Don't forget dye lasers. Those can make green too, though they aren't used that often anymore. You can even get a green He-Ne. (Yes, green. He-Ne can lase at red, orange, yellow, green, and IR. Actually, I can't think of anything else that can make yellow other than a dye laser.)
You make a good point. People knew about the Microsoft millionaire phenomenon probably before l0pht. Perhaps I should say that the press coverage of l0pht inspired me more than Bill Gates. I've known quite a few people that did the "bunch of guys hacking on the same stuff in the same apartment which also happens to be where they live" thing. One group of people that I knew actually lived in the apartment next door to the hacking apartment. One of them is even famous, (for a geek,) for having done some things similar to what the l0pht guys did. He never mentioned specifically that he was inspired by them though. Maybe it is more accurate to say that the media coverage of l0pht showed geeks a glamorous way to use the skills they have to get social respect and money. (Show people why they should care about specific technical topics and offer related assistance for a fee.)
You're right that computer skills became more in demand because computers became more entrenched in society. My main point was that geeks gained substantial social respect because the media published a bunch of stuff that glamorized geekdom. I didn't mean to imply that (social respect == ability to command more income). Geeks were already making money and their skills were already valuable. A lot of people didn't realize that at the time though. The prototype geeks the media used at the time were the l0pht guys. I think it mattered that they were independant -- they weren't working for a corporation or anything like that.
Those guys also were probably among the first to make it publically obvious that computer skills were not simply vehicles for the personal amusement of the socially inept. The press at the time always discussed how they had one apartment for themselves, and one next door for their gear. They made money being hackers, (in the old sense of the word -- not crackers.) I imagine that a substantial part of the sudden increase in society's respect for geeks, (maybe mostly their potential incomes,) was due to the glamorous press exposure l0pht received at that time. Perhaps Slashdot should thank them -- I'm not really sure. It will be interesting to see what this new l0pht is like.
I had a similar discussion with a friend a few weeks ago about something similar. We were not talking about AI taking over the unskilled jobs; we were talking about the rather tight coupling between full time corporate employment and health insurance in the US. My contention was that having those uncoupled would allow much greater economic flexibility and production efficiency for the country because the risk of leaving a corporate job and starting a venture would be greatly reduced if comparable independant health insurance was affordable. The increased production efficiency in this case would come from a better match between an individual's money making activities and the individual's strengths and preferences. Just about everyone works better and harder at things they like than things they dislike, and being good at something generally makes someone more efficient. If one could get affordable health insurance without having to work full time for an employer that offers health care, one could start one's own venture doing something that matches one's strengths and preferences. Hopefully this would cause an increase in production and efficiency.
If that effect were real, the same could be true in the case of a society where the crappy brainless jobs are done by automatons of some sort. People thought the same thing, however, back when things like washing machines and dishwashers were becoming a common part of American life. They also said the same thing when computers were becoming common in business. Even if one were to correct for the relatively poor match between people's desires and skills and the capabilities and requirements that business software often involved, I'm not sure that people became more efficient or more self-actualized as a result of having computers at their disposal. It appears that jobs that a computer could do simply went away, and the other job descriptions just expanded to include the operation of a computer to do the job that used to be done by a person. Required hours went up, required qualifications went up, productivity went up, and the people whose jobs went away either retrained for something else or retired or had financial problems. In the case of washing machines, those who used to wash clothes by hand ended up making horrific casseroles involving jello and weiners, and that *obviously* wasn't an improvement. Any way you look at it, replacing unskilled labor with machines will probably make society more productive, but without education for those who no longer have a job, society will end up with lots of hobos and laughably bad food.
The article points out that they aren't suggesting that this laser ignition system would be used in cars. Their development efforts so far have been aimed at large natural gas engines. I'm guessing that they aimed for natural gas engines with this technology because it will be expensive and a lot of the costs will be relatively fixed per engine, thereby making it a prohibitively large fraction of engine cost for a small (cheap) engine, but more reasonable for a large one.
The article also mentions that reflected light from within the cylinder could be used to measure the fuel constituents to match engine parameters to the fuel. (Natural gas engines are frequently used for things like burning waste gasses from landfills to generate electricity. Landfill gas contains all sorts of weird stuff in concentrations that vary over time; engine emissions are a serious concern when burning landfill gas, but emissions can be controlled to a large degree by varying things like timing and mixture. Being able to vary engine parameters on the fly in a more accurate way could be a big deal for the environment in places that generate a substantial amount of electricity this way.)
Running your diesel engine on gasoline is inadvisable. The reason they suggest that you have a service check if you run your diesel engine on gasoline is because gasoline tends to wear out your injection pump and some kinds of injectors very quickly. The injection pump relies on the fuel for lubrication and injection pumps are fairly expensive. Even when biodiesel costs a lot more then petro diesel, using biodiesel as a diesel additive (2-5%) can be cost effective because the better lubricity of the fuel can make injection system components last longer.
The Fed is a bastardized institution that benefits nobody, and has very little real power.
I agree with you about the "bastardized institution" part, but the Fed is actually very powerful. Their economic power comes from the fact that they have the ability to write checks for any amount of dollars that are not drawn on any account whatsoever. The checks always clear, and yet the money doesn't come from anywhere. They can buy anything they choose with dollars that they just conjure into existence -- treasury bills, mortgage backed assets, oil, gold -- whatever. Basically, their job is to "cheat" monetarily to smooth out economic crises. They have all the powers required to do that. The only other economic entity that has anywhere close to that much power is perhaps the US Treasury, and the methods it has to meddle with economies are considerably more limited. Basically, it can auction off treasury securities in attempts to borrow money on behalf of the US government. (I believe the mints are also part of the treasury, but they just make physical money -- technically that money needs to already exist on paper before they can print bills or stamp coins to represent it, I believe.)
The Fed is truly an unusual economic entity. I think part of the problem that led up to this economic crisis is that very few people actually understand what the Fed and other similar central banks do. Understanding what they do and keeping abreast of the specific things that they have done over the past two decades or so would have clued people in long ago that a huge bubble had formed.
If we had a central bank where loans were given directly to those that need them for modest interest rates, we wouldn't be in this situation.
I think it is hard to tell what might have happened if that were the case. I don't think the Fed making loans directly to individuals would have prevented banks and hedge funds and other entities from manufacturing bottom line value by creating derivatives. I do find it annoying that if I wanted to sell a treasury security, I'd have to sell it to a dealer that would charge me a fee, (one of the Fed member banks,) instead of the Fed or the treasury directly.
