Faster than serial?! Of course! I only meant to compare it to a traditional parallel procssing environment. And you can definitely write a simple parallel algorithm for any O/PDE that will work on GPUs. What I meant was that there are an awful lot of claims about how wicked fast GPU processing can be. Some people tout it as much faster than traditional computing. This can be true, but to get a GPU to actually perform at that level, it requires particular structure to your data. Unstructured meshes are known to be particularly nasty. Doesn't mean you can't compute anyway. It just may or may not be any better than traditional methods.
I don't mean to poo-poo GPU computing in general. I admittedly haven't followed this field closely in a year or two, so it's possible there have been some newer agorithms for unstructured meshes that have improved the situation. And without knowing more about your particular problem, I won't speculate and tell you how it should or shouldn't work. Maybe you figured out a decent implementation on your own. In which case, publish it already!
"There is almost no chance that a $100 board can be designed to have a memory interface that can keep 64 cores well fed at this point in time. "
I agree with you 100% on that. If the cache isn't terrible, it might be okay if you have a problem amenable to openMP. But mainly I view these low-end things as kind of fun toys.
That said, there is a market for something reasonably compact and affordable in between a 4-8 core desktop and a large scale cluster. I occasionally test and debug problems on my desktop that seem to work fine, but when I scale it to 200 processors and put it on the cluster, all hell breaks loose and it can be hard to debug. A cheapo 64 core board, even if slow, could help bridge that gap, assuming I can use mpich/openMP on this thing.
Otherwise it is for hobbyists or as a learning tool.
There is a high barrier to entry for piddling around with graphic cards. Fortunately, most home computers are already parallel (2-8 cores). I do extensive parallel programming, and I do most of the testing (for small problems, anyhow) on my desktop or laptop, which each have 8 cores.
There is simply a set of "parallel" function calls which can be built directly into your code. You then just need to compile your code with the proper libraries, usually either mpich or OpenMP. I believe both are available in the ubuntu repository. Pick the one that is most promising for your problem. They are fundamentally two distinct approaches to parallelism, each useful at times. Lawrence Livermore has a great tutorial site, including loads of example fortran and C codes of both openMP and mpich. THey are ready to compile and run on your home computer. Happy computing!
That, and GPU computing really only gives that kind of performance for a few types of problems. Namely, if you are able to structure your data arrays in memory in such a way that a GPU can operate on it efficiently. If you are solving nasty PDEs on an unstructured mesh, it's very difficult to do this. In that case, a GPU is pretty worthless. I don't know how these parallella boards work, but hopefully they would be a bit more versatile.
The post office isn't anonymous? I put stuff in mail boxes all the time without a return address.
I agree in principle with you. I think an easy method of remaining anonymous in communications can only help a democracy. I have two issues though:
1) I don't see anonymity as being legitimately threatened. Pre-paid phones exist. Mailboxes don't ask for ID. Newspapers can print an editorial under a psuedonym. You can pay for things in cash. For home internet, there are legal and effective tools like TOR which can effectively anonymize your internet traffic. I could go on. You can do pretty much anything you want to anonymously, except perhaps use Facebook. It's certainly not automatic, but then again, it was never assumed by default anyhow.
2) While I agree it is a nice feature, I still don't see why it is necessary, which is what the GP claimed. You use state-imposed censorship/repercussions as an argument for its necessity, but that is exactly what the First Amendment guarantees against. The use of anonymity to evade state harassment is only necessary in a society without a guarantee like the first amendment.
Going one step further, I have gotten into discussions with people who argue that I should be afraid of what the government might do if they one day decide they don't like my opinions, and that's why anonymity is still necessary. But I think this type of fear is exactly the type of fear that the first amendment frees us from having to experience. If you have to worry about being targetted for your views, and fear reprisal, what the heck is the point of the first amendment? Freedom means freedom from that fear. Period.
A last thought: Supposing the government suddenly starts throwing people in jail or otherwise harassing people for their political opinions, and the courts are somehow silenced. What good does a "right" to anonymity give you? Do you really think they wouldn't be listening to phone calls in any case? It's not worth anything more than the "right" to free speech which is being ignored.
The GP discussed the Bill of Rights. The last time I checked, this is an American document. It was claimed that the Bill of Rights protects anonymity in speech. I refuted that claim, and asked him why he thought that. If he were to cite a court decision where this concept is upheld, I would absolutely change my mind. If he presented a cogent argument about why it is necessary or otherwise implied, I would certainly consider it.
You could have presented such an argument as well, but instead chose to go with the nationalistic attack. I'm sure whatever country you're from is full of very enlightened people.
