I like this advice, but in the case at hand -- it's been two years! -- I doubt it's worth wasting any more time trying to get your issues resolved by that vendor. You'll have to eat some costs one way or another.
This is probably a textbook case for promoting free software. That has to be said. And since you'll need a replacement VPN solution, it's not just a pedantic argument.
I like that quote, even though it was a bit difficult to digest. The English language has evolved in the past century in a way that demands much less of the reader and conveys much less complexity and accuracy.
I wanted to add, somewhere, my $.02 about "faith." I'm told that early (1st century) Christians used what-we-translate-as-faith to be a kind of radical trust. More verb than noun. A trust in an idea, not fully understood or rationalized, that allowed them to lead lives that were unselfish, bold/foolhardy, non-violent extremists, anti-establishment, share-the-wealth sorts of people. The idea is that for them, faith was incompatible with certainty. Conviction deletes the possibility of faith. They did not have proof of deity, a consistent doctrine, etc. Reason was encouraged and appealed to, but knowledge was known to be incomplete.
What most people think about religion is that it is a doctrine (teaching or authority-based knowledge) that requires unwavering belief without question or reason. (My perspective here is Christianity rather than all religion, but I suspect that most major world religions are similar in this way.) Yet this is probably not a genuine or original form of any given religion but instead what human nature and politics have deformed religions into over time. People want to be told what to believe, and people who desire power cannot help but use fear and shame to great effect. I think modern-day Christianity is more about manipulating people and in most respects is the exact opposite of its earliest incarnations.
Science today has some of the same struggles. Science itself is an art, since the more precisely one tries to define it, the more inaccurate that definition becomes. Scientific knowledge is a little bit of an oxymoron since science can be described as a tool for disproving what is not true more than it is a means of proving what is true. This is true on all scales of complexity, but it's most evident at the reductionist frontier of particle physics and cosmology. The standard model is not logically consistent with general relativity, yet both theories are spectacularly successful. And there are problems of naturalness, etc. It is not tenable, not reasonable or scientific, to think that our most successful scientific theories are set to last. Modifications need to be made, and probably in big, fundamental, philosophically-challenging ways. The history of the development of physics is full of cases like this and physics is by no means "done." But people are eager to philosophize based on "what scientists know", and they are eager for answers from authority.
Authentic science, like authentic religion, is not authority-based. I'm not saying anything negative about consensus, just that there is always room for new theories and new experiments regardless of credentials. Data does not respect authority. And I don't believe there needs to be any contradiction between the two approaches of religion and science, as long as we are referring to religion as a searching process not a placating drug. Both science and religion address the basic problem of doing the best we can today with what little we know. Good scientists know that good questions are better than "right" answers, and good... what, "religious" folk ??? (atheists included)... know that it's better to be loving than right.
I suppose most of these ideas come from two books that might seem diametrically-opposed: The Underground Church, and Dreams of a Final Theory.
A program like the grandparent's "Hello World" is meant as a starting point and not a demonstration of how small a nearly-useless program can be. A GUI program necessarily aims to do more than just print a message, and this example gives you a small glimpse at how you the language could look and feel, and how you might go about doing something more practical. A MessageBox popup is not a good starting point since about all you can change is the text itself.
Do you have personal experience with this? Are there any data on that? How many lives are saved per year by the threat of gun violence?
In the absence of a study, imagine a world in which every citizen (maybe older than, say, the legal driving age) is carrying a firearm. Imagine the major population centers like NYC where the statistics would matter. Would there be fewer gun-related deaths in that world than in ours? I can't see it that way. I would feel safer in a world where people are more encouraged to deal with conflict in a nonviolent way.
Re: Mercury's precession, I'm still a believer in Vulcan.
Yeah, even the term "disproves" is not exactly correct. Newtonian gravity has a very hard time explaining Mercury's precession and is completely untenable with today's observational evidence. General relativity explains Mercury's orbit without having to invent new invisible planets & stuff. And today General relativity is still doing spectacularly well with many careful neutron star observations as well as experiments closer to home, like Gravity Probe B's measurements of frame dragging and more.
Funny, I read that book (which is excellent) but don't remember that analogy. But I think you're talking about special relativity, not general relativity. The best GR explanation I've seen is an article Lost in Hyperbolia. For me that explanation worked perfectly.
