Uh, ya. Now explain to me why thousands of students from those countries (usually the best ones at that) continue to pursue graduate course work in North America and Europe.
A good starting place for information is this little document by Mike Gleicher who is a graphics prof at the University of Wisconsin. He gives a lot of general advice, but also some specific stuff for graphics which is helpful and otherwise hard to find.
Here's his list of places where major graphics research is going on in North America:
The "Big" Places for Graphics: (all of these places have LOTS of
graphics students)
Washington
Stanford
Georgia Tech
UNC
Utah
MIT
Brown
British Columbia
Other Really good groups (smaller, more personable):
Princeton
Caltech
NYU
CMU (was big, but lots of people left)
Toronto
Up and Coming Groups (newer groups with a small number of newer
faculty in graphics)
Berkeley
Virginia
Illinois
USC
He also of course adds Wisconsin to the list too. I'd personally place it in the last category, since they have a small, young but respectable program. Generally, I agree with this list although there are probably a whole bunch of other places with just one faculty member doing good work.
Having spent time studying and researching in Europe, there are some good graphics groups there too. In the UK, I know of two places off hand: Cambridge (Malcolm Sabin doing stuff with surfaces and geometric modeling) and Bath (Phil Willis and some others). In Germany, there is the Max Planck Institute (which does everything in English and is a very strong group although I admit I used to work there, so I'm probably biased), TU Darmstadt (Alexa), RWTH Aachen (Kobelt) and Tuebingen (Strasser). In Switzerland, there is the ETH in Zuerich (Markus Gross), the EPFL in (I think) Lausanne (Nadia and Daniel Thalmann) and Basel (Thomas Vetter who not so long ago left Freiburg to start a new graphics group there). In France, there is a group at the INRIA in Grenoble (Marie-Paul Cani). The above is certainly not an exhaustive list, just names that come to mind off the top of my head.
Most of these programs in continental Europe probably require you to already have your MSc before beginning PhD studies, but some have MSc programs as well. I know the MPI has one and awards scholarships even to foreign students. Generally, it's easier to get funding with only a Bachelor's degree in the US than in most of Europe. I've also heard the funding situation in the UK is not really very good, which is probably another reason to consider places like the US, Canada, Germany or Switzerland.
Competition for admission to US schools is fairly intense and is a time consuming and expensive process, but there are some really amazing opportunities here. I'm currently a grad student (graphics, visualization and scientific computing) at UNC and we probably have the largest number of graphics and imaging faculty and grad students under one (academic) roof in the world. We're not as hard to get into as a lot of the top schools like Berkeley, MIT, Stanford or CMU, but we still only take about 1 in 7 applicants or something. The key is to apply to as many places as you can afford, but only apply to places you'd actually attend.
That website seems to imply that you need an additional order n amount of space to do heapsort. However, it can be done in place without too much effort:
1. Take your input array containing n elements and perform a build maxheap operation on it. This takes O(n) time and can be done in place.
2. Swap the root (the maximum element in the heap) with the last value in the part of the array still occupied by the heap, effectively removing the root from the heap and placing it in the sorted portion at the end of the array. O(1)
3. Percolate the new root (that was just swapped) down the tree to guarantee the heap property holds for it. O(log n)
4. Go back to step 2 so long as there are still elements left in the heap.
Still, quicksort usually tends to be faster in most cases because of overhead. Heapsort really pays off though when you don't want to sort the entire sequence. For example, maybe you just want the 10 largest or smallest elements. In that case, you only have to do the full O(n) build heap operation instead of a full O(n log n) sort.
UNC is most definitely NOT a liberal arts university. Although there is no real engineering program, there is a lot of natural science and medical research going on. Plus, there's the business, law, journalism and a myriad of other more professional schools.
The CS department here is a top 20 research-based program with a very strong focus in the computer graphics and imaging areas.
Just like they do for virtually every other plant. How many people grow more fruit, vegetables, grain or herbs than they buy? How about people who brew more of their own beer than buy it in the supermarket? How about people who grow their own tobacco?