There's a difference between expressing an opinion or idea (protected speech), and reneging on a contract you voluntarily signed that forbids you from disseminating particular information (not protected speech). Snowden was completely free to rail against the idea of the government collecting phone records. He instead disseminated information about the government's activities. There is an important difference.
You can feel about Snowden however you want, and I don't care. Traitor or hero, it's your opinion, and it's perfectly valid. All I will say is that he knowingly violated the law because he thought it was the right thing to do. Sometimes it is the right thing, and it's open for discussion. But arguing that what he did was covered by free speech is factually wrong. Unless you equally think that someone at a government lab should be able to post blueprints for nuclear weapons without any consequences, because you know, free speech and government transparency.
Maybe you're right. I have thought for awhile now that the Republican party is regrettably in a civil war, and is becoming unviable. I was once a registered (R), but no longer. And with every major election cycle, the average Republican candidate seems to get worse. To be fair, there will always be loonies in both parties so long as loonies continue to exist. The thing is, I think an intellectually healthy party controls such members effectively. You don't see some hippy named Moonbeam that wants to entirely ban guns and force non-gluten diets on everyone win very many primaries within the democratic party. Especially for major office. But the cartoonish, knee-jerk, pseudo-anarchist that wants to privatize highways seem to be gaining traction on the right. They haven't totally gotten control of the party, but they are definitely on a major upswing, and it's destabilizing the whole party.
And yet, the party persists. They may very well have a majority in both chambers in 2015. They have a solid base that will continue to keep them competitive for the forseeable future. They seem to be defying gravity, and I'm at a loss to explain how.
I'm curious why anonymity is necessary for free speech? Serious question.
The Constitution guarantees I can fundamentally express any idea I want to without fear of reprisals from the state, no matter how controversial or unpopular it might be. It does not guarantee, at least to my understanding, that I can express those ideas anonymously or without repercussions from my fellow citizens... If you feel anonymity is important (they wore hoods in the Klan, after all) you can do that. Take the bus, train, ride a bike, walk, etc, to your protest and don a white hood like they did in the old days. Don't bring your cellphone while your at it. Protecting your anonymity is on YOU, it's not a guaranteed right...
Anyway, it's a bit tangent to the main thread. I generally feel I have no expectation of privacy in a public area, and have no problem wih the idea of a "chip" that has the sole purpose of updating my registration automatically (some concerns about cost - what happens if it breaks?), but I'm also wary of implementing a system that could potentially be used or abused by a future administration.
Not the poster upthread, but as someone else who runs fluids codes on big machines, I will chime in:
A lot of the guys on the big NICS machines aren't using ANSYS. They're using their own research codes that are tailored for parallel performance and/or to solve specific and difficult problems that commercial codes don't do well, like fluid-structure interaction. I know there are guys that depend on licensing somehow or another and this is artificially limiting. But I never really understood it. If all you want is a basic, parallel fluids solver, there are some open-source options. Probably won't scale well, but it sure beats spending half your lab budget to get only 8 processors.
Even if you have your own in-house solver, you will of course run into problems with latency as you scale up. I usually run on around 100-200 processors, depending on the problem. I would love use more, but the communication costs start to take over. Some guys can run on 10-100,000 processors. Not sure what they are doing, but I am guess whatever they are computing requires very little communication between nodes, or has been optimized to an extreme degree. Hard to imagine those guys are running a normal fluids solver with an unstructured grid. That'd be a huge waste.
And I agree to whomever said that if someone know of a big wasted supercomputer with idle time on it, please advertise it here! All the ones I've ever seen are more-or-less utilized to their full extent.
I'm not an expert on constitutional law (or any law), but I don't think it would require much of Congress. Just a bill that says something like, "The USPS can manage itself as a completely independent entity until further notice by Congress". The USPS could then get into retail banking, or not, of their own accord at any point they so choose.
Surely all those people who talk so loud about ineffective government management would be all for a scaling back of congressional control of the Post Office, right? Right? Oh wait, they'd rather micromanage it into oblivion to make a political point. Maybe you're right.
I used to live in a stand-alone, single family home, in the middle of a major US city. The streets were gridded and well-marked, and my address was displayed. It was close to the heart of the city, not out in some suburban snake-pit of cul-de-sac roads.
I never had a problem with packages from USPS, or FedEx. But UPS... Oh holy god. I ordered something online, and waited. And waited. I checked the tracking only to see it was labelled as "undeliverable address" or something like that. I drove to some distribution point way out of my way to pick it up. I thought somehow I must have typed in a bad zip code or something, but when I picked it up, everything was correct. The driver just couldn't find it.
When it happenned a second time, I decided never to use UPS again, and damn the cost.
