I was asked why, in my opinion, homebrew on hacked PS3's was worthless, I answered. I stand by it. It's not worth the effort. Exactly as you said, XNA is a better solution. The cell is expensive, complicated, and because the whole architecture is atypical you're looking at writing all of the tools without complete information from scratch. Sony has literally dozens of people who spent years writing it with full information. By the time you get anything useful as a development framework that will do much of anything the PS4 will be half way through its lifecycle.
And even then you're developing into a technology that's dead. Not because it's bad, but because it sits between the CPU and GPU in terms of what it does conceptually, and problems tend to lend themselves much better to either the CPU or GPU than to the cell. As a result even sony is jumping ship to ARM for their PSP2.
They've ported everything off the cluster I built, I have since gone to another institution. We build the cluster about 3.5 years ago, and ported everything to GPU computing 3 years ago, because a single 400 dollar GPU (nvidia specifically) outperformed the 4 machine cluster by about a factor of 2. it was a waste of our money and sony's. Now I do some PS3 development stuff (through official channels) and that's bad enough, trying to do it without all of the libraries they provide is an enormous undertaking from scratch, to accomplish a task better done a million other ways.
Don't get me wrong. I distinguish between Sony should still be supporting Linux, and Geohot's taking that into his own hands and then some. Sony sold it with PS3 linux, they shouln't take it out (at least on the machines it was sold with, new versions without it are fine). But they shouldn't let someone hack the system either, that violates the entire integrity and the point of the system in the first place.
Sure, a small handful of people bought those machines to hack them (again, separate from running linux officially), but that shouldn't give them free reign to damage the experience for everyone else.
Because with homebrew you'd have terrible access to the hardware, and the hardware isn't all that compelling (it's not any better than a PC, which is equally tied to your living room). There's no substantial media functionality that's missing, and anything that is missing is prohibitively expensive to add when it is, again, better done on something other than a PS3.
The Cell isn't all that great. Seriously, it isn't. You need a lot of money, time and people to make proper use of it. And then there's the GPU which you don't have great access to. If you want to homebrew up a game it's far easier to just use a PC. And if you want to do cell development, well, why? The only people actually doing anything with the cell are supercomputing clusters, and even they're abandoning cell left right and centre for GPU computing (which you can't do on the PS3's gpu).
Homebrew on a handheld is a whole other ball game. You can actually take that with you which you can't do with a PC. Even if you don't get access to decent tools or all the hardware being able to 'take it with you' is pretty compelling. Being able to play on my TV is a matter of buying a $20 cable for my PC.
In their defence, the whole purpose for the vast majority of users when buying a PS3 was to have a gaming machine and some semblance of fair competition in multiplayer. The forums are naturally more full of fanboys than anywhere else. Any PS3 hack directly interferes with the notion of a fair playing session against other people. As much as it got Sony free press, PS2 linux... I'm sorry PS3 linux was only ever there to try and skirt around EU import tariffs. It didn't work. Sony really doesn't want you using PS3's for astrophysics clusters or airforce research (I helped build a cluster for astrophysics work). They sell at a loss because they want you to buy games. I'm sure the airforce cluster is great press, but the vast majority of the research just costs them money they don't get a tax break for.
All MMO's go after botters. Shouldn't I be able to run whatever software I want on my own machine?
The olympics go after people who have too much cold medication or whatever else. Shouldn't you be allowed to take whatever your doctor recommends for your health?
The US congress thought steroid use in baseball was so important they dragged barry bonds and co. to washington to talk to them. This is when they are trying to deal with a trillion dollar deficit, they're willing to waste days of peoples time on steroids in baseball. And you think sony fanboys are overreacting to a hack? At least the sony fanboys are actually participating in, and affected by cheaters directly.
Not that Sony is blameless. They should never have allowed 'other OS linux' on the PS3 in the first place, they should not have gone around threatening to sue everyone under the sun who might have been intrigued by geohot. But if you're a MMO player, you're glad to see the banhammer go out to gold farmers, botters etc. Sony is trying to balance on one hand the developers and hardcore gamers (who roughly have aligned interests in terms of security), and hacker types who should be free to toy with their own stuff, but not at the expense of the network experience of everyone else.
Lets be honest. Sure, a PS3 jailbreak hack lets you run homebrew games, and may re-enable partially functional PS2 emulation, and brings back linux support. But it also lets you mess with the memory state of your machine and hack the game as you're running and it lets you pirated games. Homebrew has no real value on a PS3. PS2 emulation, I'm not sure on, I'm betting sony didn't just take it out for the fun of it, it's probably really hard, if not impossible to do properly and provide a good experience. Linux... well it shipped with linux so I guess they should still support it. But that goes into the next problem, which is the whole network experience of 40 odd million players (obviously not all of whom actually use the network functionality) can be easily disrupted by even a handful of people hacking which is really a serious problem. Avoiding that is sort of the point of having a console in the first place, and being able to steal stuff from the playstation store isn't exactly something I support either. I don't think pirated games (blu-ray/DVD) is a huge issue, though I could be wrong, I suppose once the hack is out here some cheap PS3 game knockoff manufacturers could pose problems if they want to get into that business.
of course we also train a lot more people too (both in absolute numbers and as a percentage of the population). In practice the details of law, and the depth of knowledge that define a law degree are no less than that required to define the philosophy of morality. The few that truly grasp the latter in addition to the former can become professors. Everyone else just gets a law degree.