Usually, it just turns out to be easier and more convinient to buy most things retail. Of course hobbyists will always do it for the fun or the challenge, but there would be plenty of money to made by companies especially compared to the basically $0 they make right now.
Honestly, I have to wonder how much of a problem this really is. Over the course of a few weeks (the usual time frame for reviews to go out and come back), I imagine most websites containing this information get a fair amount of traffic. A paper I wrote last year gets downloaded about 25-30 times per month and it's not even that signficant.
I know the site for my group gets thousands of accesses every month. Is it really that easy to pick out who the reviewers are from all that traffic?
Even with an anonymous proxy, it's not terribly hard to guess who the possible reviewers are anyways. There are only so many people/groups in a field and papers chairs try to send your paper to somebody who will be able to competently evaluate it.
I'm not convinced this is true. Mostly, there are so many PhDs (and really in the big picture of society, how many are there?), because with a PhD you get to be in charge of stuff. That doesn't mean you get to be the next Einstein. It just means people with the money trust you more to lead a research project or study. As long as there are PhDs, people with an MS or BS will continue to work underneath them.
The predator is already there. How many people go to college? How many graduate? How many become graduate students? How many finish graduate school with an MS/MA? With a PhD? How many get tenure track jobs?
The numbers have already been dramatically winnowed down before somebody gets to the level of Assistant Professor, although perhaps the cut at each of the above levels needs to be tighter to get the numbers down. The problem of overproduction seems to vary a lot by field though too. There is a sizeable glut of English PhDs who have trouble finding traditional jobs that require such a degree. A number of programs are aware of the problem and intentionally admit fewer PhD students than they can support because overproduction has been such a problem.
At the same time, in my research group in a Computer Science deparment (good, but not top 5 or anything), graduates (MS and PhD) have not really had trouble getting decent jobs where a graduate degree is an asset even in today's less than stellar job market.
They really need to start doing that here at UNC-CH. There was an article a few months back about the millions of dollars in delinquent parking fines. As a student, if you don't pay your parking tickets, they'll block your registration and I believe also the registration of your car with the state. Apparently though, there are a lot unpaid fines for cars that are registered to a student's parents (and thus cannot have registration blocked) out of state (thus NC can't block car registration).
You start towing things and it's amazing how fast those fines start getting paid though!
I'm sure Intel will spend the effort to bring about a port of their C/C++ and probably Fortran compliers to x86 OS X. With this new partnership, it's obviously in their own interest for software on these Intel Macs to run as fast as possible.
It doesn't shock me too much that it only took 2 hours to port Mathematica. I mean, the API for OS X on Intel is probably exactly the same as for OS X on PPC. Probably only very, very small parts (if any at all) of Mathematica are written in assembly code. You fix those parts and anything that relies on specific processor behavior then do a recompile.
Apple has obviously got an x86 gcc for Tiger and has already begun the process of porting the frameworks, most of which will probably not require massive porting effort. Frameworks like vecLib will probably require some more work to use SSE instead of Altivec though.
Even the concerns about things like endianness are not really a problem so long as the code was written the right way in the first place.
Such a switch to Intel chips would do very little towards true Win32 binary compatability. Obviously, you can't run an x86 Linux binary on Windows, just like you can't run a Windows binary on x86 Linux (at least without WINE or other similar compatability software). The same will hold true for an x86 version of OS X.
Right now, products such as Virtual PC allow Mac users to run Windows programs under emulation. A switch to Intel chips would mean such programs could run Windows programs much more quickly, but that's still a long, long way from OS X being able to handle the complex Windows APIs natively.
Well, actually they don't pay the RIAA. They pay ASCAP, BMI and SESAC, which technically represent the song writers and performers as opposed to the big record companies. If I remember correctly, non-profit stations do not have to pay ASCAP or SESAC any royalties at all. BMI is paid, but at a dramatically reduced rate (only paid for about one week a year at a place I used to work). Right now, in the US the cost is somewhere around 8 cents per song, which for a commercial station is nothing compared to even labor costs.