Now I remember reading in various places that the solar eclipse data on GR was not actually conclusive. Bad science. The earlier work Einstein did that explained the precession of mercury's orbit was actually the first confirmation of GR. Also, of course, confirmation is not a word that is ever used correctly in science. The precession of mercury's orbit disproved Newtonian gravity but failed to disprove GR. The bending of starlight by the Sun would have been an even more impressive failure-to-disprove GR if the data were actually conclusive.
I guess I need to read through the studies some more, but I believe that most/many indicators of education show that lecturing is poor (relative to other methods) regardless of the particular talents of the students. Yes, I have felt that I've gotten a lot out of certain lectures. Some lecturers are definitely better than others. Certainly both student and professor abilities can make a huge difference, but on top of that a professor can still do better by adopting a hybrid approach.
Lecture for a bit (less than 10 minutes, maybe 15 tops), then switch it up. Have students work a problem (maybe in pairs) and vote on a solution. With smallish classes you can use simple voting methods, and with 100+ groups you can use electronic voting. Have short, focused discussions. Do a demonstration that includes volunteers. Have students research related topics and present their own mini-lecture (small classes only). And so on. Even if the professor is a great lecturer and the students are all highly disciplined and auditory-sequential learners (almost never true for an entire class), they would *still* benefit more from this kind of approach. It certainly takes more work on the part of the professor and involves some retraining, but students are paying a lot of money for their education and they ought to be treated as valued customers a little bit more. I've tried these approaches myself in undergraduate physics classes of various levels from gen-ed courses to the introductory sequence and up to the jr/sr level. I won't claim to have mastered anything but it quickly became obvious to me that lecturing for an hour is never the best way to teach. There are alternatives and a conscientious professor owes it to their students to pay some heed to the learning sciences and experiment with different approaches. It's not a one-technique-fits-all kind of thing, but anyone can improve upon pure lecturing.
I heard of one professor that met a student wandering the halls at the end of the semester, looking for their professor. He asked him where this certain professor's office was -- they were preparing for the final exam and had some questions. He said "I'm your professor. Have you seen me before?" Obviously this student didn't feel that the lectures were an efficient way for them to learn.
But studies have been finding this for the past two decades.
My thoughts exactly. This is apparently a new study, however. It's not clear to me what is new about it other than, perhaps, translating the results into letter-grade equivalents. I like the quote: "it’s almost unethical to be lecturing if you have this data."
And yet, as you point out, this kind of data has been around for decades at least. I think they knew in the 80's if not earlier that knowledge retention is terrible for students listening to lectures compared to other methods (reading, group activities, teaching, etc). But how many professors took that data to heart? Is it a matter of couching it in different terms like letter grades? Probably not because those professors who lecture today either don't know or don't care. In either case they are immune to new studies like this.
One problem is that folks have to pay Comcast for decent internet service, and also they have to pay Netflix for a subscription. Fine of course, but if Netflix has to pony up extra fast-lane and direct-lane fees, ultimately their subscription prices increase. So Comcast+Netflix customers essentially get a hidden charge for their video streaming, one directly to Comcast and the other indirectly to Comcast (through Netflix). The real problem is that the indirect fee also applies to DSL and satellite customers, so you can't even avoid this fee by choosing a Comcast competitor.
I can understand wanting a free market system to avoid tragedy-of-the-commons types of issues with Netflix customers causing other non-streaming subscribers to get worse performance, but this present "solution" is clearly broken and gives Comcast and other last-mile providers a significant economic influence over other companies like Netflix that does not derive from consumer choice.
I think there are at least two underlying problems here. You point out one: corruption in the form of campaign finances and lobbyists. The other is the outrageous effectiveness of propaganda. I don't hear much about potential solutions to this second problem, and it's perhaps more difficult to solve. It might help to have some kind of propaganda analysis/deconstruction as part of a standard curriculum in high school, but that wouldn't be enough. We also need some watchdog-type media coverage that picks apart and shames people using such blatant propaganda techniques. Obviously TFA is just such a thing, as are programs like The Daily Show etc. These are good but not yet spread widely enough across the political spectrum.