The more complex the world the more time needs to be spent to understand it at any level. I don't think schools lack depth, in fact I would argue they have tremendous depth (and choice), but it is implicitly assumed that you both study and learn something about morality when you take a degree in law. Animation is a college not university thing, so I won't hit on that (that really is job training, there are Uni level comp sci courses on animation but that is an exercise in kinematics, physics and rendering, a somewhat separate problem again). Game design, which is where I live, is a whole other ball game. There's the 'arts' side of design, which incorporates psychology and writing, essentially 'how do I make it fun'. There's no painting involved, and it's much more about storytelling and pacing but with a game development twist (along with all of the things that make a game a game, death, violence, or lack thereof, complexity and so on). And then there is the 'science' side of design (which is what I do), which really is an exercise in computer science. How do you define and test balance? How do you quantify and assess all of those things the 'arts' designers need as tools? How complex is this problem? The broad philosophical question "what is fun" is certainly covered, a lot. But in practice you need to know 'how do I define fun' 'how do I know if it is fun' rather than just 'what is fun'. The devil is in the details as it were. I can write you a 40 page paper on what is fun, but it's worthless unless I can translate that into numbers in a program that runs and actually *is* fun. The standard has been raised from a general understanding to a specific one.
Virtually all of my students * ARE* well rounded already. Admittedly I see physics, math, chemistry, engineering and CS types and the occasional wayward business student. They have a grasp of, and an internal model of how the world behaves, where they fit in it, and how to learn and understand problems. They wouldn't be in university if they didn't. What they need is depth of understanding of a specific area. Everyone of them has some high minded idea of what morality is (and, if so inclined can easily find the classic texts on whatever your topic of choice is), but if you cannot apply those principles to solve real problems they have no value. Not because the principles themselves are valueless, they aren't, but because they all get the principles already, what they need is to understand the details of real problems.
It is the difference between the people who complained a lot about democracy in libya and got nice cushy faculty positions in the UK and the US from it, and the people who are now trying to actually build a system of government. Some of them are even the same people, but the latter are just as committed to the principles of democracy as the former, they just have to spend time building electoral districts, a parliamentary system and voting rules.
Don't kid yourself for one moment, the people learning the enormous depth of information required to actually solve problems understand and appreciate the world around them. That's why they are invested in learning to actually solve real problems at all. 140 years ago you needed to worry about people who 'got the big picture' much more than today. Teachers are better, information is more accessible (and cheaper). It's much easier to a make a student reasonably well rounded now by the time they're done high school. But of course nearly everyone graduates high school. In canada it's about 90% of people by the time they are 24 have completed high school, compared to the US in 19
the question becomes: is it the same greek as being asked here. In 1869 Greece was 20 years independent from ottoman rule and still basically a nothing state on the world stage (not that it's much of anything today). Greek as taught in modern greece (or in this case 1860's greece) is not necessarily the same as the various versions of greek that would pre-date the modern world. There are classical language courses you can take at some schools, including universities, but the version of greek harvard is asking about in 1869 probably has relatively little bearing on contemporary greek of that period let alone modern greek.
Today you can find greek in a smattering of countries, italy, turkey and greece being the big ones, but armenia, ukraine, cyprus and a few others as well, but no more than a person today could do well with old english (which is nearly unintelligible), or middle english (shakespeare and KJV bible era, which is somewhat comprehensible).
I think probably if you could pass that exam today you could probably still do well in most liberal arts programmes at least. It shows an ability to grasp foreign languages (always handy), and a relatively diverse reading set. Not that there wouldn't be better measures of success today though. I think a modern scientist faced with an exam from 140 years ago might have a lot of trouble. There's language, terminology and style advances, skills that have largely been obsoleted (by for example the calculator), and then well, we know more now than they did then of course. Even if the math is the same, the way the math is written has changed quite a lot (hello matrices!). Any test carries with it the context of its time, and there's probably a ripple effect. What is today a challenging entrance exam at harvard or a PhD level topic in programming will 20 or 30 years from now be pushed into the highschool curriculum, and then 20 or 30 years later it will likely be long forgotten as it is supplanted by new problems and techniques.
Ya, I mean, servers and phones aren't exactly areas where microsoft was the market leader, or if it was, never by any meaningful amount (servers). Saying linux is dominating the phone market, when their phone product is 2 years old is like oldsmobile claiming they won the car market in 1901... Whether or not MS-Nokia will be rolling in mobile money 2 years from now remains to be seen.
The server arguement is interesting. HPC was never MS's area. At all. They had probably 1% of the market 20 years ago, and probably have 1.5% of the market today (CUDA on windows being that 0.5% growth, the other 1% is people who don't know any better, as it was 20 years ago). Sure linux did a great job cannibalizing unix marketshare, but that's not beating windows. In terms of general servers.... servers for what? Giant distributed databases (say for websites), when was that ever any area MS dominated? They've done well in that space, and I've got a few windows servers here and there but not for anything important. I don't think they've ever been the company to go to for 5 9's of uptime. Nor have they really tried to be (although you can probably achieve that on windows boxes), supposedly some companies offer 6 9's of windows server 2008 uptime. But that's as much hardware as it is software.
Embedded devices... again. When was MS anything in the embedded devices space? Sure there's roughly parity on the 360 (windows) PS3 (linux) in terms of marketshare. Only hackers care what OS those machines use, everyone else just wants to play games, and don't really care about OS, and it's not like the xbox has a big 'Windows inside' sticker. Sure MS ran 'terminals' (for bank and ticket machines), but never really consumer electronics. I don't think they've ever wanted that market, and winning it 'from' microsoft, who aren't even competing there doesn't make a lot of sense.
The only area 'open' has destroyed MS is browsers. Which is because MS made a truly horrid browser and has been desperately trying to rectify that error and kill it off for years. Firefox is a vastly better product for just about everyone.
20 years ago MS dominated 2 big markets. Desktop operating systems and desktop office applications. And not even really office back in 91 (that is windows 3.0 days). Since then office has faced real competition from free (google docs) and the end of printing for everything, recently making pdfs being free and easy, and so on. I think Open Office and it's compatriots have kept MS prices down, but today paying 150 bucks for 3 home licences of office doesn't seem unreasonable, after all, the people who develop the software should maybe get paid for it. The home and business operating system market... is still owned by microsoft. Through and through. It may face competition from apple, but linux isn't even on the radar here. Unfortunately.
