat, it might be at every point. I believe that at least one minimum exists when x!=0, but without carrying out the analysis, I can't know for certain, but until someone does the maths, I can be certain that nobody else is any more certain about what x should be than I am.
My understanding was that he wrote sections of the book alongside the movie, making the script/book a joint effort, although the book was actually finished and polished later. Well, the only two people who know for certain are now working on a prequel (not available on Earth), from the Monolith's perspective.
Slashdot Mirror
But it's a good name given the work he put into the theory of geostationary satellites. However, I would like to make a suggestion: that where a marker, be it physical or electronic, is clearly designed with the intent of indicating the presence of alien life in the galaxy, the information type be measured in a ten-point scale of Clarkes. (A basic signal is one Clarke, up to a sufficiently advanced signal that makes it indistinguishable from magic, which would be ten Clarkes. In honour of Arthur C. Clarke's series on the supernatural, complex numbers are permitted.)
The answer, to me, is that F/L/OSS is charity, a charity that produces information the same way the above charity donating to a library produces information, and is a charity that turns a bunch of metals and chemicals into a finely-honed computing tool, the same as the above charity created a park. What we do is indeed charitable, not because we deprive ourselves, but because we enrich others. The cost to ourselves is zero, because we would have scratched our itches anyway. You can't rationally add as a cost of sharing the cost of pleasing ourselves.
Charity obviously allows for return on investment, it just means that others also get a return on your investment. But it doesn't require that others give any kind of feedback at all. If you make a public park and only you visit, it's still public, it was still an act of charity, but it's an act of charity you get exclusive benefit from.
Microsoft's statement, then, is a dark one indeed. No charity, of any kind? It says that they gain no pleasure in the results of their labour, that they suffer with every release, that every enhancement and refinement is a source of pain. Quality must be endless torment (which would explain some things). It is a bleak future when everything is misery and there is an apparent determination to spread that misery.
If they wanted to spread even just contentment, through their freely-donated hot-fixes, patches and service packs, freely-donated Microsoft Research products and freely-donated e-mail service and instant messenger, they'd be guilty of charity. Since they have denounced the charitable and all their works, these things cannot be given for the use of others. But, if they are not usable, even in theory, what are they? Microsoft's comments deride and slander all who would offer service to others, so the only conclusion is that these things are intended to cause suffering and misery, which - to judge by Vista service pack 1 - is indeed what they cause.
(This is why most people stick with SIMD for local computing and MISD for grids.)
Those using MPI for parallelization are probably (hopefully) aware of PAPI, TAU, KOJAK and DAKOTA. The first two supply accurate timings to profiling software, the other two are profiling packages for MPI environments.
Unfortunately, it is limited to MPI. If you use PVM or BSP, I know of nothing similar. Likewise, if you do a remote memory-to-memory copy (skipping the CPU), profiling is of limited use. And as most MPIs use sequential operations to do collective calls, any idiot could replace an MPI collective send with a reliable multicast. Infiniband and iWarp support it. I doubt the profilers do.
This means you can't rely on high-schol techniques. You've got to plan what happens, when, and how that is to be communicated.
Ah, yes, communication. A simple cluster will work fine with a simple star topology. The bandwidth requirements rise expontially and you are likely to be forced to go with one of the Big Three: Fat Tree (masses of hotspots, prone to failure but cheap), Butterfly (medium risk of failure if the software is imabalanced, and almost any MPI program isn't just unbalanced but stark raving, Hypercube (low risk of failure, very popular with researchers, very stable, very long latencies and very high price).
You also need to weigh network performance and what price that performace is at:
Now most poor research centers can't aford a hypercube topology with 12 GB/s links onto machines capable of digesting that much data. So, instead, they learn to program efficiently, much as the old assembly programmers (like myself) had to do with sparse resources. If you want to get ahead, keep the poverty mindset but exploit the raw computing power before you.
So to discover that other nations have just walked by and left the EU and US in the dust is a little annoying, but let's face it. Their resources were kept minimal by the west, so it's no surprise they learned the golden rule of boolean logic. waste not, want nots. C'mon, don't pretend we didn't expect this. OpenMP wasn't developed for the fun of it. Parallel has been on the way in for 25+ years. 25 years of Moore's Law applying to their work. You can catch up if you like. It will mean less visible glory, but it would mean doing something real.