But, as usual, greed prevailed. Oh, and by the way, communism != socialism, no matter what Rush Limbaugh tells you.
Agreed. The fact that a substantial portion of Americans and American companies are now economically disfunctional yet there has been tremendous demand for treasury securities and dollars suggests to me that the role America plays in the global economy is more like that of a bank than a nation that produces wealth. This disturbs me, even without considering the notion that nearly all American citizens are unimportant to America's role as a bank, because it implies that a "run" on such a bank could force America to try to be internally self-sufficient in a hurry. I doubt it would be successful at doing so; I don't think America knows how any more.
There was a consumer reports article about condoms a few years back -- it reported that the ones least likely to break were the LifeStyles Extra Thin or Ultra Thin or whatever they are called. I don't remember the data from the study, so the conclusion might not have been statistically significant; It seems implausible that the thinner condoms were less likely to break than the regular condoms, but maybe they use a different rubber formulation in the thin ones or something.
In my experience, LifeStyles condoms, (even the regular ones,) seem to be a better fit than the "magnum" or "large" Trojan products for a heat seeking moisture missile of somewhat larger than average diameter. I think this is because they are more elastic than the Trojans. I once had one of the Kimono ones break, but I've never had a LifeStyles break, and I've used a lot more of the LifeStyles.
Maybe that is because more effort is required to make sex with a circumsized schlong worthwhile, so people have less sex. Perhaps I should read the article first -- it might address that.
One major reason, (and as far as I know, there isn't any scientific evidence to support either the truth or the falsity of this argument,) is that even after having extensive condom-protected sex, the brain somehow thinks that you still haven't had sex. In my experience, one can get off in any number of ways that do not actually involve ejaculating inside a woman's vagina, and the grinding "need for sex" feeling never goes away. The pipes may be clean, relaxation may have set in, arousal may be difficult for a while, but I still feel like I haven't had sex, (the same result as having a wank.) I don't know how or why that occurs, but somehow, the brain *knows*. Maybe women have a substance in their vaginal secretions or something that gets absorbed by the tallywhacker thereby setting the "successfully mated" flag somewhere. I read an article a while ago that said something about the cervix secreting something that acts as a beacon for sperm so they know which way to go to find the egg. Perhaps the schlong needs to be exposed to that in order for satisfaction to be achieved?
The other thing that I don't hear talked about that effects me with regard to condoms is that after going at it for a while, (usually not that long,) I start getting significant peehole irritation. Once that starts to set in, I know I've got only a minute or two more before I'm not going to be able to stand it anymore -- it hurts like hell. The minute or two being contingent upon any pleasurable feelings overriding the discomfort enough that the discomfort doesn't cause rapid boner deflation. I wonder if that is caused by the lubricant that comes with the condom being of poor quality. That lubricant always feels very different from any other sex lube; the lube on "lubricated for someone's hypothetical pleasure" condoms usually feels kind of gritty.
Disclosure: I am circumsized. I once asked why my folks had me circumsized, and it was reported that the hospital just did it automatically. I wonder if they lied to me -- that the hospital would do it without asking first seems so ridiculous as to be implausible, but then again, I was born in the late 70s in America where circumcision was supposedly the norm. Messed up, that is. I'm glad American society seems to at least not be doing that automatically any more.
A big part of the problem with the roads in Michigan, (and they have always sucked -- that is nothing new,) is that there are too many of them. The big three Americna auto makers were probably responsible for that. Michigan's economy has relied on them for so long that when GM says, "no busses -- build more roads so we can sell more cars," the legislature asks how many they want built and where they want them. There has been speculation amongst vehicle owners in Michigan, (probably for decades now,) that an additional reason there are so many potholes, (and some are truly epic in proportion,) is that suspension damage causes more parts sales for the auto makers. Similar speculation exists regarding the extensive use of rock salt to melt ice on the roads, since it also causes extensive corrosion of steel parts.
Michigan got a huge pile of federal money at least a decade ago to fix roads. I don't remember how large it was, but I remember thinking that it was larger than the entire annual budget of the state at the time. Anyway, the roads were in terrible shape and there was no way the state could afford to fix them all. The budget shortfall for fixing the roads was probably measured in dB. They were starting to become dangerous in places and had been causing damage to vehicles for quite a while.
The management of the "big three" were incapable of or unwilling to appreciate the fact that they needed to continuously adapt to changes in the world and their customers' desires in order to survive. Whatever mindset crippled their effectiveness in that way seems to be a ubiquitous problem with large, successful companies. It really is sad how they had the engineering talent, the manufacturing technology, the capital, and the support of the government and die-hard "buy American" customers for over two decades and still managed to cock it all up, pretty much destroying the economy of an entire state, (and damaging the economies of several others.)
I don't post very often, but I felt compelled to reply to this one because of the profoundly positive effect that a study abroad program had on me. I did my final year of a BS in Physics abroad at the University of Bath in the UK. The U of Bath is a smallish, selective school that is primarily technical in nature. It might be comparable to, say, CalTech. While I can say that the focus of the program there was substantially different from the program at my home university, (Purdue), it was also an excellent program. I knew a couple people there who were studying CS. From what I heard about it, it seemed to be a good program also. The U of Bath has lots of international programs, so the typical international student will be living with people from all over the place. In the immediate proximity of my residence where people from France, Germany, Sudan, Korea, UAE, Belgium, Mexico, Spain, Italy, Argentina, Russia, Ukraine, Greece, and Hungary. (Those are just the ones that I remember.) Having gone to a few different Universities for different purposes over the years, I can say that the biggest thing that differentiates them at the undergraduate level is the type of students that they attract. Basic CS theory is basic CS theory, and you would be studying the same stuff no matter where you go for the most part; the difference between universities is mostly who you would be studying it with.
In general, I think most Americans would be well served by the experience of going somewhere else and living for a while. It doesn't really matter that much if you do school or work -- just go be somewhere that isn't the US for long enough to forget that you aren't in the US any more. Once you've gone a few days without thinking about the fact that you aren't in the US, you're probably in a good position to get a glimpse of the US as non-Americans see it. You'll also then be in a position to evaluate beliefs you may have about the US and the "American way," as some like to call it. All of this tends to cause a huge increase in self confidence. Once you realize that everything you really need to live for a year either fits in a big suitcase or can be acquired for a relatively small amount of money and that human nature is pretty much the same no matter where you go, you'll never look at anything the same way again. For me, this has resulted in a large improvement in my overall state of mind, and it allowed me to pursue happiness much more effectively.