The other thing 'free' has done is tried to accustom us to not paying for the products we use. But being 'worth something' to advertisers who will pay the people who develop the products we use. It's sort of like regular TV. And not in a good way. Of course people prefer free with no strings attached to free with advertising. But that's beside the point. Whether you use linux or windows or mac the thousands of software developers who wrote that software are mostly getting paid, with MS you pay them for the product. With Mac you pay for the experience and with linux you get it for free because hopefully someone business will pay on your behalf, either directly or not.
Linux has done fantastically well for itself since 1991. But so has MS. And linux hasn't made any significant inroads on MS's main product portfolio. If anything MS has pushed into areas that are traditionally more unix/linuxy in servers than linux has pushed into windows areas on the desktop. And the mobile space seems a long way from being decided. It has however served as both a great relatively free source of anti-MS information (much of it deserved) and a good way to keep the monopoly from being as disastrous as it could be.
Every woman I have ever met is far more likely to forget to take pills than I am to forget to have condoms. Choosing to take the risk without one is not the same as forgetting to put one on.
In my (very limited) anecdotal experience women miss an average of one pill every 3 months. But two in a row I don't know the probability there.
In terms of preventing diseases, condoms would seem to be the only main option. I would doubt they are much better than hormones at contraception but they do deal with the disease problem.
Caveat: I don't know the error rate (or tolerance) for oral contraceptives, and I'm not even sure I've ever actually heard it. If the average woman misses say 1 pill a year, or 1 pill every six months or whatever the number is, what does that do to pregnancy risk compared to the supposed effects of taken optimally. Then you get into exposure to conflicting meds etc. which may have a similar problem. Quoting the expected error rate assuming you don't take conflicting meds, don't miss pills etc. isn't real world data on error rates. Condom error rates are probably pretty close to accurate (fall off & break are the main options I would think) neither of which count if you don't bother to wear one in the first place. You can't really 'forget' to wear a condom, well, maybe you can (sleep sex?), but I'm not really sure how that would be factored in, but you can forget pills. I suppose you could apply substances to a condom which degrade its quality (e.g. if you're trying to clean them for reuse, which I could see as an issue in the developing world), but afaik they aren't really designed to be reused.
Correlation only measures degree of relationship. Pearson's correlation coefficient (which is the most common type) only works on a linear relationship, there are matrix methods, polynomial rank methods, etc. for dealing with non linear relationships, or just generally being more robust, but generally more computationally intensive.
If your correlation function is some goofy polynomial a casual treatment of mean differences (the most common pearson method) probably isn't going to produce anything sensible. But that doesn't mean the data isn't correlated or that it doesn't show dependence, only that you need more sophisticated tools.
so you could directly import the 100 volume budget document into a flash visualization? Whether flash visualization is value for the money is open for debate.
All it seems to be is some aggregation and visualization stuff, which, if you want to parse through a 100 volume book takes time to read and make sense of the 100 volumes, decide what should and shouldn't be visualized. And of course they have a whole section the API they wrote to handle the spending data, and then they've implemented or bought some search tools etc.
True, correlation is not *necessarily* causation. But you cannot show causation without correlation.
It is equally possible that Algebra II teaches the necessary math tools and problem solving skills to be successful, or that those likely to be successful will take algebra II. Well, actually, I would be inclined to guess the former. I don't know specifically what Algebra II teaches in the US, but in canada to do well at any of the sciences and a large chunk of math/econ knowing how to do algebra makes a huge difference.
I should probably expand on how Algebra II *could* cause people to be successful. By itself a single course seems unlikely, but there's no harm in throwing out a theory. Again though, I reiterate, having not been through the US system I am not in detail familiar with Algebra I vs II vs anything else taught. Algebra II from what I can understand, in 3 minutes of research (note that searches for 'algebra II' from canada tend to produce canadian oriented results which isn't all that helpful) teaches you how to factor polynomials, deal with complex numbers, and the basics of numerical methods. These topics introduce students to a number of important concepts, first, at least that I can see, is that relationships aren't always linear, and just because they aren't linear doesn't mean they can't be quantified. Secondly it gives tools to examine how non-linear relations can relate to each other (one set of polynomials as an inverse of another set), and lastly an introduction to numerical methods is pretty handy when you deal with anything involving numbers. If Algebra II can be taught in such a way as to impart an understanding of problems that reflects a non trivial analysis, and can teach students some useful math tools it can be a driver of success. And it's of course also possible that Algebra II happens to sit just on a particular cusp of usefulness at this moment in time, and 40 years ago it wouldn't have had the same correlation effect and 40 years from now it won't either. Part of what might make it valuable is the relative number of people who can do it.
I could see the framework for a project like this easily taking a decade to set up, and that wouldn't (and probably shouldn't IMO) include a frontend. For example, we don't need a brand new organization to massage data, we need to pass laws requiring the originators to conform to a set format.
Therein probably lies much of the challenge here. 1 or 2 years into a multi year project, where most of the time has been spent figuring out what standards to use, how long it takes to convert to those standards and so on it looks like you haven't really done anything. You can't point to big spending abuses or great policy successes because you've spent all your time trying to quantify all the things you need to asses that. Of course there might be someone else trying to do the same thing, and they equally have no results and one of you is redundant too.
I don't understand the terminology they're using, since I'm not an american some of it is unique to the US government process. Even the idea a congressman vs a member of parliament imply different things and while I can superficially understand similarities, every big organization has it's own nomenclature.
It's one thing to worry about the process used to determine who gets money, rather than the mechanics of actually allocating money. Since everything in the US will have a uniform set of mechanics (whether it's some sort of bill through the congress or a Presidential signing, or a committee or a bureaucrat), which I am neither familiar with, nor have any reason to develop an interest in, I naturally reach the limit of my ability to understand what exactly they're spending the 4 million dollars on, or how difficult that job is.
