Some researchers don't bother to parallelize or optimize their applications because after the paper is written, the code isn't run any more. If you only need to run a few times to confirm your hypothesis, then why bother optimizing if the time between runs isn't terrible.
If it's not worth optimizing for parallel execution and the runtime is managable, then why the heck do you need a cluster? I can see if you're both lazy and impatient borrowing someone else's cluster for a few days, but if you actually have to build something, then why bother?
... being able to use natural math symbols and natural parenthesis makes the math so much more readable, and when 90% of your code is math, you come to appreciate it. I wouldn't have it any other way, at this point.
For example, the original Hebrew prophecy of the Messiah spoke of a "young girl," which in the Greek Septuagint -- which was the most popular "Bible" back when the New Testament was being written -- translated into a word meaning "virgin." Well, this eventually snowballed into the Immaculate Conception, but starting from the 1700s or so Christians started to recognize that what really happened was that young teenage Mary got herself knocked up.
If this is the best bible scholarship your book has, then throw it out the window and forget everything you learned from it, because the author doesn't know crap.
First, yes the Hebrew means "young girl" (culturally synonymous with a virgin) and the Greek means "virgin" (specifically). That's the prophecy. The New Testament, which records the events, uses "virgin" specifically in reference to Mary. There is no conflict - the OT word is broad enough to encompass virgin, even if it doesn't require it, so there is no need for convoluted explanations.
Second, the roots of the Immaculate Conception go back to Patristic times (200-600 AD or thereabout). To claim it arises in the 1700's is just gross ignorance.
Finally, the Immaculate Conception refers to the conception of Mary. It has nothing (directly) to do with the conception of Jesus. If you're going to write books attaching Christian teachings, at least take the trouble to learn what the hell you're talking about! Oh, but this is Slashdot, where the most idiotic crap will get modded "Insightful" as long as it bashes Christianity.
As for the quality of the students themselves, the students from the new curriculum are different. They are very fast (faster than me) at solving problems with known forms. On structured problems, similar to ones they have seen before, they are very fast. Unfortunately, they are very poor at solving unstructured problems, and problems where they have not seen the solution technique in advance.
I had a chance to teach university physics as a visiting professor at a "selective US college" and I can tell you exactly why this is. The problem is the students don't know the difference between memorizing and understanding any more. Standardized tests are, I suspect, a significant factor in this because in a multiple-choice test it's hard to test anything more than "Does the student know fact A and fact B?" and high schools have started teaching to the test (NCLB etc). Because my students had always learned that the way to get an "A" is to memorize the recipe for getting the answer the teacher wants to see, as opposed to learning and understanding why they're doing these different things, they could not apply what they've already learned to new situations. Furthermore, because they were cramming the night before for the exams (basically studying = memorize the textbook in their minds), they had poor retention of material and abysmal retention of material from past classes.
Let me give an example. Optics was part of what I taught, so they had seen both Snell's law (reflections and refractions) and thin-film interference problems (for example, anti-reflective coatings) in their homework. On the exam, I gave them a problem where I asked them to do both, one after the other, in order to design a mirror, and all but the one or two of the brightest students had not the slightest idea what to do. They had the formula for refraction (which some tried to use); they had the formula for interference (which others tried to use); but they didn't have one formula for doing both and they didn't know how to put the two formulas they did have together, because, while they're bright and motivated students, they've never been made to think before.
And don't even get me started on what they had [not] retained from the prerequisite Calculus course...
Bruno suggested that there could be an infinite number of worlds and that they could be inhabited by intelligent life.
Bruno may or may not have suggested this, but he was burned at the stake for his theological and occult views, not his scientific views
(see Wikipedia).
The Vatican is really taking their modernization seriously, aren't they?
The comments came from José Funes, the head of the Vatican Observatory, who is a Catholic priest and has a PhD in astronomy (and is a pretty cool guy, having met him last year).
Contrary to news headlines, however, he was not speaking on behalf of the Vatican; he's an employee, but not the Vatican's official spokesman. But why let the truth get in the way of a good headline...
-JS
PS. The Catholic church does not consider Hell a metaphor—Hell is real. The open (and probably unanswerable) question is whether Heaven and Hell are actually places or some kind of states of being.