They're losing money for a variety of reasons. The most important is that they are mandated to exist by congress, and are supposed to be financially autonomous, but are micromanaged by congress. You'd have to think long and hard to come up with a worse group of PHBs. Congress told them to pre-fund in full their retirement fund for the next 75 years. The USPS has basically said, "This requirement is bankrupting us. If you relax it, or let us make our own decisions we'll be fine." People wonder why they can't compete with FedEx, UPS, DSL, etc, and the answer is simply that those companies don't have to listen to Congress dictate details like telling them to pre-fund the entirety of a 20 year-old employee's pension right now. I'm all for fiscal responsibility and responsible funding of pensions, but is ten years of secure pension funding not enough? 20? 30? I mean, 75? How do you even estimate your pension needs 75 years in advance?
On another note, one idea I've heard that was intriguing would allow them to operate something like a bank. Not a financial investment house, but a low-end and low-cost branch bank. Sure, I might not switch all my finances over to it, and most people probably wouldn't either. But I might open an account and seed it with some cash if it were convenient. I could send mail and have it draw on the account without having to buy stamps or wait in line. Just drop it in a box at the post office and enter my account number/pin. It could work really nicely. There's already a branch in every city. And for a lot of working poor that have no bank affiliation, it might be the most convenient place to open an account, reducing the population of unbanked. Basically a public option for retail banking.
Believe it or not, it works like this in most other places, mostly with success. And this is the way it used to work in the US as well, but it was not FDIC insured, and was phased out in the 60s or 70s.
Fair points, all. I have no problem with using LibreOffice myself, I find it works as good or better than Office for most things.
However, the lack of interoperability with MS Office is a major sticking point. You may be correct that this is mainly because of the multitude of crappy and proprietary file formats that MS puts out, but as a practical matter, MS Office is what most people use. When I have a client or my boss that asks me to send them a few power point slides, or someone sends me some powerpoint slides, I simply can't use LibreOffice, as much as I would like that. And I can't simply tell clients and authority types that "You're doing it wrong. Use this other software that I use!" I have to pull up Office in wine, or reboot. It sucks, but that's where it's at.
So, it may not really be LibreOffice's fault, but it's still a problem. Even for people who would like to use it.
Yeah, there are DOE, NSF, DOD, etc., grants that require US citizenship. And you can't be foregin and get an NSF Fellowship or something. But I can promise you... Almost no international STEM graduate students self-fund. They are all paid. And they account for over 50% of graduate students in these fields. Most of that funding is from US based organizations or agencies.
I didn't say anything about exploitation. I'm just saying that the immigration policy is inconsistent. They make it easy to come over and be a grad student, likely funded by a US organization. But they make it difficult to stay when you are done.
I am not really that concerned with the ultimate economics of the situation. I have no real problem if we, as a country, break even, make money, or end up in the red on international students. I have a problem with the strange policy of paying for a particular individual's education and then asking them to leave. I do not believe it is good national policy, and it also disrupts real human lives after they spend 4-6 years here getting an advanced degree, and are then often told to go home. In many cases, they have already started building a life here, and they want to stay. Why not make it easy for them? If they stay, they are likely to add to our economy and culture.
If they want to go home, I have no problem with that. It's not ultimately about how much money is spent/earned on international students. It's about having a sane and consistent immigration policy. An immigration policy that says: "graduate students that cost money GOOD, educated professionals BAD" is not consistent.
Sure, I guess. If I go to someplace in Germany, supported on German taxpayer money, the same problems may exist. (At least, that's how I interpreted your post...) But I have two problems with that:
1. Without bringing up specific examples of high-barrier-to-entry-for-residency-status-after-getting-a-free-education at other locations throughout the world, the argument is a strawman. Maybe other places do the same thing, but maybe they have different policies. I don't know.
2. Even if you come up with a slew of examples of how Germany also kicks out American students that travelled there on a grant paid by the German government, it is irrelevant. The question is not: "What do other countries do?" It is: "What should we do?"
I'm not trying to advocate for less money being spent on foreign students. All I mean is that if there are educated, professional people out there who want to come to the US, then the US should make it easy for them to come in. They will add to the culture and economy. If they want to stay home, that's fine too. The point about where the money comes from and where they are educated is really secondary. But inviting people in, paying for their education, and then asking them to leave is the worst possible solution in my opinion.
If Germany also does this, then Germany should make it easier too. If I were educated in Germany at German expense, lived there for 4-6 years, and I wanted to stay and build a life for myself over there, I think they ought to let me. It wouldn't make much sense to kick me out after all that. But I don't know what they do, and it's kind of immaterial.
For undergraduate degrees, yes, we do. But the main point is for advanced degrees in STEM. For graduate students, yes, tuition is still charged. The university gets paid whether you are international or not. The question is: Who pays?