I got a lot of things out of my Physics BS, but I'd put the study abroad experience about on par with the Physics in terms of what is valuable to me now, 6 years after graduation. I even failed a couple Physics classes while I was there and had to take them again when I got back. It was still worth it. I can't recommend a year abroad highly enough. Go somewhere -- it doesn't really matter so much where you go. Try to spend a year there if you can. Be open minded and respectful of the natives when you get there, and I can pretty much promise that you won't regret it.
The cooling system was the first thing I thought about. I used to work for a company that built prototype lasers for defense and medical applications. We were working on something that could have been an eventual competitor to a laser like the one in TFA. There are lots of commercially available lasers in the 100W-1kW category that are probably similarly designed. Many of the ones that I have seen have a laser "head" which contains all the optics, the pump source, and the laser rod. I've seen ones roughly the size of a toaster. The cooling unit, (which can't be avoided for any laser that I have ever discussed with any of my colleagues, regardless of how much money you have at your disposal to commission a cooling-free one,) is typically the size of a small refrigerator. The problem with solid-state lasers is that the laser rod itself is typically about 3-5mm in diameter and maybe 10-15mm long. The quantum efficiency of the photon conversion process is not 100%. In fact, it isn't even particularly close to 100%. (The quantum efficiency is a function of the laser system in question. For example, any Nd:YAG laser with the same doping concentrations will have the same theoretical maximum quantum efficiency. Nd:YVO4 (wouldn't surprise me if Nd:YVO4 is in the laser in TFA), has a different one, etc.) Whatever light from the pump source isn't converted to photons in the output beam usually ends up being deposited as heat somewhere else in the system. With solid-state lasers, a large amount of this heat gets deposited in this rather small crystal rod. That ends up being the power-limiting factor most of the time. Whatever Northrup Grumman is doing to make a solid-state laser function at 15kW must involve *lots* of cooling. Either the cooling solution has to be really cold, (unlikely -- too much thermal deviation causes materials to change in size too much causing optical misalignment which causes all sorts of nasty things to happen), or the flow volume must be very large, and they must have developed some way to make an unusually good thermal interface between the laser rod and the cooling solution. Another possibility is that they have multiple gain stages, or that they actually have multiple solid state lasers in that box whose beams are already combined. If they do have multiple lasers in that box, then an M^2 of "nominally 1.5" is almost unbelievably excellent. (Unless that figure came from a "hero" experiment wherein they got an M^2 of 1.5 for 3us before it all went south.) (For those of you who want to know more about laser beam quality and what "nominally 1.5x diffraction limited" really means, here is a rather lackluster summary that includes most of the relevant information: http://en.wikipedia.org/wiki/Beam_parameter_product The "times diffraction limited" value in TFA is the "M-squared" number talked about in the summary.) Even with the multiple gain stage option, (which has problems as well), they still need to get a lot of heat out of that box, and they need to be able to control the temperature of the laser rod and most of the stuff that is close to it, which means they can't just connect up some large heat sink (like the body of the vehicle, for instance), and go to town. The temperature needs to be stable. Since they probably aren't running the laser all the time, (presumably they don't want to spend the power having it running and then just open a shutter when they want to melt something -- this would require dissipating all the heat the shutter would be absorbing too,) the cooling system needs to be able to provide relatively little cooling at times, and then lots when they turn the thing on. In fact, it probably needs to be able to heat too, since it would need to be usable in cold weather in a military application.
A lot of research has been done on fiber lasers for this type of application because the heat generated in the gain medium is spread out over the length of a fiber instead of concentrated in a sm
Actually, in a short time, you will probably be able to buy a refurbished one for a substantial discount. (40% is probably about right.) The refurbished ones bought through the Apple store have the same warranty as a new one. I bought a refurbished MacBook Pro from the Apple website about two years ago for $1500 when a new, similarly configured one was $2500. The refurbished one was a slightly older revision, and so it was slightly inferior to the newer revision, but the newer revision was definitely not worth $1000 more, in my opinion.
Interesting point. We won't really find out what large companies might plan to do with DNA data by releasing data on 10 people anyway. I'm sure that the expense of trying to use that data wouldn't be worth the gain that might come from making business decisions regarding 10 people. At the same time, these people are also releasing information about any offspring, parents, siblings, and other blood relatives they might have now or in the future. I keep hoping that at some point people in charge of entities that exploit other people will get the clue that they might, some day, end up at the other end of the exploitation deal, and will then stop doing it. I'm not holding my breath though.
What I mean by "make an in-house version" is that if they are concerned about new binaries causing problems, they could, in the case of something like Vim, which doesn't connect to outside machines and pose a direct security risk, simply scrutinize the source for and then build a binary and store that binary on-site and permit people to use only that one. This means that some of the benefits of open source are lost, but at least you get to use the software for the most part.
They don't necessarily have to scrutinize source -- presumably the notion that software might be dangerous is also true in the case of commercial software and if that is true, then they should have methods of qualifying specific installations of a program as safe, regardless of the type of transaction through which they would acquire the software. I realize that companies often do not have such qualifying methods and instead rely on the implied threat of a lawsuit to prevent commercial software vendors from selling them malware, (either intentionally malicious or not,) but the legal recourse is usually far inferior to just having software that does only what the users think it does. Legal recourse is an expensive and risky endeavor that often doesn't really make up for all the damage done; there are, of course, examples of where the suing entity made a killing from their victimization, but there are a lot of far less exciting outcomes where the victim still ended up taking various types of loss even if they won the lawsuit. You could point that out to them, but keep in mind that you will be essentially pointing out that their usual arguments are incorrect and that you know they are actually just engaging in ass-covering. This may go over badly.
You can still suggest that they qualify a binary, though. That is reasonable, in my opinion, if you can justify the utility of the software you want in monetary terms regardless of what arguments you may present as to why their no-open-source policy doesn't make sense.