I know that only a handful of these have been implemented for Linux or *BSD, even fewer for both. Instead of Summer of Code producing stuff nobody ever sees, how about one of the big players invest in students producing some of these meaty chunks of code?
Schemes for reducing packet loss by active queue management: REM, RED, GRED, WRED, SRED, Adaptive RED, RED-Worcester, Self-Configuring RED, Exponential RED, BLUE, SFB, GREEN, BLACK, PURPLE, WHITE
Schemes for adjusting packet queues: CBQ, Enhanced CBQ, HFSC, CSFQ, CSPFQ, PrFQ, Local Flow Separation,
Schemes for scheduling traffic: Gaussian, Least Attained Service, ABE, CSDPS
Schemes for shaping traffic flows: DSS, Constant bit Rate
Schemes for bandwidth allocation: RSVP, YESSIR, M-YESSIR
Schemes for active flow control: ECN, Mark Front ECN
Schemes for managing queues: Adaptive Virtual Queue, PRIO
(I find it sad that in America, private property is often guarded with deadly force, but private property is replaceable, whereas privacy has no protection at all and privacy can never be replaced. Once privacy is lost, it is lost forever.)
You understand the reason they're pulsating? After dropping Thrud the Barbarian, there has been no stabilizing influence.
The only result that surprised me was: PezCycling News - What's Cool In Pro Cycling
Exactly. Although I now see I could do the same with a prism that split out the red, green and blue, using three different cameras. However, this would seem to be a more expensive option and the prism must absorb some of the light energy. Nonetheless, the prism method seems to be the popular method for very high-end photography, to judge from the companies selling the components. Thoughts on how the two methods would compare in practice would be appreciated.
If, in 1915, you could take superb photos of the natural world with clunky colour filters, then a pinwheel on a stepper motor should be vastly superior. The device needs only to respond 3 times faster than you need to reduce blur to acceptable limits. (If you were using film, this would mean using 400 film in situations where 100 would be fine.)
Engineering-wise, one click of the button takes three equal-duration photographs. The stepper motor is triggered such that each photograph is through a different filter. Thus, you have three photographs, not one, and must use photographic or electronic techniques to recombine the images.
You could use a Dichroic prism to achieve a similar result, provided the output from each surface goes to a CCD of high enough resolution. This is how many high-end CCD devices work. It achieves the same result (recording each colour plane as a seperate and independent channel) and doesn't suffer the speed problem but does lose some light intensity from absorbtion and also requires manufacture to finer tolerences.
The CCD wikipedia page states: There are some high-end still cameras that use a rotating color filter to achieve both color-fidelity and high-resolution. These multi-shot cameras are rare and can only photograph objects that are not moving. I question this, as CCD is much faster than film from 1916, never mind the filter change times, and yet the library of congress photos from 1916 include riverbank scenes in which moving water is only marginally blurred. It may be limited to slow scenes, but stationary seems excessive.
Alternatively, you could place an intensifier in front of the CCD, reducing its response time to a couple of hundred picoseconds. In turn, this reduces the time for the rotating filter method to a massive 600 picoseconds. Short enough even for Indicar.
"True" (or at least a damn-good facsimilie) of high-resolution HDR photography is clearly possible with some engineering skill and a good camera. Out of curiosity, I wondered just how far you could go, if you had the money. Atmel sell a CCD (AT71201M) that supports a resolution of 4096x4096. No indication on dynamic range or speed. For a landscape, you wouldn't necessarily care. 16 megapixels isn't as much as the gigapixel project but is still damn nice. Their TH7888A offers a very nice dynamic range (10,000dB) and a superb operating temperature range (not space-worthy, but close), and a superb data rate. The resolution is a crappy 1024x1024 (ie: 1 megapixel) though. Atmel has since sold these designs to e2v, but e2v's website is a mess and I can't find their new names or how they compare with e2v's other products. I did find one e2v product of interest - CCD48-20 - which operates in a range of -120 to +50 celcius. It's only a megapixel resolution, though. The spetral range (200nm - 1100nm) is extremely nice.