4 million dollars would pay 30-40 people. That's not a whole lot, considering all of the data that has to be collected, checked, massaged into the right format, made compliant with accessibility rules, press dealings, server support, IT support for staff, and so on. I'm not an american so I'm not all that familiar with how funding is allocated in detail, but the site seems to spend a lot of time on awards, and sub award reporting. Presumably 'awards' could be easily extracted from regular budget documents but sub awards can't? There's seems to be a lot of time devoted to analysis of the data as well (which could drive costs up a lot if you have a few PhD's in stats or econ doing the analysis), in addition to building the flash visualization stuff.
On top of all of the sort of obvious stuff I'm sure there's a lot of legal there too. You can't always just go and blab what contractors you're giving money to, or if you can you need to verify the information you're going to say about the company. There can be a big difference between a deal with a company that is myurl.net and myurl.com, and you don't want to say they got 10 million dollars when they got 1, or 100.
As with all any large outfit, the more money you spend accounting for the money you're spending, the less is available for the actually things you're trying to do. It becomes a balance between the legitimate need to know where money is going, and the equally legitimate need to not waste 50 cents on every dollar documenting where you spent the other 50 cents. It seems like most everything on this website is available elsewhere, not necessarily easily. Whether or not a few millions of dollars in data aggregation on top of billions in accounting for trillions in spending is providing good value, especially when it's not my money, is beyond me.
Theories as to what caused the crash are not the same as having the blackbox data and being able to confirm any given theory, or decide that you cannot confirm any theory given the state of the blackbox.
Don't get me wrong. We know enough about aircraft, and the environment factors to make decent educated guesses. But if it flew into a storm it *should* have been able to handle and failed, that's very different than flying into a storm it shouldn't have been able to handle.
I was more responding to the anonymous posters implication that he's saying this now that he has a book to sell. He clearly doesn't need to sell the book, nor will he notice any more or less revenues from it given the controversy.
I wonder why he would need to sell a book. I mean.. he's got 13 billion dollars or something. JK rowling would be lucky to have cleared a billion, and that's for 7 of the best selling books + movies etc. Paul allen's Yachts combine to cost hundreds of millions of dollars, and he'd be lucky to make millions from his book.
What's amazing to me is that google, being not quite 13 years old, is being slapped with requirements that will extend for 20 years. Who knows, by then they could be a completely different company.
Research papers, talks, and grant proposals aren't ego. They're what you get paid for. As a tenure track (around here) you have to average about 1 paper a year as your own, or a talk (depends on your field), or both, + supervise grad students who also publish papers. And you pay for all of that with grants which you get from having successful grant proposals. Once you have tenure the 'papers per year' metric drops a bit but the basic 'publish or perish' mantra applies.
Research and writing are work, they take time to do well. If I'm not going to get credit for it I have to do it 'on my own time'. I don't know a lot of people that work for 8 hours a day and then go home and try and do the same thing for another 6 hours for the fun of it. Some profs eat sleep and breathe their work though, but even then, if you have things like families an
With OSS you can contribute, and then write about your contributions or you can 'give it away' (say host on some website) for free. And the author gets credit for both the software and papers written about it. With wikipedia your changes could be tossed if some random admin doesn't like them, or if someone else comes along and decides to change it. Your name never shows up, and you don't get credit for it in any way that would go on a grant proposal or that you can say at a promotion and tenure meeting as meaningful work you've done.
I'm sure if there was a good way to give academic credit for contributions to things like wikipedia it would be a great place for people to start publishing work.
I'm in canada. We're happy to take all those jobs. Because we give back 40% of the salary paid to game developers here in ontario. If you have payroll over 1 million dollars you can get that money faster, if not, you have to wait until the end of the project.
The UK is stuck between a rock and hard place on this one. A desperate need for revenue from corporate taxes, balanced by a desperate need for income taxes, and being no where near a competitive labour market for either of those.
Ideally, in an age of austerity, no one would be offering 40% tax breaks to any industry. But here in canada, and ontario in particular, we're trying to build a games industry. So we're offering huge tax incentives to get people to either move here, or startup here. Either the UK can try and claim unfair trade and subsidy (probably legitimately), and demand places with low taxes on game developers stop doing that, which would take years and might fail. Or you can match them, and keep what you can of income taxes and so on.
A 5 or 10% difference probably isn't enough to persuade someone to move. But 40%, which is more or less what they'd get in Ontario or Quebec is pretty hard to refuse.
Imagine for a minute you could build a generically unlicenced device, which shouldn't interfere with licenced ones. So it works on unlicenced space (or on licenced space but it can't find a free channel), now you have two choices, either the device can fail to operate, since there are too many devices. Or it can automatically go hunting for new channel space.
Ideally a device should be able to hunt around for free wireless spectrum, and then resolve if it can stay there when something else shows up. I can see a lot of problems with this system, but in a low powered device the harm it could do is hopefully minimal.
Overall it's a good idea, and a useful research project. But it requires the government to manage its use, and well, depending on the party in power (and this applies everywhere, not just he US) the idea of more government controlling what your devices can and cannot do might offend people.
I would love to see the flak from trying to get faculty members to use an RSA token. 'What if I lose it', 'isn't this dumping responsibility on me because IT can't build a secure system.' I argue this regularly with both my bank and department chair. My world of warcraft account (with authenticator) is at least in principle more secure than they are. And honestly, there's nothing on my WoW account that cannot be replaced by a rollback.
I'd probably just require an RSA token (or equivalent) for everything, and integrate it into the employee ID card. I'd have a hard time arguing any systems should be less secure than any other. HR: tax information, and where your pay goes, while, around here at least, your salary is public people probably don't want everything in HR public. E-mail: medical information on students with disabilities asking for accommodations etc. Room bookings, a student could reschedule an exam/class on top of something else to try and get out of things. But I dunno, it sounds like the system at penn state is a better compromise than what we have.