I really don't think AMD is hosed. AMD still has lots of assets, not all of which are centered around processors and (now) GPUs. They could unload some of these assets to stay afloat if need be. Additionally, I think a lot of the press about Phenom being completely owned by the Core 2 architecture is a bit biased and exaggerated. One thing that people don't seem to talk much about is the fact that on a Phenom die, all cores share a common L3 cache. Core 2 Quad processors do not share *any* cache among all 4 cores. This means that if a process gets moved to a core on the other die of a Core 2 Quad, the cache needs to be reloaded, which means going through the relatively high latency memory access process again. As I understand it, "processor affinity" code in OS kernels is still fairly non-optimal, so this probably happens fairly often. I'm sure someone out there in Slashdot land knows some value about how often this actually happens. If so, it would be good if you'd post it, since I'm curious to know what it is.:)
With regard to benchmarks, situations where multiple processes share a chunk of memory and situations where processes get shuffled around to different cores frequently are examples of when a Phenom might be a Core 2, other things being equal. I have not been able to find any benchmarks of these two architectures that include, for example, PostgreSQL serving a large number of queries simultaneously, (thereby using all cores most of the time.) The benchmarks all seem to revolve, at best, around measuring the performance of a single process, possibly while something else is also running. It wouldn't surprise me if Phenom spanks Core 2 in some situations that aren't talked about in the reviews much.
Reviews that consider some of these other situations that identify some places where Phenom beats Core 2 could do a world of good for AMD.
I have not done a complete MSc level course in astrophysics, but I have done most of a complete MSc level course in Physics, and I have taken a course on astrophysics. In my opinion, you will get the majority of the background that you need from a "modern physics" textbook. (There is a class that most physics students in the US take called "modern physics" that includes a wide variety of topics that is meant to be preparation for some more specific advanced courses. The textbook that I studied from is called "Modern Physics from (alpha) to (Z0)", (with the alpha and z0 substituted for their greek renderings). The author is James William Rohlf. While I haven't looked extensively at other texts, this one is pretty decent in that it gives a good overview of most of what you will need. It is several years old, (at least the version I have), so you might be able to find it cheaply.
If there is a one single topic that I would suggest you study for background, it would be statistical mechanics. I would probably say that about any scientific discipline, however, so perhaps you will find that the core of astrophysics is really something else.
I used to work for a company that was attempting to manufacture fiber-based AWG (Arrayed Waveguide Grating) devices back in about 2000. At that time, the fraction of fiber in the ground that was dark was thought to be about 99%. The devices we were testing were capable of multiplexing 16 channels together on to one fiber. The standard speed for a fiber link over single mode fiber is 2.5 Gbit/s, and a fiber link requires a pair of fibers, (for bi-directional traffic.. I suppose if you only wanted to send data one way, you could use a single one.) At that time, there were multiple competitors that had 40 channel devices based on some different technologies. When I stopped paying attention to what was available, 160 channel devices were being talked about and 80 channel devices were on the market. The cost of one of these AWGs was about $20k, (to buy as a customer, not the cost of production), and they have since come down in price by a large amount. You would need one on each end of the fiber. If we assume that 80 channel devices are available, and 1% of the fiber in the ground (the portion that was used) was 1 pair, then there were at least 8000 2.5 Gbit/s channels available in whatever segment of the network contained "99% dark fiber".
I haven't been able, in the last few minutes, to find stats on current backbone traffic levels, but I seriously doubt that the amount of potential long-haul fiber capacity is the reason for laying these cables. The only valid reasons I can see are that the existing ones are owned/controlled by entities that aren't cooperating or utilizing their cables very well or that redundancy is desired. The article states that Google is planning on running a cable from the US to Japan. I have to assume that this is more because the owners of existing cables are not cooperating. This might be the start of investment in a highly fractured network which does not have the redundancy that the internet was originally designed to provide.
I've been curious about the direct streaming feature. On the website, it says XP or Vista is required with certain service packs, etc. What format do they use for this and is Windows really necessary? Requiring Windows is sure to irritate customers as well.
I've been learning the same kinds of things with one of the ADD stimulants and an antidepressant for the last few years. One of the best things about these drugs is that they show you what "the other side" feels like. When you've been predominantly in one mode every day since puberty, it can be impossible to even understand what people mean when they ask simple questions about what you are going to do. For instance, there was always cognitive dissonance when a manager would ask me something like, "can you do thing X by deadline D?" My mental answer to a rhetorical question like that was always something like: "maybe." I usually answered verbally in the affirmative because I was aware that it was expected that I would, despite knowing that the actual answer was not so clear. In situations like that, I always felt like I had entered another dimension where people continuously behave in ways that don't make sense. The reason for this, I later found out, was because most of the time, normal people can say "yes" or "no" to a question like that and be sure that unless something extremely unusual were to happen, they would be correct; they either can do it, or they can't, and they know ahead of time which is true. Their reasons for saying "yes" or "no" didn't usually include thoughts like: "technically yes, and I've done it before -- it is actually pretty easy, but my track record for managing to do it is dismal for reasons that I don't understand, so an objective interpretation of the data suggests that a long-winded and unsatisfactory answer that indicates that I really don't know if I will do X by deadline D, and that the reasons for this are beyond my realm of comprehension, is what my reply *should* be, but I'm going to say yes anyway because any other response is going to piss people off. I absolutely hate corporate America -- this kind of weird asking of questions that are obviously unanswerable in an honest fashion must mean that either people are screwing with me or they are intensely stupid."
I didn't realize all this in a concrete manner until somewhere in my late 20s after trying some of these drugs that made things like mental crises, and the utter inability to turn my brain off to focus or sleep optional. I've since taken them on and off as necessary, but being able to intuitively understand what it means to be able to cause one's actions to align with one's intentions on a regular basis is invaluable. I can say with complete honesty that I really didn't understand how the world worked before.
Agreed. Using a "scripting" or "rapid prototyping" language to figure out what you need to write is definitely the way to go. If then, you need a substantial performance improvement, rewriting the important parts in something like C is a good solution. One of the reasons that I prefer Unix-like OSs to others is that the whole shell/pipe thing makes doing this sort of stuff really easy. For instance, it is easy to write a Ruby script that parses a file, makes a data set, feeds it to a C program to crunch via a pipe and then returns the results via a pipe.
Simple data partitioning multiprocessing is easy to do as well this way; just divide up the data set and feed it to multiple instances of the C program.
I am aware that a great majority of parallel operations cannot really be done this way, but when they can, the "divide and pipe" method leaves little to screw up. One also loses little efficiency if the amount of computation done by the C program is large compared to the cost of spawning another process. Parsing the file in Ruby is usually relatively computationally cheap since Ruby mostly does this kind of thing using C libraries anyway. (depends on the format of the file, of course, but most of the common ones are done with C libraries).