Stable is relative. If by stable you mean "have half-lives on the order of one second" then yes, it's been expected for a long time. To claim a half-life of >100^8 years, though, strains credibility. All these super-heavy nuclei are presumably produced by supernovas. If an isotope has a lifetime on par with Uranium, then it will have a natural abundance on par with Uranium, which means we would have found it either chemically or spectroscopically decades ago. That we haven't found any naturally occurring elements past U238 means they all must have half-lives very short compared to geological time scales.
Name one plausible environmental damage scenario (other than full-out nuclear war) that would cause a significant proportion of human extinction.
Depends how you define significant, of course, but the consequences of running out of oil, including the corresponding loss of fertilizer and industrial infrastructure, would seem to fit the bill. Starvation of even 15% of the "surplus population" (to pull a WAG) is still significant in my book.
You're talking about average temperatures for most of history, which is MAYBE 2000 years.
Between tree rings, ice cores, and samples from lake and ocean beds, we have an excellent record going back a few tens of thousands of years and reasonable records going back over 100,000 years. And, as you so quickly point out, geologists have indirect indications of climate going back hundreds of millions of years (otherwise, they would just be talking out their ass, right?). Your 2000 years is just the written record.
All Stein is doing is asking scientists to act like it. They should acknowledge the weak spots in any theory and look to finding the explanations.
Scientists are acting like scientists. When experimental evidence is contrary to a theory, they label the theory false and move on.
ID is contradicted by experimental evidence (several things [eyes, some enzymes, etc] that, according to ID, are so complex they "could not" have evolved have been shown, through a combination of fossil and DNA studies to be, in fact, the result of evolutionary processes). ID makes basic predictions that have been proven to be false and therefore ID cannot be considered a meaningful explanation of anything.
What about those of us who love our jobs and love to excel in them, but don't want to make work our entire life?
That suggests to me that you've chosen a job that you don't *really* love, since you see a clean break between going to work as a necessary chore and returning home to enjoy life. That's not uncommon: it's called 9-to-5'ism, and it's the bane of company life because it creates shoddy, uncommitted workforces full of people whose main concern is leaving the office.
It's the bane of company life for three entirely different reasons:
1] 9-to-5'ers will quit and go work somewhere else if you treat them badly enough.
2] 9-to-5'ers won't let you exploit them by paying industry standard wages for 60-hour weeks. If a 9-to-5'er works 50% more hours than average, he/she actually expects to be paid 50% more than average.
3] 9-to-5'ers usually have an life outside of work. If your jobs is your life, then that gives your manager a lot of power over you. Managers like that. Employees shouldn't.
Despite what your students say, my experience has always been it's easier to debug Matlab code than compiled code, if you know anything about programming. Being interpreted and having good graphics routines built-in, I've found it especially handy for prototyping algorithms where the code changes quickly and most changes are one-offs.
It's got library routines for 99% of what I need.
It's fast enough for 85% of what I need.
Data types of float, int, and string are 99% of what I need (this is a research lab, not a software shop, after all). Object-oriented is just another place for the bugs to hide.
Why should I wait for someone to re-write all this in Python, and then put up with some else's buggy numerical code? Matlab is available, well-supported, actively devloped and maintained, and it lets me do the job I need to do faster than anything else.
It's really quite amazing how the programmer first has to express "add these two arrays" on a really low level, then the compiler needs to figure out what the programmer is really trying to do in order to generate code that's actually taking advantage of SSE primitives.
gcc -msse2 -mfpmath=sse *.c
It may be hard, but it can't be that hard because I've been using this flag for a good five years now...
Name a single real world problem that doesn't parallelize.
How about event-driven programming (or input-driven, same thing) where the interval between events is larger than the time to service an event? In other words, web browsers, email readers, word processors, PowerPoint...
Or, if you prefer scientific research, problems that map to Markov chains usually don't parallelize because of the strong temporal dependence between update steps (individual update steps are fast, but they're strongly correlated to the previous state in some non-deterministic fashion).
Get more memory. The more, the better. I'd rather have 8G of DDR2-533 than 2G of DDR3 uber-awesome overclocked OMGBBQ ram, because caching works.
OK, I'll admit I've managed to entirely avoid Vista
so far, but eight flipping gigs of memory?
My laptop (Mac, OSX/Leopard) with three days of uptime and the usual applications running is nice and responsive, and still never comes close to using all the 2GB installed. My previous Linux box ran acceptably on 384MB of RAM.
Modern day religions typically have a single benevolent deity...
There are
920 million Hindus would disagree with that statement, as would the 300 million Buddhists (give or take). Monotheism is unique to the Abrahamic religions; it is not an essential feature of "modern" religions.