Germany does have a tiered high school system, but it's not as if the education system only serves the elite. Everybody is entitled to an education there. If anything, the average job in Germany requires substantially more schooling that the average American job. For example, if you want to be a flower dealer or a hair stylist, there's a formal (and usually long) practical education path to be followed.
I always thought people buying computers at supermarkets (like Aldi) was one of the weirdest things computerwise about Germany. One time, I saw an acquaintance of mine return a computer at Plus even. The cashier gave him 2 500EUR notes taken from the normal register for the refund even.
I honestly can't imagine anybody in the US buying such an expensive item cash (especially at a supermarket) or the supermarket even having that amount (especially in that denomination) in a normal register (maybe in a safe in the manager's office or something). Personally, I wouldn't even feel safe walking around with 1000EUR in my pocket. Germany's definitely much more of a cash culture than the US (where people can and do use credit cards for virtually everything) and this all happened in a very safe suburb, so it does make sense I guess.
Moreover, it doesn't even necessarily mean that much inside a given architecture, especially these days with entire system performance often being more dependent on cache, memory, disk and in certain cases GPU.
Oh, there definitely are advantages to having an MS...
People working on an MS at a major research university will most likely be involved in some sort of research and will be required to take some more advanced coursework. More than anything, I think the process solidifies the student's knowledge of the field and allows him or her to focus a bit on one or two specific areas of interest in order to gain some depth not usually possible in an undergraduate program. Additionally, students are expected to work independently to a much greater degree. The more detailed knowledge, a demonstrated ability (at least in comparison to undergrad) to work independently, and the fact that generally stronger students go to graduate school means MS grads are attractive to many companies. Also, for a lot of jobs that require research skills, a MS can often demonstrate a basic competency. There are a good number of jobs out there (more than for PhDs) that require only this level of experience.
The advisor situation can play a role as well. Many profs have contacts (often former students) at companies doing interesting work. When a student graduates, the advisor often turns to their network to start asking around about what's available for this student with a particular skill set. In particular, during an MS every student (at least at research-oriented programs) tends to specialize more than they did as an undergrad. In many areas, this means you have specialized knowledge. A lot of MS grads use this to get jobs where more interesting or cutting-edge work is being done.
Personally, I'm not entirely convinced MS grads have more job opportunities. However, they do in general tend to have *better* ones. Now, a lot of the above is probably not true for people who are already working and doing a coursework based MS at night. My impression is that a lot of those programs give the student a more solid background in CS, but mainly serve as a way to demand a higher wage or help in getting into management.
Perhaps the saddest part about France going to Microsoft for this project is that whatever data is produced is more likely to be locked into some proprietary format. That could be particularly unfortunate, since these cultural resources really belong to all the people of France and should therefore be made as accessible as possible.
I dunno. The older I get, the less interested I find myself in younger generations. They're all just in a different stage of their lives or something. Of course, if it's just a year like the article author, that's no big deal. However, once you're 21, it's just not as cool to be buying beer for and hanging out with only 18 year olds.
It always amazes me how many Germans will claim that everybody in their country (or Europe, although I haven't heard that so much) can speak English. It's simply not true. The ability of people to speak even basic English can get very spotty.
First, more educated people (like those who went or current go to a regular university) are more likely to be able to speak English, while a lot of blue collar workers don't speak it at all.
I've found that region also plays a big difference. I lived in the Saarland for two years and more or less only people associated with the university could speak English. Even at the foreigners office which handled visas, I think only one worker spoke English (and not that well) and these people deal with foreigners all day, every day. Apparently, only recently has the school curriculum changed, so that English (as opposed to French) is taught there as the first foreign language (or so I'm told). A guy I worked with (a German) was telling me about a job offer he had considered in Dresden (former East Germany, where learning English was very rare before the reunification), where two of his assistants (who he would have to keep because of civil service employment laws) could not speak English at all, even though they worked in a Computer Science department. On the other hand, it seems like many, many more people in the whole Rhine/Ruhr area (Cologne/Duesseldorf/Essen) speak good English.