Fairchild's CCD486 is a full 16 megapixel device, has a low-end operating temperature suitable for space or Alaska, and a very respectable dynamic range. Hamamatsu produce two interesting CCDs - the S7030-1008 and the S7031-1008. The resolution is roughly half a megapixel, but they see clearly into both the ultraviolet and infrared. To give you an idea, the minimum wavelength they respond to is 200 nm, placing it right in the middle of the UV range. An maximum of 1200 nm is the majority of IR-A. It would be hard - and expensive - to build up the resolutions of these devices,
If you use a high-res 16bpp b/w digital camera, you can produce "true" HDR images by using the same technique as an early Russian photographer - simply rotate between red, green and blue filters. You now have a 48bpp colour image. If you now apply the 3D techniques, you would get a far more realistic 3D image (as you have far better data to work with).
Remember that device resources have to be scheduled. Make sure that your Relational Operating System isn't suffering from interrupt exhaustion. You should probably verify that your signal handlers are compatible, or the handshake won't go through correctly. Some devices appear to have a shared buffer, so input to them is rejected if they have not completed a communications transaction. Others seem to operate over transactional TCP, rejecting any session that is not complete to specifiction. So long asthe I/O is RFC-compliant, you shouldn't have any issues.
I thought everyone knew that hotnwetware operates strictly on a point-to-point protocol, and that broadscast protocols were inefficient on bandwidth and were less stable.
That is "for nerds" (such as four more bands releasing freely-downloadble albums, with one band intentionally uploading to the pirate bay) gets relatively little coverage.
Hell, they've got a server, all they need to do is multihome it and have the second webserver called wikinews, or something. Declare it a sister site for things that are major but not technically leaks.
Didn't he also have a Law of SciFi that said that science fiction should be true to science known, bar any one law that you must violate in a consistant and explained manner?
I have to say that geek writing is vastly superior to academic writing. Academia teaches methods that are astonishingly unclear. Sir Humphrey would be proud of how papers are written. Your writing, as is almost always the case with Slashdot posters who take the time to write serious stuff, is vastly superior in style. (This is one reason I'm surprised geeks don't often band up and do science journalism. They could do so with vastly greater accuracy than science journalists currently do, and with vastly superior presentation than academics ever manage.)
His attempts at nuclear fusion research by accelerating chairs to a high velocity have inspired countless thousands of comic strip writers, and I think it likely that he will always be remembered for this most valuable contribution to society... no matter how hard he tries to make us forget...
So, at least in those two States, we can be certain that even the State government ordering a test of voting machines is likely to be declared illegal under the laws as they stand. If I ever move to either, I'm most certainly not voting on unaudited computers. Hell, I'd prefer an honest dictatorship over what would essentially be a dishonest de-facto dictatorship. That's probably why the British who've stayed in Britain are fine with the draconian system there. However perverse it gets, it's honest and open about it. Honest, sincere depravity.
You've got me curious - what was the method you were thinking of? Particularly in noisy images, my method may produce an impractical set of sets of possible results, simply because the uncertainty is too high. In such a system, a simpler method that is less influenced by noise may be far more suitable.
What I think you're saying is that people will themselves reflect what they believe the logo reflects when they subliminally see the logo. The association becomes a suggestion. You can't control the association (and therefore cannot control the suggestion) but what can be controlled is how blurred the line is between the two. Would that be correct?
In general, though, yes, there is a lot of stability in job requirements over time. Certainly enough to be sure that if you learned all that is current now, you'd know almost all that was current five, ten, fifteen years down the road. Those who only learn two or three languages have the most need to speculate, as they're at greatest risk of getting left behind, but those who've mastered the bulk of common languages today are in excellent shape for the future.
Nonetheless, I believe the advice holds true. If you keep learning, you will be the stronger for it, and if you second-guess correctly which way the market will swing in the future, you will be the more secure for it.
Common sense says that for any function, there must be some number of points (between 1 and infinity) where that function is at its minimum value. It might be that the minimum is achieved when there is no recycling (x=0), it might be at some other point or points, and if the function turns out flat, it might be at every point. I believe that at least one minimum exists when x!=0, but without carrying out the analysis, I can't know for certain, but until someone does the maths, I can be certain that nobody else is any more certain about what x should be than I am.
My understanding was that he wrote sections of the book alongside the movie, making the script/book a joint effort, although the book was actually finished and polished later. Well, the only two people who know for certain are now working on a prequel (not available on Earth), from the Monolith's perspective.