Slashdot Mirror
I was asked why, in my opinion, homebrew on hacked PS3's was worthless, I answered. I stand by it. It's not worth the effort. Exactly as you said, XNA is a better solution. The cell is expensive, complicated, and because the whole architecture is atypical you're looking at writing all of the tools without complete information from scratch. Sony has literally dozens of people who spent years writing it with full information. By the time you get anything useful as a development framework that will do much of anything the PS4 will be half way through its lifecycle.
And even then you're developing into a technology that's dead. Not because it's bad, but because it sits between the CPU and GPU in terms of what it does conceptually, and problems tend to lend themselves much better to either the CPU or GPU than to the cell. As a result even sony is jumping ship to ARM for their PSP2.
They've ported everything off the cluster I built, I have since gone to another institution. We build the cluster about 3.5 years ago, and ported everything to GPU computing 3 years ago, because a single 400 dollar GPU (nvidia specifically) outperformed the 4 machine cluster by about a factor of 2. it was a waste of our money and sony's. Now I do some PS3 development stuff (through official channels) and that's bad enough, trying to do it without all of the libraries they provide is an enormous undertaking from scratch, to accomplish a task better done a million other ways.
Don't get me wrong. I distinguish between Sony should still be supporting Linux, and Geohot's taking that into his own hands and then some. Sony sold it with PS3 linux, they shouln't take it out (at least on the machines it was sold with, new versions without it are fine). But they shouldn't let someone hack the system either, that violates the entire integrity and the point of the system in the first place.
Sure, a small handful of people bought those machines to hack them (again, separate from running linux officially), but that shouldn't give them free reign to damage the experience for everyone else.
Because with homebrew you'd have terrible access to the hardware, and the hardware isn't all that compelling (it's not any better than a PC, which is equally tied to your living room). There's no substantial media functionality that's missing, and anything that is missing is prohibitively expensive to add when it is, again, better done on something other than a PS3.
The Cell isn't all that great. Seriously, it isn't. You need a lot of money, time and people to make proper use of it. And then there's the GPU which you don't have great access to. If you want to homebrew up a game it's far easier to just use a PC. And if you want to do cell development, well, why? The only people actually doing anything with the cell are supercomputing clusters, and even they're abandoning cell left right and centre for GPU computing (which you can't do on the PS3's gpu).
Homebrew on a handheld is a whole other ball game. You can actually take that with you which you can't do with a PC. Even if you don't get access to decent tools or all the hardware being able to 'take it with you' is pretty compelling. Being able to play on my TV is a matter of buying a $20 cable for my PC.
In their defence, the whole purpose for the vast majority of users when buying a PS3 was to have a gaming machine and some semblance of fair competition in multiplayer. The forums are naturally more full of fanboys than anywhere else. Any PS3 hack directly interferes with the notion of a fair playing session against other people. As much as it got Sony free press, PS2 linux... I'm sorry PS3 linux was only ever there to try and skirt around EU import tariffs. It didn't work. Sony really doesn't want you using PS3's for astrophysics clusters or airforce research (I helped build a cluster for astrophysics work). They sell at a loss because they want you to buy games. I'm sure the airforce cluster is great press, but the vast majority of the research just costs them money they don't get a tax break for.
All MMO's go after botters. Shouldn't I be able to run whatever software I want on my own machine?
The olympics go after people who have too much cold medication or whatever else. Shouldn't you be allowed to take whatever your doctor recommends for your health?
The US congress thought steroid use in baseball was so important they dragged barry bonds and co. to washington to talk to them. This is when they are trying to deal with a trillion dollar deficit, they're willing to waste days of peoples time on steroids in baseball. And you think sony fanboys are overreacting to a hack? At least the sony fanboys are actually participating in, and affected by cheaters directly.
Not that Sony is blameless. They should never have allowed 'other OS linux' on the PS3 in the first place, they should not have gone around threatening to sue everyone under the sun who might have been intrigued by geohot. But if you're a MMO player, you're glad to see the banhammer go out to gold farmers, botters etc. Sony is trying to balance on one hand the developers and hardcore gamers (who roughly have aligned interests in terms of security), and hacker types who should be free to toy with their own stuff, but not at the expense of the network experience of everyone else.
Lets be honest. Sure, a PS3 jailbreak hack lets you run homebrew games, and may re-enable partially functional PS2 emulation, and brings back linux support. But it also lets you mess with the memory state of your machine and hack the game as you're running and it lets you pirated games. Homebrew has no real value on a PS3. PS2 emulation, I'm not sure on, I'm betting sony didn't just take it out for the fun of it, it's probably really hard, if not impossible to do properly and provide a good experience. Linux... well it shipped with linux so I guess they should still support it. But that goes into the next problem, which is the whole network experience of 40 odd million players (obviously not all of whom actually use the network functionality) can be easily disrupted by even a handful of people hacking which is really a serious problem. Avoiding that is sort of the point of having a console in the first place, and being able to steal stuff from the playstation store isn't exactly something I support either. I don't think pirated games (blu-ray/DVD) is a huge issue, though I could be wrong, I suppose once the hack is out here some cheap PS3 game knockoff manufacturers could pose problems if they want to get into that business.
of course we also train a lot more people too (both in absolute numbers and as a percentage of the population). In practice the details of law, and the depth of knowledge that define a law degree are no less than that required to define the philosophy of morality. The few that truly grasp the latter in addition to the former can become professors. Everyone else just gets a law degree.