Ideally a large fraction of voters need to be able to identify propaganda whenever they see it. And they need to react negatively to it, more than logic demands, overcompensating for the rest who get suckered by it all. Unfortunately, nearly all political campaigning is so densely infused with propaganda that we get accustomed to it. If all advertisers and all politicians do this all the time, there's no person more shamed than any other.
If one cannot be "redeemed," why even serve long sentences? I can only think of four possible goals or outcomes of a penal system. One is revenge, which seems entirely without merit (to put it mildly). The second is deterrent, to reduce crime by fear of punishment. The third is protection, to remove harmful elements from an otherwise-healthy society. The fourth is reformation/rehabilitation, so that individuals can once again participate in society in a positive way. Both of these latter goals get difficult to define when you have to get detailed or specific, but in a broad, general sense they seem reasonable to me.
I don't see the fourth goal being very effectively accomplished by penal systems today. The first two on this list are at odds with the fourth to a large degree. If you torture or drug prisoners to think they've served a 1000-year sentence, there is probably no way for them to return to society after having been driven mad with isolation, boredom, and anger. Capital punishment would better serve the first three goals unless, like the article's author, you feel that death is too good. If redemption is ever not a realistic possibility, I don't see a better alternative than death.
On the other hand, death is irreversible and courts do not always correctly determine guilt. A disturbingly-large large fraction of death-row inmates have been proven innocent.
Just to counterbalance this, I recently switched from doing academic physics research to a programming career. Now I'm a senior software developer with a wide variety of experiences (currently Android internals). I've taken exactly one comp sci course: data structures. That course was helpful but habitually writing, reading, and debugging code has taught me most everything I've needed. Reading code is very important, as is starting programming at an early age and applying it to whatever you're doing. Learning a variety of languages is also important as each one brings its own paradigms. I read a patterns textbook once but didn't see anything that I wasn't already familiar with. I just learned terminology from that.
Still, if one is new to programming I'd have to think that getting a degree one way or another would teach a lot in a short period of time. I had never really thought of programming as a profession but rather as a tool that was necessary to whatever else I wanted to do. Economic pressures turned my hobby into a profession. I'm still just a bit grumpy about that even though I'm having fun.
I would naively have expected that cuts to NIH programs would have low impact relative to programs funded by NASA and DoE. I don't have much knowledge of the NIH programs, but I feel like they are shorter term studies with less specialized infrastructure involved. So a 5% cut in funding there has closer to a 5% reduction in output. But DoE and NASA programs are often 10 to 20-year projects. If you cut funding one year (and they have cut funding 5-10% every year for a long time), you start cancelling programs and losing the money already spent. An oversimplified model would be that if you fund one of these agencies for, say, $5B for 8 years, $2B the 9th year, and then $8B the 10th year you'd get $20B worth of "output" not the $50B you've spent. In year 9 programs get cancelled, equipment mothballed, and people leave the field.
I agree with most of this, but I still use emacs a lot.
Sometimes an IDE can become an impediment if it keeps you from looking at hard-to-find code, makes you in any way lazy about debugging, or helps you read through code too quickly (making happy assumptions). I'm currently working on an Android system and regularly have to jump between application code, Android internal services and components, various JNI or hardware abstraction layers, and kernel code. Sometimes I need to read through standard C library implementations too. So while I like doing some things in eclipse, most of the time eclipse is not configured to let me dig deeply enough or follow calls all the way through the framework. Similarly the interactive debugging utilities are often limited and I resort to reading code and adding print/logging statements. It feels old-fashioned sometimes but on the other hand I have no excuses or reluctance to work hard enough to get a deep understanding of the whole system.
So I'd judge an IDE by its ability to help you quickly navigate around a large and maybe unfamiliar code base. A good programmer's #1 job is probably reading code carefully and accurately. Features like auto-completion and live error checking are much less important. Also keep a tab on the amount of time it takes to keep your IDE configured correctly and able to compile independently from a more automated build system. Don't let it become a timesink, crutch or an excuse!