It may surprise you that most STEM graduate students don't pay for their own tuition. In fact, most get paid out of some grant money somewhere. So, in effect, the American Heart Association, or the National Institutes of Health, or the National Science Foundation, etc, etc, will pay a professor at a university to study a problem. The professor then hires a graduate student to work on said problem. The professor takes the grant money and pays the student's tuition and a small salary. So, in effect, US organizations and taxpayer dollars fund an overwhelming amount of international students. This is fine, the professors, universities, and various agencies want to attract the best talent, and it's a worldwide marketplace.
Now, the real kicker is that after they graduate with a masters or doctorate, we make it difficult for them to stay here if they want. There should be an easy path in place for recipients of advanced degrees at US universities to stay here if they want. There's not. An awful lot of them are sent back home against their will. So I ask you: What is the point of bringing someone to this country, funding their education, and then demaning that they return home?
I find the notion that reality conforming to a semi-obscure prediction market based in Ireland to be patently ridiculous. Something like an election is decided based on millions of individuals who go out and vote. An example from the most recent election in the US:
Every statitician worth their salt was predicting Obama as the likely victor. Nate Silver was only one, but he had the biggest microphone. There was also Andy Tannenbaum (electoral-vote.com), Larry Sabato (UVA center for politics), and even one with a conservative bent, but I can't recall the guy's name (electionprojection.com).
However, despite that, you'd NEVER have known that Obama was likely to win if you watched the news. The media drummed it up as the "closest election EVAR!". I have relatives who only watch the news, and they all told me afterwards that either they were relieved and surprised Obama won handily, or upset and surprised he won handily, based on their political preference. The bottom line: Reality matched what the polls said, not what everybody was told on the news. The media is an indisputably powerful force for changing perception, but it didn't seem to actually make the election any closer. So why should we expect something like intrade to be able to move the needle of reality? If intrade had predicted a Romney win due to some serious market manipulation, it would have been roundly mocked afterwards, and Obama bettors would have laughed all the way to the bank. So I guess I still don't really see what the issue is...
That said, if you have an academic paper that might explain the concerns in a more proper form, please feel free to send me a link. I'm an amateur in this field, and I know that my ideas can have gigantic holes in them.
I see. Well, in that case, he might be right, but I still doubt it... He posits then that the effect of the market manipulator will dilute across all markets. Okay, but so what?
I would argue that a heavily skewed market would then attract heavier volume, which will dilute the manipulators influence until we reach an equilibrium. Going with the same example, lets assume the two markets have closed the gap, and finally reach the same odds, say 5:1. That doesn't mean that equilibrium has been reached. That only means they will now more or less move in tandem (ignoring further manipulation) to the equilibrium, presumably 4:1.
Analogy: Suppose there is an active weather prediction market, and someone manipulated it heavily for whatever reason. You notice one morning that there are 2:1 odds that it will rain tomorrow. You check several markets, and they all have the same odds. You check, and the weatherman says there is a 90% of rain. Do you take the bet? I bet the markets see some big bets and heavy volume until the odds go up to approximately 9:1...
TL;DR: If the odds don't match what the "crowd" thinks, then the "crowd" will bet untill it does. The crowd will exploit the easy money left on the table by the manipulator, and if the odds aren't right, the "crowd" will continue to bet until it matches the groupthink, regardless of the manipulator.
Caveat: Of course, with deep enough pockets, manipulators can have an effect regardless of volume. That's true in any market. So again, I fail to see what the point is... It's certainly not peculiar to prediction markets.
Not disputing that two sites might have different odds. Just curious as to why it would be considered a problem? The disparaity should be self-correcting in at least the folloiwing two ways:
1) Savvy bettors will help even it out. If the odds are different, as you point out, a risk-free gain can be made by clever wagering. With free money on the table, people will write automated scripts to detect this sort of thing. They will bet as much as they can when this scenario happens, because hey, why not? This will in effect bring the odds closer together.
2) Typical bettors will help even it out. Suppose you think Obama is going to win. You're pretty sure of it. You check the two sites, and and see odds of 4:1 and 6:1, respectively. Why on EARTH would you place a bet on the 6:1 site? If you only have $100 to bet, it will go further on the 4:1 market. When this situation occurs, the average Obama bettors will flock to the 4:1 site, while the Romney bettors will go to the 6:1 site. The odds will converge.
If, in a scenario where someone has money to burn and continually manipulates the market, then these markets lose their predictive value, yes. However, in that case, there is easy money to be had... Might as well stop complaining and ride that train.
Slashdot Mirror
Faster than serial ?! Of course! I only meant to compare it to a traditional parallel procssing environment. And you can definitely write a simple parallel algorithm for any O/PDE that will work on GPUs. What I meant was that there are an awful lot of claims about how wicked fast GPU processing can be. Some people tout it as much faster than traditional computing. This can be true, but to get a GPU to actually perform at that level, it requires particular structure to your data. Unstructured meshes are known to be particularly nasty. Doesn't mean you can't compute anyway. It just may or may not be any better than traditional methods.