Finally, there is the age gap. Younger people are WAY more likely to be able to speak English than the elderly, who probably were last in school before English enjoyed such broad status as the default international language.
On the plus side, there are a lot of people who can sort of understand basic English, even if they can't speak it themselves. Most Germans tend to be helpful and polite even if you can't speak a common language, but of course that's not true for everybody.
I'm convinced that if they had more subtitled television and movies, it would improve the situation over time. In a smallish city (definitely a city and not a town) like Saarbruecken, there was still (and probably still is) only one movie theater which showed English movies in the original version (not dubbed) and it was second-run place that showed only one per week. Television is just as bad, except of course, it's the same bad situation throughout the country. Pretty much the only English programs that were subtitled instead of dubbed when I lived there were Jackass, some dating show on MTV and Crank Yankers. All of them cater to a young audience, so it probably wasn't financially worth it to pay for the dubbing. I did get international CNN and BBC World though, which were both nice, but most people want more entertainment than just the news.
Please, don't take this the wrong way, but your post makes a lot of sweeping generalizations about how things are in the US. Yes, laws have changed and so has the bureaucracy that implements them. I'm not convinced that's so much a reaction to how foreigners living/studying/visiting/working in the US are viewed now, but rather politicians feeling they need to do something to 'fix' the embarrasment of people like the 9/11 hijackers getting into the country when there were all sorts of warning signs indicating they should not have been allowed to. Moreover, I'm not convinced that the impression most Americans have of foreigners has changed all that much. Sure, there are people who don't want any foreigners here, but they were around long before 9/11.
Most of all, it just saddens me to hear people say that the US is not as nice and friendly as it used to be. The government has stricter policies now, but that not necessarily what I perceive as less friendly towards foreigners. Rather, there's a feeling of allowing people to come here like before, but expending more effort on making sure they're not going to cause trouble. Unfortunately, the mess that is US immigration (cost, waiting time, inflexible process) has only gotten worse since these new policies were put into effect.
That said, there was a dramatic drop off in the number of foreigners applying to study in the US a couple years ago. However, I was reading something not too long ago about how the number of such applicants is growing again. The new numbers still don't make up for a certain permanent drop in applicants though. A couple of years ago, when I was studying in Germany, I met a lot of students (primarily from India and Pakistan) who had decided to study there instead of the US, simply because it's so much of a hassle to get a US student visa nowdays.
It sounds like Paul Graham himself has done really well for himself with the cram 40 into 4 method. However, for every person who does as well as him by taking that path, there have to be hundreds or thousands of burned-out workaholics in stressful, financially unstable work situations. If you're the type of person to do that sort of thing anyways, you might as well. I don't think it makes sense though to encourage others to do that.
Rarely is there a Computer Science graduate student who didn't wish he or she had more mathematical background. Now, if you're just going to be a "hacker" ala Graham, I admit more math is probably not necessary.
Slashdot Mirror
Uh, ya. Now explain to me why thousands of students from those countries (usually the best ones at that) continue to pursue graduate course work in North America and Europe.
Here's his list of places where major graphics research is going on in North America:
He also of course adds Wisconsin to the list too. I'd personally place it in the last category, since they have a small, young but respectable program. Generally, I agree with this list although there are probably a whole bunch of other places with just one faculty member doing good work.Having spent time studying and researching in Europe, there are some good graphics groups there too. In the UK, I know of two places off hand: Cambridge (Malcolm Sabin doing stuff with surfaces and geometric modeling) and Bath (Phil Willis and some others). In Germany, there is the Max Planck Institute (which does everything in English and is a very strong group although I admit I used to work there, so I'm probably biased), TU Darmstadt (Alexa), RWTH Aachen (Kobelt) and Tuebingen (Strasser). In Switzerland, there is the ETH in Zuerich (Markus Gross), the EPFL in (I think) Lausanne (Nadia and Daniel Thalmann) and Basel (Thomas Vetter who not so long ago left Freiburg to start a new graphics group there). In France, there is a group at the INRIA in Grenoble (Marie-Paul Cani). The above is certainly not an exhaustive list, just names that come to mind off the top of my head.