I read "...and improve their understanding of math concepts" with a lot of skepticism. I think that schools love to teach computation skills because they are easy to teach and because success there is very easy to measure. But this skill is relatively unimportant compared with what I would consider "math concepts": How you apply mathematical abstractions to real-world situations (beyond making correct change at a cash register). How you break down a hard problem into less-hard pieces. How to visualize quantitative relationships, develop and use algebraic systems, and so on. These are rarely taught in schools because they are relatively difficult to teach and difficult to measure gains. So computation skills are taught instead, regardless of the fact that cheap computers are billions or trillions of times faster than any human.
Can electric current apply to this kind of conceptual learning? If so, it would have application to nearly all kinds of education, not just math.
I think one of the major problems with standardized testing is that the scores are over-used. Scores are used for evaluating students, teachers, and schools. Each group has incentives attached: scholarships and admissions for students, salaries and job opportunities for teachers, and higher funding levels for high-performing schools. If these incentive systems were decoupled instead somehow, then meaningful analyses of scores could be made.
I agree that assessment is important, but it's a very difficult problem that we have not solved. Is it better to use a broken approach than have no assessment at all? I personally think that testimonials and self-evaluations are currently more meaningful than test scores in general. Standardized tests usually measure only those outcomes that are easy to teach and test but are unimportant in real life.
Mathematics is an easy example of this. My 6th-grade daughter gets homework sheets that are mostly repetitive exercises to boost speed. A recent example was computing the circumferences and surface areas of circles. Calculators were allowed in this case, so it was an exercise in key punching. Then we went to a pizzeria a few days later and she was happy to compute the circumference of a pizza (without prompting -- her idea). So I asked her "about how many bites would it take to eat a whole pizza?" She didn't know how to work that out, thinking that she was not taught how to solve that kind of question. I wasn't giving her one of the needed inputs, the size of a bite. And also her understanding of surface area was not strong. The problem here is fairly universal: Math students are taught the abstract computation skills but not synthesis of this knowledge. It is very easy to test how quickly students can compute things and even how fast they can look up formulas that match the given knowns with the required unknown, but these are meaningless skills without also learning how to bridge from the abstract to the real, and how to form connections between the principles they learn. Students should be given harder problems that they have to puzzle over, even starting in grade school. It's okay if they don't get the right answer or if they get stumped, as long as they were able to begin to think productively about them. This builds genuinely-useful math skills and hopefully a certain discipline of thought that is not usually innate. This ability to apply math to real-world situations, and by algebra to think of math as a language, is much harder to assess than timed multiplication tests and so it is generally avoided. If a teacher were to spend time on this kind of development in place of the repetitive speed drills, their students would perform worse on standardized tests. The teacher's students would suffer, as would the teacher and their whole school to some degree.
So, is standardized testing better than nothing? I think that yes, it is, but on the other hand almost any other form of assessment would be better still even if it were not standardized.
Mathematics, especially in this context, is just a language for expressing ideas. So I think that in some sense it is possible that some particular string theory really does describe what's going on. I don't know if it's very likely, but the idea of a "final theory" is that it is, in some sense, a complete and accurate description of our universe's mechanics. (The holographic principle is nice because it states that two seemingly very different theories can actually be equivalent.)
I think there are two main caveats here. One is that you never really know when you're "done", and have a theory that is indeed final. We know now that we are not done today because of inconsistencies, but science also does not have even the capability of perfect validation. The other is that a microscopic, reductionist description of physics is not useful or even the correct language for describing more macroscopic effects since basically "more is different." Chemists don't use quantum field theory because it's just not helpful.
And while it's great (even important) to consider philosophical ramifications of theoretical work like this, we have to remember that it's all still conjecture and it will probably always be conjecture. The philosophical spin-offs, so to speak, should never be taken as a way of either supporting or condemning the theory.
I think I'm basically summarizing some of what Wienberg describes in his book Dreams of a Final Theory (1993). That seems old but I highly recommend it!
That depends on what you mean by "tangible benefits." One argument I've heard for practical, what's-in-it-for-me-today benefits is that the technology produces spin-offs such as techniques to mass-produce rare-earth magnets, the world-wide web, etc. But that's honestly a weak argument because there's a lot of research going on that has similar chances to produce spin-off tech.