As some others have previously mentioned, dealing with formatted data is one of the things that Fortran is good at, and so I would imagine that this technique could work quite well using Fortran instead of C for a data-crunching subprogram. I haven't tried Fortran in that capacity however.
Most of the comparisons I have seen have shown that in many situations, Java can approach the speed of C. That being said, I can understand that there may be some situations where the C compiler in question gets itself into an optimization corner and produces code that is less efficient than a reasonable Java counterpart. I'd be really really surprised, however, if you were to take a cross section of typical programs, write them in Java and C, (in a fairly normal manner), and had the Java programs beat the C programs on average. I don't really see how that is possible unless you were specifically trying to use examples that the exploited the best properties of the JVM and the worst properties of the C compiler and processor in question.
Slashdot Mirror
Don't forget dye lasers. Those can make green too, though they aren't used that often anymore. You can even get a green He-Ne. (Yes, green. He-Ne can lase at red, orange, yellow, green, and IR. Actually, I can't think of anything else that can make yellow other than a dye laser.)
You make a good point. People knew about the Microsoft millionaire phenomenon probably before l0pht. Perhaps I should say that the press coverage of l0pht inspired me more than Bill Gates. I've known quite a few people that did the "bunch of guys hacking on the same stuff in the same apartment which also happens to be where they live" thing. One group of people that I knew actually lived in the apartment next door to the hacking apartment. One of them is even famous, (for a geek,) for having done some things similar to what the l0pht guys did. He never mentioned specifically that he was inspired by them though. Maybe it is more accurate to say that the media coverage of l0pht showed geeks a glamorous way to use the skills they have to get social respect and money. (Show people why they should care about specific technical topics and offer related assistance for a fee.)
You're right that computer skills became more in demand because computers became more entrenched in society. My main point was that geeks gained substantial social respect because the media published a bunch of stuff that glamorized geekdom. I didn't mean to imply that (social respect == ability to command more income). Geeks were already making money and their skills were already valuable. A lot of people didn't realize that at the time though. The prototype geeks the media used at the time were the l0pht guys. I think it mattered that they were independant -- they weren't working for a corporation or anything like that.
Those guys also were probably among the first to make it publically obvious that computer skills were not simply vehicles for the personal amusement of the socially inept. The press at the time always discussed how they had one apartment for themselves, and one next door for their gear. They made money being hackers, (in the old sense of the word -- not crackers.) I imagine that a substantial part of the sudden increase in society's respect for geeks, (maybe mostly their potential incomes,) was due to the glamorous press exposure l0pht received at that time. Perhaps Slashdot should thank them -- I'm not really sure. It will be interesting to see what this new l0pht is like.
I had a similar discussion with a friend a few weeks ago about something similar. We were not talking about AI taking over the unskilled jobs; we were talking about the rather tight coupling between full time corporate employment and health insurance in the US. My contention was that having those uncoupled would allow much greater economic flexibility and production efficiency for the country because the risk of leaving a corporate job and starting a venture would be greatly reduced if comparable independant health insurance was affordable. The increased production efficiency in this case would come from a better match between an individual's money making activities and the individual's strengths and preferences. Just about everyone works better and harder at things they like than things they dislike, and being good at something generally makes someone more efficient. If one could get affordable health insurance without having to work full time for an employer that offers health care, one could start one's own venture doing something that matches one's strengths and preferences. Hopefully this would cause an increase in production and efficiency.
If that effect were real, the same could be true in the case of a society where the crappy brainless jobs are done by automatons of some sort. People thought the same thing, however, back when things like washing machines and dishwashers were becoming a common part of American life. They also said the same thing when computers were becoming common in business. Even if one were to correct for the relatively poor match between people's desires and skills and the capabilities and requirements that business software often involved, I'm not sure that people became more efficient or more self-actualized as a result of having computers at their disposal. It appears that jobs that a computer could do simply went away, and the other job descriptions just expanded to include the operation of a computer to do the job that used to be done by a person. Required hours went up, required qualifications went up, productivity went up, and the people whose jobs went away either retrained for something else or retired or had financial problems. In the case of washing machines, those who used to wash clothes by hand ended up making horrific casseroles involving jello and weiners, and that *obviously* wasn't an improvement. Any way you look at it, replacing unskilled labor with machines will probably make society more productive, but without education for those who no longer have a job, society will end up with lots of hobos and laughably bad food.
The article points out that they aren't suggesting that this laser ignition system would be used in cars. Their development efforts so far have been aimed at large natural gas engines. I'm guessing that they aimed for natural gas engines with this technology because it will be expensive and a lot of the costs will be relatively fixed per engine, thereby making it a prohibitively large fraction of engine cost for a small (cheap) engine, but more reasonable for a large one.
The article also mentions that reflected light from within the cylinder could be used to measure the fuel constituents to match engine parameters to the fuel. (Natural gas engines are frequently used for things like burning waste gasses from landfills to generate electricity. Landfill gas contains all sorts of weird stuff in concentrations that vary over time; engine emissions are a serious concern when burning landfill gas, but emissions can be controlled to a large degree by varying things like timing and mixture. Being able to vary engine parameters on the fly in a more accurate way could be a big deal for the environment in places that generate a substantial amount of electricity this way.)
Running your diesel engine on gasoline is inadvisable. The reason they suggest that you have a service check if you run your diesel engine on gasoline is because gasoline tends to wear out your injection pump and some kinds of injectors very quickly. The injection pump relies on the fuel for lubrication and injection pumps are fairly expensive. Even when biodiesel costs a lot more then petro diesel, using biodiesel as a diesel additive (2-5%) can be cost effective because the better lubricity of the fuel can make injection system components last longer.
The Fed is a bastardized institution that benefits nobody, and has very little real power.
I agree with you about the "bastardized institution" part, but the Fed is actually very powerful. Their economic power comes from the fact that they have the ability to write checks for any amount of dollars that are not drawn on any account whatsoever. The checks always clear, and yet the money doesn't come from anywhere. They can buy anything they choose with dollars that they just conjure into existence -- treasury bills, mortgage backed assets, oil, gold -- whatever. Basically, their job is to "cheat" monetarily to smooth out economic crises. They have all the powers required to do that. The only other economic entity that has anywhere close to that much power is perhaps the US Treasury, and the methods it has to meddle with economies are considerably more limited. Basically, it can auction off treasury securities in attempts to borrow money on behalf of the US government. (I believe the mints are also part of the treasury, but they just make physical money -- technically that money needs to already exist on paper before they can print bills or stamp coins to represent it, I believe.)