I don't mean to poo-poo GPU computing in general. I admittedly haven't followed this field closely in a year or two, so it's possible there have been some newer agorithms for unstructured meshes that have improved the situation. And without knowing more about your particular problem, I won't speculate and tell you how it should or shouldn't work. Maybe you figured out a decent implementation on your own. In which case, publish it already!
"There is almost no chance that a $100 board can be designed to have a memory interface that can keep 64 cores well fed at this point in time. "
I agree with you 100% on that. If the cache isn't terrible, it might be okay if you have a problem amenable to openMP. But mainly I view these low-end things as kind of fun toys.
That said, there is a market for something reasonably compact and affordable in between a 4-8 core desktop and a large scale cluster. I occasionally test and debug problems on my desktop that seem to work fine, but when I scale it to 200 processors and put it on the cluster, all hell breaks loose and it can be hard to debug. A cheapo 64 core board, even if slow, could help bridge that gap, assuming I can use mpich/openMP on this thing.
Otherwise it is for hobbyists or as a learning tool.
Sorry, forgot to also post link to the OpenMP tutorial:
https://computing.llnl.gov/tutorials/openMP/
There is a high barrier to entry for piddling around with graphic cards. Fortunately, most home computers are already parallel (2-8 cores). I do extensive parallel programming, and I do most of the testing (for small problems, anyhow) on my desktop or laptop, which each have 8 cores.
There is simply a set of "parallel" function calls which can be built directly into your code. You then just need to compile your code with the proper libraries, usually either mpich or OpenMP. I believe both are available in the ubuntu repository. Pick the one that is most promising for your problem. They are fundamentally two distinct approaches to parallelism, each useful at times. Lawrence Livermore has a great tutorial site, including loads of example fortran and C codes of both openMP and mpich. THey are ready to compile and run on your home computer. Happy computing!
Tutorial: https://computing.llnl.gov/tutorials/mpi/
Example codes: https://computing.llnl.gov/tutorials/mpi/exercise.html#Exercise1
That, and GPU computing really only gives that kind of performance for a few types of problems. Namely, if you are able to structure your data arrays in memory in such a way that a GPU can operate on it efficiently. If you are solving nasty PDEs on an unstructured mesh, it's very difficult to do this. In that case, a GPU is pretty worthless. I don't know how these parallella boards work, but hopefully they would be a bit more versatile.
Thank you. If that is how they ruled, I accept that anonymity is implied in the 1st amendment.
Do you have a link to that case?
The post office isn't anonymous? I put stuff in mail boxes all the time without a return address.
I agree in principle with you. I think an easy method of remaining anonymous in communications can only help a democracy. I have two issues though:
1) I don't see anonymity as being legitimately threatened. Pre-paid phones exist. Mailboxes don't ask for ID. Newspapers can print an editorial under a psuedonym. You can pay for things in cash. For home internet, there are legal and effective tools like TOR which can effectively anonymize your internet traffic. I could go on. You can do pretty much anything you want to anonymously, except perhaps use Facebook. It's certainly not automatic, but then again, it was never assumed by default anyhow.
2) While I agree it is a nice feature, I still don't see why it is necessary, which is what the GP claimed. You use state-imposed censorship/repercussions as an argument for its necessity, but that is exactly what the First Amendment guarantees against. The use of anonymity to evade state harassment is only necessary in a society without a guarantee like the first amendment.
Going one step further, I have gotten into discussions with people who argue that I should be afraid of what the government might do if they one day decide they don't like my opinions, and that's why anonymity is still necessary. But I think this type of fear is exactly the type of fear that the first amendment frees us from having to experience. If you have to worry about being targetted for your views, and fear reprisal, what the heck is the point of the first amendment? Freedom means freedom from that fear. Period.
A last thought: Supposing the government suddenly starts throwing people in jail or otherwise harassing people for their political opinions, and the courts are somehow silenced. What good does a "right" to anonymity give you? Do you really think they wouldn't be listening to phone calls in any case? It's not worth anything more than the "right" to free speech which is being ignored.
Ad hominem much?
The GP discussed the Bill of Rights. The last time I checked, this is an American document. It was claimed that the Bill of Rights protects anonymity in speech. I refuted that claim, and asked him why he thought that. If he were to cite a court decision where this concept is upheld, I would absolutely change my mind. If he presented a cogent argument about why it is necessary or otherwise implied, I would certainly consider it.
You could have presented such an argument as well, but instead chose to go with the nationalistic attack. I'm sure whatever country you're from is full of very enlightened people.