The more complex the world the more time needs to be spent to understand it at any level. I don't think schools lack depth, in fact I would argue they have tremendous depth (and choice), but it is implicitly assumed that you both study and learn something about morality when you take a degree in law. Animation is a college not university thing, so I won't hit on that (that really is job training, there are Uni level comp sci courses on animation but that is an exercise in kinematics, physics and rendering, a somewhat separate problem again). Game design, which is where I live, is a whole other ball game. There's the 'arts' side of design, which incorporates psychology and writing, essentially 'how do I make it fun'. There's no painting involved, and it's much more about storytelling and pacing but with a game development twist (along with all of the things that make a game a game, death, violence, or lack thereof, complexity and so on). And then there is the 'science' side of design (which is what I do), which really is an exercise in computer science. How do you define and test balance? How do you quantify and assess all of those things the 'arts' designers need as tools? How complex is this problem? The broad philosophical question "what is fun" is certainly covered, a lot. But in practice you need to know 'how do I define fun' 'how do I know if it is fun' rather than just 'what is fun'. The devil is in the details as it were. I can write you a 40 page paper on what is fun, but it's worthless unless I can translate that into numbers in a program that runs and actually *is* fun. The standard has been raised from a general understanding to a specific one.
Virtually all of my students * ARE* well rounded already. Admittedly I see physics, math, chemistry, engineering and CS types and the occasional wayward business student. They have a grasp of, and an internal model of how the world behaves, where they fit in it, and how to learn and understand problems. They wouldn't be in university if they didn't. What they need is depth of understanding of a specific area. Everyone of them has some high minded idea of what morality is (and, if so inclined can easily find the classic texts on whatever your topic of choice is), but if you cannot apply those principles to solve real problems they have no value. Not because the principles themselves are valueless, they aren't, but because they all get the principles already, what they need is to understand the details of real problems.
It is the difference between the people who complained a lot about democracy in libya and got nice cushy faculty positions in the UK and the US from it, and the people who are now trying to actually build a system of government. Some of them are even the same people, but the latter are just as committed to the principles of democracy as the former, they just have to spend time building electoral districts, a parliamentary system and voting rules.
Don't kid yourself for one moment, the people learning the enormous depth of information required to actually solve problems understand and appreciate the world around them. That's why they are invested in learning to actually solve real problems at all. 140 years ago you needed to worry about people who 'got the big picture' much more than today. Teachers are better, information is more accessible (and cheaper). It's much easier to a make a student reasonably well rounded now by the time they're done high school. But of course nearly everyone graduates high school. In canada it's about 90% of people by the time they are 24 have completed high school, compared to the US in 19
the question becomes: is it the same greek as being asked here. In 1869 Greece was 20 years independent from ottoman rule and still basically a nothing state on the world stage (not that it's much of anything today). Greek as taught in modern greece (or in this case 1860's greece) is not necessarily the same as the various versions of greek that would pre-date the modern world. There are classical language courses you can take at some schools, including universities, but the version of greek harvard is asking about in 1869 probably has relatively little bearing on contemporary greek of that period let alone modern greek.
Today you can find greek in a smattering of countries, italy, turkey and greece being the big ones, but armenia, ukraine, cyprus and a few others as well, but no more than a person today could do well with old english (which is nearly unintelligible), or middle english (shakespeare and KJV bible era, which is somewhat comprehensible).
I think probably if you could pass that exam today you could probably still do well in most liberal arts programmes at least. It shows an ability to grasp foreign languages (always handy), and a relatively diverse reading set. Not that there wouldn't be better measures of success today though. I think a modern scientist faced with an exam from 140 years ago might have a lot of trouble. There's language, terminology and style advances, skills that have largely been obsoleted (by for example the calculator), and then well, we know more now than they did then of course. Even if the math is the same, the way the math is written has changed quite a lot (hello matrices!). Any test carries with it the context of its time, and there's probably a ripple effect. What is today a challenging entrance exam at harvard or a PhD level topic in programming will 20 or 30 years from now be pushed into the highschool curriculum, and then 20 or 30 years later it will likely be long forgotten as it is supplanted by new problems and techniques.
Ya, I mean, servers and phones aren't exactly areas where microsoft was the market leader, or if it was, never by any meaningful amount (servers). Saying linux is dominating the phone market, when their phone product is 2 years old is like oldsmobile claiming they won the car market in 1901... Whether or not MS-Nokia will be rolling in mobile money 2 years from now remains to be seen.
The server arguement is interesting. HPC was never MS's area. At all. They had probably 1% of the market 20 years ago, and probably have 1.5% of the market today (CUDA on windows being that 0.5% growth, the other 1% is people who don't know any better, as it was 20 years ago). Sure linux did a great job cannibalizing unix marketshare, but that's not beating windows. In terms of general servers.... servers for what? Giant distributed databases (say for websites), when was that ever any area MS dominated? They've done well in that space, and I've got a few windows servers here and there but not for anything important. I don't think they've ever been the company to go to for 5 9's of uptime. Nor have they really tried to be (although you can probably achieve that on windows boxes), supposedly some companies offer 6 9's of windows server 2008 uptime. But that's as much hardware as it is software.
Embedded devices... again. When was MS anything in the embedded devices space? Sure there's roughly parity on the 360 (windows) PS3 (linux) in terms of marketshare. Only hackers care what OS those machines use, everyone else just wants to play games, and don't really care about OS, and it's not like the xbox has a big 'Windows inside' sticker. Sure MS ran 'terminals' (for bank and ticket machines), but never really consumer electronics. I don't think they've ever wanted that market, and winning it 'from' microsoft, who aren't even competing there doesn't make a lot of sense.
The only area 'open' has destroyed MS is browsers. Which is because MS made a truly horrid browser and has been desperately trying to rectify that error and kill it off for years. Firefox is a vastly better product for just about everyone.
20 years ago MS dominated 2 big markets. Desktop operating systems and desktop office applications. And not even really office back in 91 (that is windows 3.0 days). Since then office has faced real competition from free (google docs) and the end of printing for everything, recently making pdfs being free and easy, and so on. I think Open Office and it's compatriots have kept MS prices down, but today paying 150 bucks for 3 home licences of office doesn't seem unreasonable, after all, the people who develop the software should maybe get paid for it. The home and business operating system market... is still owned by microsoft. Through and through. It may face competition from apple, but linux isn't even on the radar here. Unfortunately.