You're dating a person, not an ideology. Who cares what (s)he believes, what matters is how you relate to him/her as a person. If you're going to let something as silly as astrology is going to get in the way of your relationships, you're in for a very lonely life.
As a matter of interest what would be the consequences to modern physics if Gravity waves do not exist?
Gravity waves are predicted by Einstein's theory of general relativity. Thus, if gravity waves do not exist, then general relativity is fundamentally flawed (hence the interest in testing this particular prediction of the theory).
(see also Easter, corresponding to the Summer Solstice and many cultures fertility holidays [ever wonder about easter eggs?]).
Easter corresponds to the Jewish feast of Passover, and has since the earliest days of Christianity. Even its name in most Romance languages (Pascha, etc) is a variation the Greek Hebrew words for Passover. The crazy formula we now use (Easter is the first Sunday after the first full moon after the spring equinox) was also devised as a way to approximately calculate the date of Passover (which always begins on 14 Nisan) and came into use as Christianity moved away from its Jewish roots (and calendar).
You also mean spring equinox (late March). The summer solstice is always in June. Thus, the holidays displaced would have been the New Year's festivals (on the old Roman calendar) and not pagan Midsummer's night festivals.
Slashdot Mirror
If it's not worth optimizing for parallel execution and the runtime is managable, then why the heck do you need a cluster? I can see if you're both lazy and impatient borrowing someone else's cluster for a few days, but if you actually have to build something, then why bother?
-JS
That's why God invented Fortran!
-JS
If this is the best bible scholarship your book has, then throw it out the window and forget everything you learned from it, because the author doesn't know crap.
First, yes the Hebrew means "young girl" (culturally synonymous with a virgin) and the Greek means "virgin" (specifically). That's the prophecy. The New Testament, which records the events, uses "virgin" specifically in reference to Mary. There is no conflict - the OT word is broad enough to encompass virgin, even if it doesn't require it, so there is no need for convoluted explanations.
Second, the roots of the Immaculate Conception go back to Patristic times (200-600 AD or thereabout). To claim it arises in the 1700's is just gross ignorance.
Finally, the Immaculate Conception refers to the conception of Mary. It has nothing (directly) to do with the conception of Jesus. If you're going to write books attaching Christian teachings, at least take the trouble to learn what the hell you're talking about! Oh, but this is Slashdot, where the most idiotic crap will get modded "Insightful" as long as it bashes Christianity.
-JS
I had a chance to teach university physics as a visiting professor at a "selective US college" and I can tell you exactly why this is. The problem is the students don't know the difference between memorizing and understanding any more. Standardized tests are, I suspect, a significant factor in this because in a multiple-choice test it's hard to test anything more than "Does the student know fact A and fact B?" and high schools have started teaching to the test (NCLB etc). Because my students had always learned that the way to get an "A" is to memorize the recipe for getting the answer the teacher wants to see, as opposed to learning and understanding why they're doing these different things, they could not apply what they've already learned to new situations. Furthermore, because they were cramming the night before for the exams (basically studying = memorize the textbook in their minds), they had poor retention of material and abysmal retention of material from past classes.
Let me give an example. Optics was part of what I taught, so they had seen both Snell's law (reflections and refractions) and thin-film interference problems (for example, anti-reflective coatings) in their homework. On the exam, I gave them a problem where I asked them to do both, one after the other, in order to design a mirror, and all but the one or two of the brightest students had not the slightest idea what to do. They had the formula for refraction (which some tried to use); they had the formula for interference (which others tried to use); but they didn't have one formula for doing both and they didn't know how to put the two formulas they did have together, because, while they're bright and motivated students, they've never been made to think before.
And don't even get me started on what they had [not] retained from the prerequisite Calculus course...
-JS
Last time I checked, the head of the Vatican Observatory was not the head of the Catholic church either.
-JS
Bruno may or may not have suggested this, but he was burned at the stake for his theological and occult views, not his scientific views (see Wikipedia).
-JS
The comments came from José Funes, the head of the Vatican Observatory, who is a Catholic priest and has a PhD in astronomy (and is a pretty cool guy, having met him last year).
Contrary to news headlines, however, he was not speaking on behalf of the Vatican; he's an employee, but not the Vatican's official spokesman. But why let the truth get in the way of a good headline...