There's a difference between expressing an opinion or idea (protected speech), and reneging on a contract you voluntarily signed that forbids you from disseminating particular information (not protected speech). Snowden was completely free to rail against the idea of the government collecting phone records. He instead disseminated information about the government's activities. There is an important difference.
You can feel about Snowden however you want, and I don't care. Traitor or hero, it's your opinion, and it's perfectly valid. All I will say is that he knowingly violated the law because he thought it was the right thing to do. Sometimes it is the right thing, and it's open for discussion. But arguing that what he did was covered by free speech is factually wrong. Unless you equally think that someone at a government lab should be able to post blueprints for nuclear weapons without any consequences, because you know, free speech and government transparency.
Maybe you're right. I have thought for awhile now that the Republican party is regrettably in a civil war, and is becoming unviable. I was once a registered (R), but no longer. And with every major election cycle, the average Republican candidate seems to get worse. To be fair, there will always be loonies in both parties so long as loonies continue to exist. The thing is, I think an intellectually healthy party controls such members effectively. You don't see some hippy named Moonbeam that wants to entirely ban guns and force non-gluten diets on everyone win very many primaries within the democratic party. Especially for major office. But the cartoonish, knee-jerk, pseudo-anarchist that wants to privatize highways seem to be gaining traction on the right. They haven't totally gotten control of the party, but they are definitely on a major upswing, and it's destabilizing the whole party.
And yet, the party persists. They may very well have a majority in both chambers in 2015. They have a solid base that will continue to keep them competitive for the forseeable future. They seem to be defying gravity, and I'm at a loss to explain how.
I'm curious why anonymity is necessary for free speech? Serious question.
The Constitution guarantees I can fundamentally express any idea I want to without fear of reprisals from the state, no matter how controversial or unpopular it might be. It does not guarantee, at least to my understanding, that I can express those ideas anonymously or without repercussions from my fellow citizens... If you feel anonymity is important (they wore hoods in the Klan, after all) you can do that. Take the bus, train, ride a bike, walk, etc, to your protest and don a white hood like they did in the old days. Don't bring your cellphone while your at it. Protecting your anonymity is on YOU, it's not a guaranteed right...
Anyway, it's a bit tangent to the main thread. I generally feel I have no expectation of privacy in a public area, and have no problem wih the idea of a "chip" that has the sole purpose of updating my registration automatically (some concerns about cost - what happens if it breaks?), but I'm also wary of implementing a system that could potentially be used or abused by a future administration.
Not the poster upthread, but as someone else who runs fluids codes on big machines, I will chime in:
A lot of the guys on the big NICS machines aren't using ANSYS. They're using their own research codes that are tailored for parallel performance and/or to solve specific and difficult problems that commercial codes don't do well, like fluid-structure interaction. I know there are guys that depend on licensing somehow or another and this is artificially limiting. But I never really understood it. If all you want is a basic, parallel fluids solver, there are some open-source options. Probably won't scale well, but it sure beats spending half your lab budget to get only 8 processors.
Even if you have your own in-house solver, you will of course run into problems with latency as you scale up. I usually run on around 100-200 processors, depending on the problem. I would love use more, but the communication costs start to take over. Some guys can run on 10-100,000 processors. Not sure what they are doing, but I am guess whatever they are computing requires very little communication between nodes, or has been optimized to an extreme degree. Hard to imagine those guys are running a normal fluids solver with an unstructured grid. That'd be a huge waste.
And I agree to whomever said that if someone know of a big wasted supercomputer with idle time on it, please advertise it here! All the ones I've ever seen are more-or-less utilized to their full extent.
Your post just reminded me about the second time it happened. The exception they gave that time was: "No apartment number given"
Again, I lived in a single family home.
I'm not an expert on constitutional law (or any law), but I don't think it would require much of Congress. Just a bill that says something like, "The USPS can manage itself as a completely independent entity until further notice by Congress". The USPS could then get into retail banking, or not, of their own accord at any point they so choose.
Surely all those people who talk so loud about ineffective government management would be all for a scaling back of congressional control of the Post Office, right? Right? Oh wait, they'd rather micromanage it into oblivion to make a political point. Maybe you're right.
I used to live in a stand-alone, single family home, in the middle of a major US city. The streets were gridded and well-marked, and my address was displayed. It was close to the heart of the city, not out in some suburban snake-pit of cul-de-sac roads.
I never had a problem with packages from USPS, or FedEx. But UPS... Oh holy god. I ordered something online, and waited. And waited. I checked the tracking only to see it was labelled as "undeliverable address" or something like that. I drove to some distribution point way out of my way to pick it up. I thought somehow I must have typed in a bad zip code or something, but when I picked it up, everything was correct. The driver just couldn't find it.