The other thing 'free' has done is tried to accustom us to not paying for the products we use. But being 'worth something' to advertisers who will pay the people who develop the products we use. It's sort of like regular TV. And not in a good way. Of course people prefer free with no strings attached to free with advertising. But that's beside the point. Whether you use linux or windows or mac the thousands of software developers who wrote that software are mostly getting paid, with MS you pay them for the product. With Mac you pay for the experience and with linux you get it for free because hopefully someone business will pay on your behalf, either directly or not.
Linux has done fantastically well for itself since 1991. But so has MS. And linux hasn't made any significant inroads on MS's main product portfolio. If anything MS has pushed into areas that are traditionally more unix/linuxy in servers than linux has pushed into windows areas on the desktop. And the mobile space seems a long way from being decided. It has however served as both a great relatively free source of anti-MS information (much of it deserved) and a good way to keep the monopoly from being as disastrous as it could be.
Every woman I have ever met is far more likely to forget to take pills than I am to forget to have condoms. Choosing to take the risk without one is not the same as forgetting to put one on.
In my (very limited) anecdotal experience women miss an average of one pill every 3 months. But two in a row I don't know the probability there.
surgery.
In terms of preventing diseases, condoms would seem to be the only main option. I would doubt they are much better than hormones at contraception but they do deal with the disease problem.
Caveat: I don't know the error rate (or tolerance) for oral contraceptives, and I'm not even sure I've ever actually heard it. If the average woman misses say 1 pill a year, or 1 pill every six months or whatever the number is, what does that do to pregnancy risk compared to the supposed effects of taken optimally. Then you get into exposure to conflicting meds etc. which may have a similar problem. Quoting the expected error rate assuming you don't take conflicting meds, don't miss pills etc. isn't real world data on error rates. Condom error rates are probably pretty close to accurate (fall off & break are the main options I would think) neither of which count if you don't bother to wear one in the first place. You can't really 'forget' to wear a condom, well, maybe you can (sleep sex?), but I'm not really sure how that would be factored in, but you can forget pills. I suppose you could apply substances to a condom which degrade its quality (e.g. if you're trying to clean them for reuse, which I could see as an issue in the developing world), but afaik they aren't really designed to be reused.
Correlation only measures degree of relationship. Pearson's correlation coefficient (which is the most common type) only works on a linear relationship, there are matrix methods, polynomial rank methods, etc. for dealing with non linear relationships, or just generally being more robust, but generally more computationally intensive.
If your correlation function is some goofy polynomial a casual treatment of mean differences (the most common pearson method) probably isn't going to produce anything sensible. But that doesn't mean the data isn't correlated or that it doesn't show dependence, only that you need more sophisticated tools.
so you could directly import the 100 volume budget document into a flash visualization? Whether flash visualization is value for the money is open for debate.
All it seems to be is some aggregation and visualization stuff, which, if you want to parse through a 100 volume book takes time to read and make sense of the 100 volumes, decide what should and shouldn't be visualized. And of course they have a whole section the API they wrote to handle the spending data, and then they've implemented or bought some search tools etc.
True, correlation is not *necessarily* causation. But you cannot show causation without correlation.
It is equally possible that Algebra II teaches the necessary math tools and problem solving skills to be successful, or that those likely to be successful will take algebra II. Well, actually, I would be inclined to guess the former. I don't know specifically what Algebra II teaches in the US, but in canada to do well at any of the sciences and a large chunk of math/econ knowing how to do algebra makes a huge difference.
I should probably expand on how Algebra II *could* cause people to be successful. By itself a single course seems unlikely, but there's no harm in throwing out a theory. Again though, I reiterate, having not been through the US system I am not in detail familiar with Algebra I vs II vs anything else taught. Algebra II from what I can understand, in 3 minutes of research (note that searches for 'algebra II' from canada tend to produce canadian oriented results which isn't all that helpful) teaches you how to factor polynomials, deal with complex numbers, and the basics of numerical methods. These topics introduce students to a number of important concepts, first, at least that I can see, is that relationships aren't always linear, and just because they aren't linear doesn't mean they can't be quantified. Secondly it gives tools to examine how non-linear relations can relate to each other (one set of polynomials as an inverse of another set), and lastly an introduction to numerical methods is pretty handy when you deal with anything involving numbers. If Algebra II can be taught in such a way as to impart an understanding of problems that reflects a non trivial analysis, and can teach students some useful math tools it can be a driver of success. And it's of course also possible that Algebra II happens to sit just on a particular cusp of usefulness at this moment in time, and 40 years ago it wouldn't have had the same correlation effect and 40 years from now it won't either. Part of what might make it valuable is the relative number of people who can do it.
I could see the framework for a project like this easily taking a decade to set up, and that wouldn't (and probably shouldn't IMO) include a frontend. For example, we don't need a brand new organization to massage data, we need to pass laws requiring the originators to conform to a set format.
Therein probably lies much of the challenge here. 1 or 2 years into a multi year project, where most of the time has been spent figuring out what standards to use, how long it takes to convert to those standards and so on it looks like you haven't really done anything. You can't point to big spending abuses or great policy successes because you've spent all your time trying to quantify all the things you need to asses that. Of course there might be someone else trying to do the same thing, and they equally have no results and one of you is redundant too.
I don't understand the terminology they're using, since I'm not an american some of it is unique to the US government process. Even the idea a congressman vs a member of parliament imply different things and while I can superficially understand similarities, every big organization has it's own nomenclature.
It's one thing to worry about the process used to determine who gets money, rather than the mechanics of actually allocating money. Since everything in the US will have a uniform set of mechanics (whether it's some sort of bill through the congress or a Presidential signing, or a committee or a bureaucrat), which I am neither familiar with, nor have any reason to develop an interest in, I naturally reach the limit of my ability to understand what exactly they're spending the 4 million dollars on, or how difficult that job is.
salary + benefits + building + equipment. Easily 100k/employee.