For particle physics, the feeling is that we are on the verge of some kind of revolution! Admittedly it's been that way for a few decades now, but the current working theory (the standard model) has a number of deep problems (thanks wikipedia!). Most new theories, and there are a whole lot of them, predict new phenomena just at the edge of our experimental reach. Part of that is because well-meaning theorists prefer to propose theories that are either presently or soon-to-be testable. But part of it is because the experimental frontier has advanced to energies at roughly the electroweak unification point and lots of theories have interesting behavior to predict at this point, broadly speaking.
So it's not just a more-is-better kind of effort that won't stop until we build solar-system-sized accelerators. There really is a sense that a major shift, possibly even a philosophically-challenging development, is nearly within our grasp, within our lifetimes. This is not a "practical" argument for basic science, but only history can tell us what has had short and long-term practical benefits. History does tell us that this sort of pursuit has in the past been enormously beneficial. Maybe we are in a whole new era where new physics will be completely impractical, but that would honestly be surprising if true.
I don't think in these cases you have multiple labs bidding for the job. You have multiple countries wanting to host the lab, but that's a different story.
The biggest problem for high energy physics is establishing multi-year funding. The US government cannot promise anything beyond a single year of funding. If say $8 B has been spent over 10 years and one year congress says "but I promised to cut spending", then that's the end of the road for that lab. This happened for the SSC in 1993, but also a lot of times since then on lower-profile, some $500 M experiments that were, yes, in construction already.
Now say 15 years later the $10B has been spent, but its not quite done, another $2B would let you finish the project. Do you really throw away $10B to save 2B? There is no fraud, just a mis-estimation of the costs of building a beyond state-of-the-art machine and slightly larger technical problems than were expected.
Most of the cases I'm familiar with, including the SSC, were not actually budget overruns even though they were politicized that way. If you're a politician who wants to (a) publicly demonstrate how fiscally conservative you are and (b) not actually cut spending on items that might affect the bulk of your constituency, then you cut big science every time. Even if the budget grows on the whole, you've made a statement and some headlines.
I also wanted to mention the failed SSC in Texas, cancelled in 1993. That would have been running at double the LHC's energy about a decade earlier. In 1993 congress seats were won by senators promising budget cuts, and Big Science had a large target painted on its back. Killing the SSC was a big-profile way of appearing to reduce spending while at the same time not damaging something that many people understood or cared about.
Since that time, the US has proved time and time again that they are incapable of sustaining funding for a long-term science project. All of the high-energy accelerators in the US are operationally shut down, and almost no proposals in the past 20 years or so have survived all the way to producing results before getting scrapped by some budget shortfall in a particular fiscal year. The LHC survives because the US is not such a major (or critical) contributor.
Well, many of these tunnels, including the one the LHC uses, have been refurbished multiple times already. Cern's main ring was built to be somewhat future-proof, but that was a long time ago. A google search came up with The history of CERN, which dates the groundbreaking to 1954.
In accelerators you have two basic designs: linear and circular(ish). In linear accelerators each boosting element (RF cavity or whatnot) gets one chance to give the beam particles a kick, so the energy is limited to how hard you kick (limited by technology) and how many elements / how long (limited by budget).
In circular accelerators you are limited by synchrotron radiation. At some point the energy pumped into the beam matches the energy lost via synchrotron radiation. To move in a circle you have to accelerate inwardly, and an accelerating charged particle radiates light. At particle accelerator energies, this radiation is in the x-ray spectrum. You can reduce the loss by using a larger ring -- a smaller curvature requires less centripetal acceleration and hence less radiation loss. You can also of course build stronger boosting elements, but the radiation also heats the beamline and surrounding superconducting magnets, so it's not "that simple."
The other thing to vary is the kind of particle accelerated. Electrons have a very small mass and lose a larger fraction of their momentum to synchrotron radiation. SLAC and KEK are linear accelerators that use electrons. (Cornell's CESR is a ring that accelerates electrons too, but at lower energies compared to these others.) Protons are the other obvious choice, which is what Fermilab and CERN's LHC (after the upgrade) are accelerating. Being much more massive, the protons slough off less of their momentum to synchrotron radiation and can be accelerated to higher energies given the same size ring. The disadvantage of protons is that the energy of the proton is shared among its three quarks (and gluons I think) whereas the electron is truly singular as far as can be told.