Most of these programs in continental Europe probably require you to already have your MSc before beginning PhD studies, but some have MSc programs as well. I know the MPI has one and awards scholarships even to foreign students. Generally, it's easier to get funding with only a Bachelor's degree in the US than in most of Europe. I've also heard the funding situation in the UK is not really very good, which is probably another reason to consider places like the US, Canada, Germany or Switzerland.
Competition for admission to US schools is fairly intense and is a time consuming and expensive process, but there are some really amazing opportunities here. I'm currently a grad student (graphics, visualization and scientific computing) at UNC and we probably have the largest number of graphics and imaging faculty and grad students under one (academic) roof in the world. We're not as hard to get into as a lot of the top schools like Berkeley, MIT, Stanford or CMU, but we still only take about 1 in 7 applicants or something. The key is to apply to as many places as you can afford, but only apply to places you'd actually attend.
Happy searching!
That website seems to imply that you need an additional order n amount of space to do heapsort. However, it can be done in place without too much effort:
1. Take your input array containing n elements and perform a build maxheap operation on it. This takes O(n) time and can be done in place.
2. Swap the root (the maximum element in the heap) with the last value in the part of the array still occupied by the heap, effectively removing the root from the heap and placing it in the sorted portion at the end of the array. O(1)
3. Percolate the new root (that was just swapped) down the tree to guarantee the heap property holds for it. O(log n)
4. Go back to step 2 so long as there are still elements left in the heap.
Still, quicksort usually tends to be faster in most cases because of overhead. Heapsort really pays off though when you don't want to sort the entire sequence. For example, maybe you just want the 10 largest or smallest elements. In that case, you only have to do the full O(n) build heap operation instead of a full O(n log n) sort.
The CS department here is a top 20 research-based program with a very strong focus in the computer graphics and imaging areas.
Just like they do for virtually every other plant. How many people grow more fruit, vegetables, grain or herbs than they buy? How about people who brew more of their own beer than buy it in the supermarket? How about people who grow their own tobacco?
Usually, it just turns out to be easier and more convinient to buy most things retail. Of course hobbyists will always do it for the fun or the challenge, but there would be plenty of money to made by companies especially compared to the basically $0 they make right now.
Even with an anonymous proxy, it's not terribly hard to guess who the possible reviewers are anyways. There are only so many people/groups in a field and papers chairs try to send your paper to somebody who will be able to competently evaluate it.
I'm not convinced this is true. Mostly, there are so many PhDs (and really in the big picture of society, how many are there?), because with a PhD you get to be in charge of stuff. That doesn't mean you get to be the next Einstein. It just means people with the money trust you more to lead a research project or study. As long as there are PhDs, people with an MS or BS will continue to work underneath them.
The predator is already there. How many people go to college? How many graduate? How many become graduate students? How many finish graduate school with an MS/MA? With a PhD? How many get tenure track jobs?
The numbers have already been dramatically winnowed down before somebody gets to the level of Assistant Professor, although perhaps the cut at each of the above levels needs to be tighter to get the numbers down. The problem of overproduction seems to vary a lot by field though too. There is a sizeable glut of English PhDs who have trouble finding traditional jobs that require such a degree. A number of programs are aware of the problem and intentionally admit fewer PhD students than they can support because overproduction has been such a problem.
At the same time, in my research group in a Computer Science deparment (good, but not top 5 or anything), graduates (MS and PhD) have not really had trouble getting decent jobs where a graduate degree is an asset even in today's less than stellar job market.
You start towing things and it's amazing how fast those fines start getting paid though!