The Fed is truly an unusual economic entity. I think part of the problem that led up to this economic crisis is that very few people actually understand what the Fed and other similar central banks do. Understanding what they do and keeping abreast of the specific things that they have done over the past two decades or so would have clued people in long ago that a huge bubble had formed.
If we had a central bank where loans were given directly to those that need them for modest interest rates, we wouldn't be in this situation.
I think it is hard to tell what might have happened if that were the case. I don't think the Fed making loans directly to individuals would have prevented banks and hedge funds and other entities from manufacturing bottom line value by creating derivatives. I do find it annoying that if I wanted to sell a treasury security, I'd have to sell it to a dealer that would charge me a fee, (one of the Fed member banks,) instead of the Fed or the treasury directly.
But, as usual, greed prevailed. Oh, and by the way, communism != socialism, no matter what Rush Limbaugh tells you.
Agreed. The fact that a substantial portion of Americans and American companies are now economically disfunctional yet there has been tremendous demand for treasury securities and dollars suggests to me that the role America plays in the global economy is more like that of a bank than a nation that produces wealth. This disturbs me, even without considering the notion that nearly all American citizens are unimportant to America's role as a bank, because it implies that a "run" on such a bank could force America to try to be internally self-sufficient in a hurry. I doubt it would be successful at doing so; I don't think America knows how any more.
There was a consumer reports article about condoms a few years back -- it reported that the ones least likely to break were the LifeStyles Extra Thin or Ultra Thin or whatever they are called. I don't remember the data from the study, so the conclusion might not have been statistically significant; It seems implausible that the thinner condoms were less likely to break than the regular condoms, but maybe they use a different rubber formulation in the thin ones or something.
In my experience, LifeStyles condoms, (even the regular ones,) seem to be a better fit than the "magnum" or "large" Trojan products for a heat seeking moisture missile of somewhat larger than average diameter. I think this is because they are more elastic than the Trojans. I once had one of the Kimono ones break, but I've never had a LifeStyles break, and I've used a lot more of the LifeStyles.
Maybe that is because more effort is required to make sex with a circumsized schlong worthwhile, so people have less sex. Perhaps I should read the article first -- it might address that.
One major reason, (and as far as I know, there isn't any scientific evidence to support either the truth or the falsity of this argument,) is that even after having extensive condom-protected sex, the brain somehow thinks that you still haven't had sex. In my experience, one can get off in any number of ways that do not actually involve ejaculating inside a woman's vagina, and the grinding "need for sex" feeling never goes away. The pipes may be clean, relaxation may have set in, arousal may be difficult for a while, but I still feel like I haven't had sex, (the same result as having a wank.) I don't know how or why that occurs, but somehow, the brain *knows*. Maybe women have a substance in their vaginal secretions or something that gets absorbed by the tallywhacker thereby setting the "successfully mated" flag somewhere. I read an article a while ago that said something about the cervix secreting something that acts as a beacon for sperm so they know which way to go to find the egg. Perhaps the schlong needs to be exposed to that in order for satisfaction to be achieved?
The other thing that I don't hear talked about that effects me with regard to condoms is that after going at it for a while, (usually not that long,) I start getting significant peehole irritation. Once that starts to set in, I know I've got only a minute or two more before I'm not going to be able to stand it anymore -- it hurts like hell. The minute or two being contingent upon any pleasurable feelings overriding the discomfort enough that the discomfort doesn't cause rapid boner deflation. I wonder if that is caused by the lubricant that comes with the condom being of poor quality. That lubricant always feels very different from any other sex lube; the lube on "lubricated for someone's hypothetical pleasure" condoms usually feels kind of gritty.
Disclosure: I am circumsized. I once asked why my folks had me circumsized, and it was reported that the hospital just did it automatically. I wonder if they lied to me -- that the hospital would do it without asking first seems so ridiculous as to be implausible, but then again, I was born in the late 70s in America where circumcision was supposedly the norm. Messed up, that is. I'm glad American society seems to at least not be doing that automatically any more.
I grew up in Michigan.
A big part of the problem with the roads in Michigan, (and they have always sucked -- that is nothing new,) is that there are too many of them. The big three Americna auto makers were probably responsible for that. Michigan's economy has relied on them for so long that when GM says, "no busses -- build more roads so we can sell more cars," the legislature asks how many they want built and where they want them. There has been speculation amongst vehicle owners in Michigan, (probably for decades now,) that an additional reason there are so many potholes, (and some are truly epic in proportion,) is that suspension damage causes more parts sales for the auto makers. Similar speculation exists regarding the extensive use of rock salt to melt ice on the roads, since it also causes extensive corrosion of steel parts.
Michigan got a huge pile of federal money at least a decade ago to fix roads. I don't remember how large it was, but I remember thinking that it was larger than the entire annual budget of the state at the time. Anyway, the roads were in terrible shape and there was no way the state could afford to fix them all. The budget shortfall for fixing the roads was probably measured in dB. They were starting to become dangerous in places and had been causing damage to vehicles for quite a while.
The management of the "big three" were incapable of or unwilling to appreciate the fact that they needed to continuously adapt to changes in the world and their customers' desires in order to survive. Whatever mindset crippled their effectiveness in that way seems to be a ubiquitous problem with large, successful companies. It really is sad how they had the engineering talent, the manufacturing technology, the capital, and the support of the government and die-hard "buy American" customers for over two decades and still managed to cock it all up, pretty much destroying the economy of an entire state, (and damaging the economies of several others.)
I don't post very often, but I felt compelled to reply to this one because of the profoundly positive effect that a study abroad program had on me. I did my final year of a BS in Physics abroad at the University of Bath in the UK. The U of Bath is a smallish, selective school that is primarily technical in nature. It might be comparable to, say, CalTech. While I can say that the focus of the program there was substantially different from the program at my home university, (Purdue), it was also an excellent program. I knew a couple people there who were studying CS. From what I heard about it, it seemed to be a good program also. The U of Bath has lots of international programs, so the typical international student will be living with people from all over the place. In the immediate proximity of my residence where people from France, Germany, Sudan, Korea, UAE, Belgium, Mexico, Spain, Italy, Argentina, Russia, Ukraine, Greece, and Hungary. (Those are just the ones that I remember.) Having gone to a few different Universities for different purposes over the years, I can say that the biggest thing that differentiates them at the undergraduate level is the type of students that they attract. Basic CS theory is basic CS theory, and you would be studying the same stuff no matter where you go for the most part; the difference between universities is mostly who you would be studying it with.