When it happenned a second time, I decided never to use UPS again, and damn the cost.
They're losing money for a variety of reasons. The most important is that they are mandated to exist by congress, and are supposed to be financially autonomous, but are micromanaged by congress. You'd have to think long and hard to come up with a worse group of PHBs. Congress told them to pre-fund in full their retirement fund for the next 75 years. The USPS has basically said, "This requirement is bankrupting us. If you relax it, or let us make our own decisions we'll be fine." People wonder why they can't compete with FedEx, UPS, DSL, etc, and the answer is simply that those companies don't have to listen to Congress dictate details like telling them to pre-fund the entirety of a 20 year-old employee's pension right now. I'm all for fiscal responsibility and responsible funding of pensions, but is ten years of secure pension funding not enough? 20? 30? I mean, 75? How do you even estimate your pension needs 75 years in advance?
On another note, one idea I've heard that was intriguing would allow them to operate something like a bank. Not a financial investment house, but a low-end and low-cost branch bank. Sure, I might not switch all my finances over to it, and most people probably wouldn't either. But I might open an account and seed it with some cash if it were convenient. I could send mail and have it draw on the account without having to buy stamps or wait in line. Just drop it in a box at the post office and enter my account number/pin. It could work really nicely. There's already a branch in every city. And for a lot of working poor that have no bank affiliation, it might be the most convenient place to open an account, reducing the population of unbanked. Basically a public option for retail banking.
Believe it or not, it works like this in most other places, mostly with success. And this is the way it used to work in the US as well, but it was not FDIC insured, and was phased out in the 60s or 70s.
Fair points, all. I have no problem with using LibreOffice myself, I find it works as good or better than Office for most things.
However, the lack of interoperability with MS Office is a major sticking point. You may be correct that this is mainly because of the multitude of crappy and proprietary file formats that MS puts out, but as a practical matter, MS Office is what most people use. When I have a client or my boss that asks me to send them a few power point slides, or someone sends me some powerpoint slides, I simply can't use LibreOffice, as much as I would like that. And I can't simply tell clients and authority types that "You're doing it wrong. Use this other software that I use!" I have to pull up Office in wine, or reboot. It sucks, but that's where it's at.
So, it may not really be LibreOffice's fault, but it's still a problem. Even for people who would like to use it.
Yeah, there are DOE, NSF, DOD, etc., grants that require US citizenship. And you can't be foregin and get an NSF Fellowship or something. But I can promise you... Almost no international STEM graduate students self-fund. They are all paid. And they account for over 50% of graduate students in these fields. Most of that funding is from US based organizations or agencies.
I didn't say anything about exploitation. I'm just saying that the immigration policy is inconsistent. They make it easy to come over and be a grad student, likely funded by a US organization. But they make it difficult to stay when you are done.
I am not really that concerned with the ultimate economics of the situation. I have no real problem if we, as a country, break even, make money, or end up in the red on international students. I have a problem with the strange policy of paying for a particular individual's education and then asking them to leave. I do not believe it is good national policy, and it also disrupts real human lives after they spend 4-6 years here getting an advanced degree, and are then often told to go home. In many cases, they have already started building a life here, and they want to stay. Why not make it easy for them? If they stay, they are likely to add to our economy and culture.
If they want to go home, I have no problem with that. It's not ultimately about how much money is spent/earned on international students. It's about having a sane and consistent immigration policy. An immigration policy that says: "graduate students that cost money GOOD, educated professionals BAD" is not consistent.
Sure, I guess. If I go to someplace in Germany, supported on German taxpayer money, the same problems may exist. (At least, that's how I interpreted your post...) But I have two problems with that:
1. Without bringing up specific examples of high-barrier-to-entry-for-residency-status-after-getting-a-free-education at other locations throughout the world, the argument is a strawman. Maybe other places do the same thing, but maybe they have different policies. I don't know.
2. Even if you come up with a slew of examples of how Germany also kicks out American students that travelled there on a grant paid by the German government, it is irrelevant. The question is not: "What do other countries do?" It is: "What should we do?"
I'm not trying to advocate for less money being spent on foreign students. All I mean is that if there are educated, professional people out there who want to come to the US, then the US should make it easy for them to come in. They will add to the culture and economy. If they want to stay home, that's fine too. The point about where the money comes from and where they are educated is really secondary. But inviting people in, paying for their education, and then asking them to leave is the worst possible solution in my opinion.
If Germany also does this, then Germany should make it easier too. If I were educated in Germany at German expense, lived there for 4-6 years, and I wanted to stay and build a life for myself over there, I think they ought to let me. It wouldn't make much sense to kick me out after all that. But I don't know what they do, and it's kind of immaterial.