I'm sure Intel will spend the effort to bring about a port of their C/C++ and probably Fortran compliers to x86 OS X. With this new partnership, it's obviously in their own interest for software on these Intel Macs to run as fast as possible.
Apple has obviously got an x86 gcc for Tiger and has already begun the process of porting the frameworks, most of which will probably not require massive porting effort. Frameworks like vecLib will probably require some more work to use SSE instead of Altivec though.
Even the concerns about things like endianness are not really a problem so long as the code was written the right way in the first place.
Such a switch to Intel chips would do very little towards true Win32 binary compatability. Obviously, you can't run an x86 Linux binary on Windows, just like you can't run a Windows binary on x86 Linux (at least without WINE or other similar compatability software). The same will hold true for an x86 version of OS X. Right now, products such as Virtual PC allow Mac users to run Windows programs under emulation. A switch to Intel chips would mean such programs could run Windows programs much more quickly, but that's still a long, long way from OS X being able to handle the complex Windows APIs natively.
Well, actually they don't pay the RIAA. They pay ASCAP, BMI and SESAC, which technically represent the song writers and performers as opposed to the big record companies. If I remember correctly, non-profit stations do not have to pay ASCAP or SESAC any royalties at all. BMI is paid, but at a dramatically reduced rate (only paid for about one week a year at a place I used to work). Right now, in the US the cost is somewhere around 8 cents per song, which for a commercial station is nothing compared to even labor costs.
Germany does have a tiered high school system, but it's not as if the education system only serves the elite. Everybody is entitled to an education there. If anything, the average job in Germany requires substantially more schooling that the average American job. For example, if you want to be a flower dealer or a hair stylist, there's a formal (and usually long) practical education path to be followed.
It can help prevent prostate cancer.
I always thought people buying computers at supermarkets (like Aldi) was one of the weirdest things computerwise about Germany. One time, I saw an acquaintance of mine return a computer at Plus even. The cashier gave him 2 500EUR notes taken from the normal register for the refund even.
I honestly can't imagine anybody in the US buying such an expensive item cash (especially at a supermarket) or the supermarket even having that amount (especially in that denomination) in a normal register (maybe in a safe in the manager's office or something). Personally, I wouldn't even feel safe walking around with 1000EUR in my pocket. Germany's definitely much more of a cash culture than the US (where people can and do use credit cards for virtually everything) and this all happened in a very safe suburb, so it does make sense I guess.
Moreover, it doesn't even necessarily mean that much inside a given architecture, especially these days with entire system performance often being more dependent on cache, memory, disk and in certain cases GPU.
SIGGRAPH even generally waves the conference fees for people who have a paper accepted.
Oh, there definitely are advantages to having an MS...
People working on an MS at a major research university will most likely be involved in some sort of research and will be required to take some more advanced coursework. More than anything, I think the process solidifies the student's knowledge of the field and allows him or her to focus a bit on one or two specific areas of interest in order to gain some depth not usually possible in an undergraduate program. Additionally, students are expected to work independently to a much greater degree. The more detailed knowledge, a demonstrated ability (at least in comparison to undergrad) to work independently, and the fact that generally stronger students go to graduate school means MS grads are attractive to many companies. Also, for a lot of jobs that require research skills, a MS can often demonstrate a basic competency. There are a good number of jobs out there (more than for PhDs) that require only this level of experience.
The advisor situation can play a role as well. Many profs have contacts (often former students) at companies doing interesting work. When a student graduates, the advisor often turns to their network to start asking around about what's available for this student with a particular skill set. In particular, during an MS every student (at least at research-oriented programs) tends to specialize more than they did as an undergrad. In many areas, this means you have specialized knowledge. A lot of MS grads use this to get jobs where more interesting or cutting-edge work is being done.
Personally, I'm not entirely convinced MS grads have more job opportunities. However, they do in general tend to have *better* ones. Now, a lot of the above is probably not true for people who are already working and doing a coursework based MS at night. My impression is that a lot of those programs give the student a more solid background in CS, but mainly serve as a way to demand a higher wage or help in getting into management.