-JS
PS. The Catholic church does not consider Hell a metaphor—Hell is real. The open (and probably unanswerable) question is whether Heaven and Hell are actually places or some kind of states of being.
Stable is relative. If by stable you mean "have half-lives on the order of one second" then yes, it's been expected for a long time. To claim a half-life of >100^8 years, though, strains credibility. All these super-heavy nuclei are presumably produced by supernovas. If an isotope has a lifetime on par with Uranium, then it will have a natural abundance on par with Uranium, which means we would have found it either chemically or spectroscopically decades ago. That we haven't found any naturally occurring elements past U238 means they all must have half-lives very short compared to geological time scales.
-JS
Both rational and informed about a religious topic? When did we start allowing this kind of nonsense on slashdot?
-JS
When you can take a bacterium and make anything without "natural selection," then I'll accept that ID might be a science.
-JS
Depends how you define significant, of course, but the consequences of running out of oil, including the corresponding loss of fertilizer and industrial infrastructure, would seem to fit the bill. Starvation of even 15% of the "surplus population" (to pull a WAG) is still significant in my book.
-JS
Between tree rings, ice cores, and samples from lake and ocean beds, we have an excellent record going back a few tens of thousands of years and reasonable records going back over 100,000 years. And, as you so quickly point out, geologists have indirect indications of climate going back hundreds of millions of years (otherwise, they would just be talking out their ass, right?). Your 2000 years is just the written record.
-JS
Ummm, from the counterfiter's point of view, isn't that sorta the point?
-JS
Scientists are acting like scientists. When experimental evidence is contrary to a theory, they label the theory false and move on.
ID is contradicted by experimental evidence (several things [eyes, some enzymes, etc] that, according to ID, are so complex they "could not" have evolved have been shown, through a combination of fossil and DNA studies to be, in fact, the result of evolutionary processes). ID makes basic predictions that have been proven to be false and therefore ID cannot be considered a meaningful explanation of anything.
-JS
It's the bane of company life for three entirely different reasons:
-JS
In space, no one can hear you scream.
-JS
Researchers (including myself) us Matlab because:
Why should I wait for someone to re-write all this in Python, and then put up with some else's buggy numerical code? Matlab is available, well-supported, actively devloped and maintained, and it lets me do the job I need to do faster than anything else.
-JS
gcc -msse2 -mfpmath=sse *.c
It may be hard, but it can't be that hard because I've been using this flag for a good five years now...
-JS
How about event-driven programming (or input-driven, same thing) where the interval between events is larger than the time to service an event? In other words, web browsers, email readers, word processors, PowerPoint...
Or, if you prefer scientific research, problems that map to Markov chains usually don't parallelize because of the strong temporal dependence between update steps (individual update steps are fast, but they're strongly correlated to the previous state in some non-deterministic fashion).
-JS
OK, I'll admit I've managed to entirely avoid Vista so far, but eight flipping gigs of memory?
My laptop (Mac, OSX/Leopard) with three days of uptime and the usual applications running is nice and responsive, and still never comes close to using all the 2GB installed. My previous Linux box ran acceptably on 384MB of RAM.
Eight !?!
-JS
There are 920 million Hindus would disagree with that statement, as would the 300 million Buddhists (give or take). Monotheism is unique to the Abrahamic religions; it is not an essential feature of "modern" religions.
-JS
You're dating a person, not an ideology. Who cares what (s)he believes, what matters is how you relate to him/her as a person. If you're going to let something as silly as astrology is going to get in the way of your relationships, you're in for a very lonely life.
-JS
Gravity waves are predicted by Einstein's theory of general relativity. Thus, if gravity waves do not exist, then general relativity is fundamentally flawed (hence the interest in testing this particular prediction of the theory).
-JS
Given the quality of most of what I read, forget Creators—I'll settle for being able to write in complete sentences.
-JS
Easter corresponds to the Jewish feast of Passover, and has since the earliest days of Christianity. Even its name in most Romance languages (Pascha, etc) is a variation the Greek Hebrew words for Passover. The crazy formula we now use (Easter is the first Sunday after the first full moon after the spring equinox) was also devised as a way to approximately calculate the date of Passover (which always begins on 14 Nisan) and came into use as Christianity moved away from its Jewish roots (and calendar).
You also mean spring equinox (late March). The summer solstice is always in June. Thus, the holidays displaced would have been the New Year's festivals (on the old Roman calendar) and not pagan Midsummer's night festivals.
-JS