In general, I think most Americans would be well served by the experience of going somewhere else and living for a while. It doesn't really matter that much if you do school or work -- just go be somewhere that isn't the US for long enough to forget that you aren't in the US any more. Once you've gone a few days without thinking about the fact that you aren't in the US, you're probably in a good position to get a glimpse of the US as non-Americans see it. You'll also then be in a position to evaluate beliefs you may have about the US and the "American way," as some like to call it. All of this tends to cause a huge increase in self confidence. Once you realize that everything you really need to live for a year either fits in a big suitcase or can be acquired for a relatively small amount of money and that human nature is pretty much the same no matter where you go, you'll never look at anything the same way again. For me, this has resulted in a large improvement in my overall state of mind, and it allowed me to pursue happiness much more effectively.
I got a lot of things out of my Physics BS, but I'd put the study abroad experience about on par with the Physics in terms of what is valuable to me now, 6 years after graduation. I even failed a couple Physics classes while I was there and had to take them again when I got back. It was still worth it. I can't recommend a year abroad highly enough. Go somewhere -- it doesn't really matter so much where you go. Try to spend a year there if you can. Be open minded and respectful of the natives when you get there, and I can pretty much promise that you won't regret it.
The cooling system was the first thing I thought about. I used to work for a company that built prototype lasers for defense and medical applications. We were working on something that could have been an eventual competitor to a laser like the one in TFA. There are lots of commercially available lasers in the 100W-1kW category that are probably similarly designed. Many of the ones that I have seen have a laser "head" which contains all the optics, the pump source, and the laser rod. I've seen ones roughly the size of a toaster. The cooling unit, (which can't be avoided for any laser that I have ever discussed with any of my colleagues, regardless of how much money you have at your disposal to commission a cooling-free one,) is typically the size of a small refrigerator. The problem with solid-state lasers is that the laser rod itself is typically about 3-5mm in diameter and maybe 10-15mm long. The quantum efficiency of the photon conversion process is not 100%. In fact, it isn't even particularly close to 100%. (The quantum efficiency is a function of the laser system in question. For example, any Nd:YAG laser with the same doping concentrations will have the same theoretical maximum quantum efficiency. Nd:YVO4 (wouldn't surprise me if Nd:YVO4 is in the laser in TFA), has a different one, etc.) Whatever light from the pump source isn't converted to photons in the output beam usually ends up being deposited as heat somewhere else in the system. With solid-state lasers, a large amount of this heat gets deposited in this rather small crystal rod. That ends up being the power-limiting factor most of the time. Whatever Northrup Grumman is doing to make a solid-state laser function at 15kW must involve *lots* of cooling. Either the cooling solution has to be really cold, (unlikely -- too much thermal deviation causes materials to change in size too much causing optical misalignment which causes all sorts of nasty things to happen), or the flow volume must be very large, and they must have developed some way to make an unusually good thermal interface between the laser rod and the cooling solution. Another possibility is that they have multiple gain stages, or that they actually have multiple solid state lasers in that box whose beams are already combined. If they do have multiple lasers in that box, then an M^2 of "nominally 1.5" is almost unbelievably excellent. (Unless that figure came from a "hero" experiment wherein they got an M^2 of 1.5 for 3us before it all went south.) (For those of you who want to know more about laser beam quality and what "nominally 1.5x diffraction limited" really means, here is a rather lackluster summary that includes most of the relevant information: http://en.wikipedia.org/wiki/Beam_parameter_product The "times diffraction limited" value in TFA is the "M-squared" number talked about in the summary.) Even with the multiple gain stage option, (which has problems as well), they still need to get a lot of heat out of that box, and they need to be able to control the temperature of the laser rod and most of the stuff that is close to it, which means they can't just connect up some large heat sink (like the body of the vehicle, for instance), and go to town. The temperature needs to be stable. Since they probably aren't running the laser all the time, (presumably they don't want to spend the power having it running and then just open a shutter when they want to melt something -- this would require dissipating all the heat the shutter would be absorbing too,) the cooling system needs to be able to provide relatively little cooling at times, and then lots when they turn the thing on. In fact, it probably needs to be able to heat too, since it would need to be usable in cold weather in a military application.
A lot of research has been done on fiber lasers for this type of application because the heat generated in the gain medium is spread out over the length of a fiber instead of concentrated in a sm
Actually, in a short time, you will probably be able to buy a refurbished one for a substantial discount. (40% is probably about right.) The refurbished ones bought through the Apple store have the same warranty as a new one. I bought a refurbished MacBook Pro from the Apple website about two years ago for $1500 when a new, similarly configured one was $2500. The refurbished one was a slightly older revision, and so it was slightly inferior to the newer revision, but the newer revision was definitely not worth $1000 more, in my opinion.
Interesting point. We won't really find out what large companies might plan to do with DNA data by releasing data on 10 people anyway. I'm sure that the expense of trying to use that data wouldn't be worth the gain that might come from making business decisions regarding 10 people. At the same time, these people are also releasing information about any offspring, parents, siblings, and other blood relatives they might have now or in the future. I keep hoping that at some point people in charge of entities that exploit other people will get the clue that they might, some day, end up at the other end of the exploitation deal, and will then stop doing it. I'm not holding my breath though.
What I mean by "make an in-house version" is that if they are concerned about new binaries causing problems, they could, in the case of something like Vim, which doesn't connect to outside machines and pose a direct security risk, simply scrutinize the source for and then build a binary and store that binary on-site and permit people to use only that one. This means that some of the benefits of open source are lost, but at least you get to use the software for the most part.
They don't necessarily have to scrutinize source -- presumably the notion that software might be dangerous is also true in the case of commercial software and if that is true, then they should have methods of qualifying specific installations of a program as safe, regardless of the type of transaction through which they would acquire the software. I realize that companies often do not have such qualifying methods and instead rely on the implied threat of a lawsuit to prevent commercial software vendors from selling them malware, (either intentionally malicious or not,) but the legal recourse is usually far inferior to just having software that does only what the users think it does. Legal recourse is an expensive and risky endeavor that often doesn't really make up for all the damage done; there are, of course, examples of where the suing entity made a killing from their victimization, but there are a lot of far less exciting outcomes where the victim still ended up taking various types of loss even if they won the lawsuit. You could point that out to them, but keep in mind that you will be essentially pointing out that their usual arguments are incorrect and that you know they are actually just engaging in ass-covering. This may go over badly.