Perhaps the saddest part about France going to Microsoft for this project is that whatever data is produced is more likely to be locked into some proprietary format. That could be particularly unfortunate, since these cultural resources really belong to all the people of France and should therefore be made as accessible as possible.
I dunno. The older I get, the less interested I find myself in younger generations. They're all just in a different stage of their lives or something. Of course, if it's just a year like the article author, that's no big deal. However, once you're 21, it's just not as cool to be buying beer for and hanging out with only 18 year olds.
It always amazes me how many Germans will claim that everybody in their country (or Europe, although I haven't heard that so much) can speak English. It's simply not true. The ability of people to speak even basic English can get very spotty.
First, more educated people (like those who went or current go to a regular university) are more likely to be able to speak English, while a lot of blue collar workers don't speak it at all.
I've found that region also plays a big difference. I lived in the Saarland for two years and more or less only people associated with the university could speak English. Even at the foreigners office which handled visas, I think only one worker spoke English (and not that well) and these people deal with foreigners all day, every day. Apparently, only recently has the school curriculum changed, so that English (as opposed to French) is taught there as the first foreign language (or so I'm told). A guy I worked with (a German) was telling me about a job offer he had considered in Dresden (former East Germany, where learning English was very rare before the reunification), where two of his assistants (who he would have to keep because of civil service employment laws) could not speak English at all, even though they worked in a Computer Science department. On the other hand, it seems like many, many more people in the whole Rhine/Ruhr area (Cologne/Duesseldorf/Essen) speak good English.
Finally, there is the age gap. Younger people are WAY more likely to be able to speak English than the elderly, who probably were last in school before English enjoyed such broad status as the default international language.
On the plus side, there are a lot of people who can sort of understand basic English, even if they can't speak it themselves. Most Germans tend to be helpful and polite even if you can't speak a common language, but of course that's not true for everybody.
I'm convinced that if they had more subtitled television and movies, it would improve the situation over time. In a smallish city (definitely a city and not a town) like Saarbruecken, there was still (and probably still is) only one movie theater which showed English movies in the original version (not dubbed) and it was second-run place that showed only one per week. Television is just as bad, except of course, it's the same bad situation throughout the country. Pretty much the only English programs that were subtitled instead of dubbed when I lived there were Jackass, some dating show on MTV and Crank Yankers. All of them cater to a young audience, so it probably wasn't financially worth it to pay for the dubbing. I did get international CNN and BBC World though, which were both nice, but most people want more entertainment than just the news.
Most of all, it just saddens me to hear people say that the US is not as nice and friendly as it used to be. The government has stricter policies now, but that not necessarily what I perceive as less friendly towards foreigners. Rather, there's a feeling of allowing people to come here like before, but expending more effort on making sure they're not going to cause trouble. Unfortunately, the mess that is US immigration (cost, waiting time, inflexible process) has only gotten worse since these new policies were put into effect.
That said, there was a dramatic drop off in the number of foreigners applying to study in the US a couple years ago. However, I was reading something not too long ago about how the number of such applicants is growing again. The new numbers still don't make up for a certain permanent drop in applicants though. A couple of years ago, when I was studying in Germany, I met a lot of students (primarily from India and Pakistan) who had decided to study there instead of the US, simply because it's so much of a hassle to get a US student visa nowdays.
It sounds like Paul Graham himself has done really well for himself with the cram 40 into 4 method. However, for every person who does as well as him by taking that path, there have to be hundreds or thousands of burned-out workaholics in stressful, financially unstable work situations. If you're the type of person to do that sort of thing anyways, you might as well. I don't think it makes sense though to encourage others to do that.
Rarely is there a Computer Science graduate student who didn't wish he or she had more mathematical background. Now, if you're just going to be a "hacker" ala Graham, I admit more math is probably not necessary.