I've been out of touch lately but as of at least 8 years ago three proposals were being discussed: VLHC -- big ring accelerating protons. Next Linear Collider (NLC) -- long linear accelerator for electrons. Muon collider -- a smaller ring (actually with straight sections like a track&field track) that produces and accelerates muons. Muons are just like electrons only 200 times more massive and is unstable with a half-life of 2 microseconds. The muon collider was thought to be an ideal Higgs factory, but with a lot of design challenges. One of the main challenges is to not only accelerate the muons before they decay, but also collimate, or "cool", the beam very fast as well so that you can create as many head-on collisions as possible.
So the news that the VLHC design is currently in favor is interesting, but this is hardly the first time the issue has been discussed and I doubt it will be the last. Several years ago the NLC design seemed most favorable, but this would, by its length, be limited to a specific design energy and probably be built to produce Higgs, Higgs, and more Higgs. It seems to me like a VLHC would have more discovery potential for more massive Higgs particles, signs of supersymmetry, or whatever else might exist.
Slashdot Mirror
I like this advice, but in the case at hand -- it's been two years! -- I doubt it's worth wasting any more time trying to get your issues resolved by that vendor. You'll have to eat some costs one way or another.
This is probably a textbook case for promoting free software. That has to be said. And since you'll need a replacement VPN solution, it's not just a pedantic argument.
I like that quote, even though it was a bit difficult to digest. The English language has evolved in the past century in a way that demands much less of the reader and conveys much less complexity and accuracy.
I wanted to add, somewhere, my $.02 about "faith." I'm told that early (1st century) Christians used what-we-translate-as-faith to be a kind of radical trust. More verb than noun. A trust in an idea, not fully understood or rationalized, that allowed them to lead lives that were unselfish, bold/foolhardy, non-violent extremists, anti-establishment, share-the-wealth sorts of people. The idea is that for them, faith was incompatible with certainty. Conviction deletes the possibility of faith. They did not have proof of deity, a consistent doctrine, etc. Reason was encouraged and appealed to, but knowledge was known to be incomplete.
What most people think about religion is that it is a doctrine (teaching or authority-based knowledge) that requires unwavering belief without question or reason. (My perspective here is Christianity rather than all religion, but I suspect that most major world religions are similar in this way.) Yet this is probably not a genuine or original form of any given religion but instead what human nature and politics have deformed religions into over time. People want to be told what to believe, and people who desire power cannot help but use fear and shame to great effect. I think modern-day Christianity is more about manipulating people and in most respects is the exact opposite of its earliest incarnations.
Science today has some of the same struggles. Science itself is an art, since the more precisely one tries to define it, the more inaccurate that definition becomes. Scientific knowledge is a little bit of an oxymoron since science can be described as a tool for disproving what is not true more than it is a means of proving what is true. This is true on all scales of complexity, but it's most evident at the reductionist frontier of particle physics and cosmology. The standard model is not logically consistent with general relativity, yet both theories are spectacularly successful. And there are problems of naturalness, etc. It is not tenable, not reasonable or scientific, to think that our most successful scientific theories are set to last. Modifications need to be made, and probably in big, fundamental, philosophically-challenging ways. The history of the development of physics is full of cases like this and physics is by no means "done." But people are eager to philosophize based on "what scientists know", and they are eager for answers from authority.
Authentic science, like authentic religion, is not authority-based. I'm not saying anything negative about consensus, just that there is always room for new theories and new experiments regardless of credentials. Data does not respect authority. And I don't believe there needs to be any contradiction between the two approaches of religion and science, as long as we are referring to religion as a searching process not a placating drug. Both science and religion address the basic problem of doing the best we can today with what little we know. Good scientists know that good questions are better than "right" answers, and good ... what, "religious" folk ??? (atheists included) ... know that it's better to be loving than right.
I suppose most of these ideas come from two books that might seem diametrically-opposed: The Underground Church, and Dreams of a Final Theory.
A program like the grandparent's "Hello World" is meant as a starting point and not a demonstration of how small a nearly-useless program can be. A GUI program necessarily aims to do more than just print a message, and this example gives you a small glimpse at how you the language could look and feel, and how you might go about doing something more practical. A MessageBox popup is not a good starting point since about all you can change is the text itself.
But ... We ASKED him!