4 million dollars would pay 30-40 people. That's not a whole lot, considering all of the data that has to be collected, checked, massaged into the right format, made compliant with accessibility rules, press dealings, server support, IT support for staff, and so on. I'm not an american so I'm not all that familiar with how funding is allocated in detail, but the site seems to spend a lot of time on awards, and sub award reporting. Presumably 'awards' could be easily extracted from regular budget documents but sub awards can't? There's seems to be a lot of time devoted to analysis of the data as well (which could drive costs up a lot if you have a few PhD's in stats or econ doing the analysis), in addition to building the flash visualization stuff.
On top of all of the sort of obvious stuff I'm sure there's a lot of legal there too. You can't always just go and blab what contractors you're giving money to, or if you can you need to verify the information you're going to say about the company. There can be a big difference between a deal with a company that is myurl.net and myurl.com, and you don't want to say they got 10 million dollars when they got 1, or 100.
As with all any large outfit, the more money you spend accounting for the money you're spending, the less is available for the actually things you're trying to do. It becomes a balance between the legitimate need to know where money is going, and the equally legitimate need to not waste 50 cents on every dollar documenting where you spent the other 50 cents. It seems like most everything on this website is available elsewhere, not necessarily easily. Whether or not a few millions of dollars in data aggregation on top of billions in accounting for trillions in spending is providing good value, especially when it's not my money, is beyond me.
Theories as to what caused the crash are not the same as having the blackbox data and being able to confirm any given theory, or decide that you cannot confirm any theory given the state of the blackbox.
Don't get me wrong. We know enough about aircraft, and the environment factors to make decent educated guesses. But if it flew into a storm it *should* have been able to handle and failed, that's very different than flying into a storm it shouldn't have been able to handle.
I was more responding to the anonymous posters implication that he's saying this now that he has a book to sell. He clearly doesn't need to sell the book, nor will he notice any more or less revenues from it given the controversy.
I wonder why he would need to sell a book. I mean.. he's got 13 billion dollars or something. JK rowling would be lucky to have cleared a billion, and that's for 7 of the best selling books + movies etc. Paul allen's Yachts combine to cost hundreds of millions of dollars, and he'd be lucky to make millions from his book.
What's amazing to me is that google, being not quite 13 years old, is being slapped with requirements that will extend for 20 years. Who knows, by then they could be a completely different company.
Research papers, talks, and grant proposals aren't ego. They're what you get paid for. As a tenure track (around here) you have to average about 1 paper a year as your own, or a talk (depends on your field), or both, + supervise grad students who also publish papers. And you pay for all of that with grants which you get from having successful grant proposals. Once you have tenure the 'papers per year' metric drops a bit but the basic 'publish or perish' mantra applies.
Research and writing are work, they take time to do well. If I'm not going to get credit for it I have to do it 'on my own time'. I don't know a lot of people that work for 8 hours a day and then go home and try and do the same thing for another 6 hours for the fun of it. Some profs eat sleep and breathe their work though, but even then, if you have things like families an
With OSS you can contribute, and then write about your contributions or you can 'give it away' (say host on some website) for free. And the author gets credit for both the software and papers written about it. With wikipedia your changes could be tossed if some random admin doesn't like them, or if someone else comes along and decides to change it. Your name never shows up, and you don't get credit for it in any way that would go on a grant proposal or that you can say at a promotion and tenure meeting as meaningful work you've done.
I'm sure if there was a good way to give academic credit for contributions to things like wikipedia it would be a great place for people to start publishing work.
I'm in canada. We're happy to take all those jobs. Because we give back 40% of the salary paid to game developers here in ontario. If you have payroll over 1 million dollars you can get that money faster, if not, you have to wait until the end of the project.
The UK is stuck between a rock and hard place on this one. A desperate need for revenue from corporate taxes, balanced by a desperate need for income taxes, and being no where near a competitive labour market for either of those.
Ideally, in an age of austerity, no one would be offering 40% tax breaks to any industry. But here in canada, and ontario in particular, we're trying to build a games industry. So we're offering huge tax incentives to get people to either move here, or startup here. Either the UK can try and claim unfair trade and subsidy (probably legitimately), and demand places with low taxes on game developers stop doing that, which would take years and might fail. Or you can match them, and keep what you can of income taxes and so on.
A 5 or 10% difference probably isn't enough to persuade someone to move. But 40%, which is more or less what they'd get in Ontario or Quebec is pretty hard to refuse.
Or the total war series (creative assembly). Or a whole lot of the Lego games ( http://en.wikipedia.org/wiki/Traveller's_Tales)
Imagine for a minute you could build a generically unlicenced device, which shouldn't interfere with licenced ones. So it works on unlicenced space (or on licenced space but it can't find a free channel), now you have two choices, either the device can fail to operate, since there are too many devices. Or it can automatically go hunting for new channel space.
Ideally a device should be able to hunt around for free wireless spectrum, and then resolve if it can stay there when something else shows up. I can see a lot of problems with this system, but in a low powered device the harm it could do is hopefully minimal.
Overall it's a good idea, and a useful research project. But it requires the government to manage its use, and well, depending on the party in power (and this applies everywhere, not just he US) the idea of more government controlling what your devices can and cannot do might offend people.
I would love to see the flak from trying to get faculty members to use an RSA token. 'What if I lose it', 'isn't this dumping responsibility on me because IT can't build a secure system.' I argue this regularly with both my bank and department chair. My world of warcraft account (with authenticator) is at least in principle more secure than they are. And honestly, there's nothing on my WoW account that cannot be replaced by a rollback.
I'd probably just require an RSA token (or equivalent) for everything, and integrate it into the employee ID card. I'd have a hard time arguing any systems should be less secure than any other. HR: tax information, and where your pay goes, while, around here at least, your salary is public people probably don't want everything in HR public. E-mail: medical information on students with disabilities asking for accommodations etc. Room bookings, a student could reschedule an exam/class on top of something else to try and get out of things. But I dunno, it sounds like the system at penn state is a better compromise than what we have.