You can still suggest that they qualify a binary, though. That is reasonable, in my opinion, if you can justify the utility of the software you want in monetary terms regardless of what arguments you may present as to why their no-open-source policy doesn't make sense.
I really don't think AMD is hosed. AMD still has lots of assets, not all of which are centered around processors and (now) GPUs. They could unload some of these assets to stay afloat if need be. Additionally, I think a lot of the press about Phenom being completely owned by the Core 2 architecture is a bit biased and exaggerated. One thing that people don't seem to talk much about is the fact that on a Phenom die, all cores share a common L3 cache. Core 2 Quad processors do not share *any* cache among all 4 cores. This means that if a process gets moved to a core on the other die of a Core 2 Quad, the cache needs to be reloaded, which means going through the relatively high latency memory access process again. As I understand it, "processor affinity" code in OS kernels is still fairly non-optimal, so this probably happens fairly often. I'm sure someone out there in Slashdot land knows some value about how often this actually happens. If so, it would be good if you'd post it, since I'm curious to know what it is. :)
With regard to benchmarks, situations where multiple processes share a chunk of memory and situations where processes get shuffled around to different cores frequently are examples of when a Phenom might be a Core 2, other things being equal. I have not been able to find any benchmarks of these two architectures that include, for example, PostgreSQL serving a large number of queries simultaneously, (thereby using all cores most of the time.) The benchmarks all seem to revolve, at best, around measuring the performance of a single process, possibly while something else is also running. It wouldn't surprise me if Phenom spanks Core 2 in some situations that aren't talked about in the reviews much.
Reviews that consider some of these other situations that identify some places where Phenom beats Core 2 could do a world of good for AMD.
I have not done a complete MSc level course in astrophysics, but I have done most of a complete MSc level course in Physics, and I have taken a course on astrophysics. In my opinion, you will get the majority of the background that you need from a "modern physics" textbook. (There is a class that most physics students in the US take called "modern physics" that includes a wide variety of topics that is meant to be preparation for some more specific advanced courses. The textbook that I studied from is called "Modern Physics from (alpha) to (Z0)", (with the alpha and z0 substituted for their greek renderings). The author is James William Rohlf. While I haven't looked extensively at other texts, this one is pretty decent in that it gives a good overview of most of what you will need. It is several years old, (at least the version I have), so you might be able to find it cheaply.
If there is a one single topic that I would suggest you study for background, it would be statistical mechanics. I would probably say that about any scientific discipline, however, so perhaps you will find that the core of astrophysics is really something else.
Good luck!
I used to work for a company that was attempting to manufacture fiber-based AWG (Arrayed Waveguide Grating) devices back in about 2000. At that time, the fraction of fiber in the ground that was dark was thought to be about 99%. The devices we were testing were capable of multiplexing 16 channels together on to one fiber. The standard speed for a fiber link over single mode fiber is 2.5 Gbit/s, and a fiber link requires a pair of fibers, (for bi-directional traffic.. I suppose if you only wanted to send data one way, you could use a single one.) At that time, there were multiple competitors that had 40 channel devices based on some different technologies. When I stopped paying attention to what was available, 160 channel devices were being talked about and 80 channel devices were on the market. The cost of one of these AWGs was about $20k, (to buy as a customer, not the cost of production), and they have since come down in price by a large amount. You would need one on each end of the fiber. If we assume that 80 channel devices are available, and 1% of the fiber in the ground (the portion that was used) was 1 pair, then there were at least 8000 2.5 Gbit/s channels available in whatever segment of the network contained "99% dark fiber".
I haven't been able, in the last few minutes, to find stats on current backbone traffic levels, but I seriously doubt that the amount of potential long-haul fiber capacity is the reason for laying these cables. The only valid reasons I can see are that the existing ones are owned/controlled by entities that aren't cooperating or utilizing their cables very well or that redundancy is desired. The article states that Google is planning on running a cable from the US to Japan. I have to assume that this is more because the owners of existing cables are not cooperating. This might be the start of investment in a highly fractured network which does not have the redundancy that the internet was originally designed to provide.
I've been curious about the direct streaming feature. On the website, it says XP or Vista is required with certain service packs, etc. What format do they use for this and is Windows really necessary? Requiring Windows is sure to irritate customers as well.
I didn't realize all this in a concrete manner until somewhere in my late 20s after trying some of these drugs that made things like mental crises, and the utter inability to turn my brain off to focus or sleep optional. I've since taken them on and off as necessary, but being able to intuitively understand what it means to be able to cause one's actions to align with one's intentions on a regular basis is invaluable. I can say with complete honesty that I really didn't understand how the world worked before.
I'd assume that they use a fairly normal balance shaft type setup to counteract the vibrations. ( http://en.wikipedia.org/wiki/Balance_shaft )
Agreed. Using a "scripting" or "rapid prototyping" language to figure out what you need to write is definitely the way to go. If then, you need a substantial performance improvement, rewriting the important parts in something like C is a good solution. One of the reasons that I prefer Unix-like OSs to others is that the whole shell/pipe thing makes doing this sort of stuff really easy. For instance, it is easy to write a Ruby script that parses a file, makes a data set, feeds it to a C program to crunch via a pipe and then returns the results via a pipe.
Simple data partitioning multiprocessing is easy to do as well this way; just divide up the data set and feed it to multiple instances of the C program.
I am aware that a great majority of parallel operations cannot really be done this way, but when they can, the "divide and pipe" method leaves little to screw up. One also loses little efficiency if the amount of computation done by the C program is large compared to the cost of spawning another process. Parsing the file in Ruby is usually relatively computationally cheap since Ruby mostly does this kind of thing using C libraries anyway. (depends on the format of the file, of course, but most of the common ones are done with C libraries).
As some others have previously mentioned, dealing with formatted data is one of the things that Fortran is good at, and so I would imagine that this technique could work quite well using Fortran instead of C for a data-crunching subprogram. I haven't tried Fortran in that capacity however.
Most of the comparisons I have seen have shown that in many situations, Java can approach the speed of C. That being said, I can understand that there may be some situations where the C compiler in question gets itself into an optimization corner and produces code that is less efficient than a reasonable Java counterpart. I'd be really really surprised, however, if you were to take a cross section of typical programs, write them in Java and C, (in a fairly normal manner), and had the Java programs beat the C programs on average. I don't really see how that is possible unless you were specifically trying to use examples that the exploited the best properties of the JVM and the worst properties of the C compiler and processor in question.