For undergraduate degrees, yes, we do. But the main point is for advanced degrees in STEM. For graduate students, yes, tuition is still charged. The university gets paid whether you are international or not. The question is: Who pays?
It may surprise you that most STEM graduate students don't pay for their own tuition. In fact, most get paid out of some grant money somewhere. So, in effect, the American Heart Association, or the National Institutes of Health, or the National Science Foundation, etc, etc, will pay a professor at a university to study a problem. The professor then hires a graduate student to work on said problem. The professor takes the grant money and pays the student's tuition and a small salary. So, in effect, US organizations and taxpayer dollars fund an overwhelming amount of international students. This is fine, the professors, universities, and various agencies want to attract the best talent, and it's a worldwide marketplace.
Now, the real kicker is that after they graduate with a masters or doctorate, we make it difficult for them to stay here if they want. There should be an easy path in place for recipients of advanced degrees at US universities to stay here if they want. There's not. An awful lot of them are sent back home against their will. So I ask you: What is the point of bringing someone to this country, funding their education, and then demaning that they return home?
I find the notion that reality conforming to a semi-obscure prediction market based in Ireland to be patently ridiculous. Something like an election is decided based on millions of individuals who go out and vote. An example from the most recent election in the US:
Every statitician worth their salt was predicting Obama as the likely victor. Nate Silver was only one, but he had the biggest microphone. There was also Andy Tannenbaum (electoral-vote.com), Larry Sabato (UVA center for politics), and even one with a conservative bent, but I can't recall the guy's name (electionprojection.com).
However, despite that, you'd NEVER have known that Obama was likely to win if you watched the news. The media drummed it up as the "closest election EVAR!". I have relatives who only watch the news, and they all told me afterwards that either they were relieved and surprised Obama won handily, or upset and surprised he won handily, based on their political preference. The bottom line: Reality matched what the polls said, not what everybody was told on the news. The media is an indisputably powerful force for changing perception, but it didn't seem to actually make the election any closer. So why should we expect something like intrade to be able to move the needle of reality? If intrade had predicted a Romney win due to some serious market manipulation, it would have been roundly mocked afterwards, and Obama bettors would have laughed all the way to the bank. So I guess I still don't really see what the issue is...
That said, if you have an academic paper that might explain the concerns in a more proper form, please feel free to send me a link. I'm an amateur in this field, and I know that my ideas can have gigantic holes in them.
I see. Well, in that case, he might be right, but I still doubt it... He posits then that the effect of the market manipulator will dilute across all markets. Okay, but so what?
I would argue that a heavily skewed market would then attract heavier volume, which will dilute the manipulators influence until we reach an equilibrium. Going with the same example, lets assume the two markets have closed the gap, and finally reach the same odds, say 5:1. That doesn't mean that equilibrium has been reached. That only means they will now more or less move in tandem (ignoring further manipulation) to the equilibrium, presumably 4:1.
Analogy: Suppose there is an active weather prediction market, and someone manipulated it heavily for whatever reason. You notice one morning that there are 2:1 odds that it will rain tomorrow. You check several markets, and they all have the same odds. You check, and the weatherman says there is a 90% of rain. Do you take the bet? I bet the markets see some big bets and heavy volume until the odds go up to approximately 9:1...
TL;DR: If the odds don't match what the "crowd" thinks, then the "crowd" will bet untill it does. The crowd will exploit the easy money left on the table by the manipulator, and if the odds aren't right, the "crowd" will continue to bet until it matches the groupthink, regardless of the manipulator.
Caveat: Of course, with deep enough pockets, manipulators can have an effect regardless of volume. That's true in any market. So again, I fail to see what the point is... It's certainly not peculiar to prediction markets.
Not disputing that two sites might have different odds. Just curious as to why it would be considered a problem? The disparaity should be self-correcting in at least the folloiwing two ways:
1) Savvy bettors will help even it out. If the odds are different, as you point out, a risk-free gain can be made by clever wagering. With free money on the table, people will write automated scripts to detect this sort of thing. They will bet as much as they can when this scenario happens, because hey, why not? This will in effect bring the odds closer together.
2) Typical bettors will help even it out. Suppose you think Obama is going to win. You're pretty sure of it. You check the two sites, and and see odds of 4:1 and 6:1, respectively. Why on EARTH would you place a bet on the 6:1 site? If you only have $100 to bet, it will go further on the 4:1 market. When this situation occurs, the average Obama bettors will flock to the 4:1 site, while the Romney bettors will go to the 6:1 site. The odds will converge.
If, in a scenario where someone has money to burn and continually manipulates the market, then these markets lose their predictive value, yes. However, in that case, there is easy money to be had... Might as well stop complaining and ride that train.