Do you have personal experience with this? Are there any data on that? How many lives are saved per year by the threat of gun violence?
In the absence of a study, imagine a world in which every citizen (maybe older than, say, the legal driving age) is carrying a firearm. Imagine the major population centers like NYC where the statistics would matter. Would there be fewer gun-related deaths in that world than in ours? I can't see it that way. I would feel safer in a world where people are more encouraged to deal with conflict in a nonviolent way.
Re: Mercury's precession, I'm still a believer in Vulcan.
Yeah, even the term "disproves" is not exactly correct. Newtonian gravity has a very hard time explaining Mercury's precession and is completely untenable with today's observational evidence. General relativity explains Mercury's orbit without having to invent new invisible planets & stuff. And today General relativity is still doing spectacularly well with many careful neutron star observations as well as experiments closer to home, like Gravity Probe B's measurements of frame dragging and more.
Oh, what I do remember from Bairn Greene's The Elegant Universe was his analogy for Bell's Inequality. Looks like that has been put up on Wikipedia.
Funny, I read that book (which is excellent) but don't remember that analogy. But I think you're talking about special relativity, not general relativity. The best GR explanation I've seen is an article Lost in Hyperbolia. For me that explanation worked perfectly.
Now I remember reading in various places that the solar eclipse data on GR was not actually conclusive. Bad science. The earlier work Einstein did that explained the precession of mercury's orbit was actually the first confirmation of GR. Also, of course, confirmation is not a word that is ever used correctly in science. The precession of mercury's orbit disproved Newtonian gravity but failed to disprove GR. The bending of starlight by the Sun would have been an even more impressive failure-to-disprove GR if the data were actually conclusive.
I guess I need to read through the studies some more, but I believe that most/many indicators of education show that lecturing is poor (relative to other methods) regardless of the particular talents of the students. Yes, I have felt that I've gotten a lot out of certain lectures. Some lecturers are definitely better than others. Certainly both student and professor abilities can make a huge difference, but on top of that a professor can still do better by adopting a hybrid approach.
Lecture for a bit (less than 10 minutes, maybe 15 tops), then switch it up. Have students work a problem (maybe in pairs) and vote on a solution. With smallish classes you can use simple voting methods, and with 100+ groups you can use electronic voting. Have short, focused discussions. Do a demonstration that includes volunteers. Have students research related topics and present their own mini-lecture (small classes only). And so on. Even if the professor is a great lecturer and the students are all highly disciplined and auditory-sequential learners (almost never true for an entire class), they would *still* benefit more from this kind of approach. It certainly takes more work on the part of the professor and involves some retraining, but students are paying a lot of money for their education and they ought to be treated as valued customers a little bit more. I've tried these approaches myself in undergraduate physics classes of various levels from gen-ed courses to the introductory sequence and up to the jr/sr level. I won't claim to have mastered anything but it quickly became obvious to me that lecturing for an hour is never the best way to teach. There are alternatives and a conscientious professor owes it to their students to pay some heed to the learning sciences and experiment with different approaches. It's not a one-technique-fits-all kind of thing, but anyone can improve upon pure lecturing.
I heard of one professor that met a student wandering the halls at the end of the semester, looking for their professor. He asked him where this certain professor's office was -- they were preparing for the final exam and had some questions. He said "I'm your professor. Have you seen me before?" Obviously this student didn't feel that the lectures were an efficient way for them to learn.
But studies have been finding this for the past two decades.
My thoughts exactly. This is apparently a new study, however. It's not clear to me what is new about it other than, perhaps, translating the results into letter-grade equivalents. I like the quote: "it’s almost unethical to be lecturing if you have this data."
And yet, as you point out, this kind of data has been around for decades at least. I think they knew in the 80's if not earlier that knowledge retention is terrible for students listening to lectures compared to other methods (reading, group activities, teaching, etc). But how many professors took that data to heart? Is it a matter of couching it in different terms like letter grades? Probably not because those professors who lecture today either don't know or don't care. In either case they are immune to new studies like this.
One problem is that folks have to pay Comcast for decent internet service, and also they have to pay Netflix for a subscription. Fine of course, but if Netflix has to pony up extra fast-lane and direct-lane fees, ultimately their subscription prices increase. So Comcast+Netflix customers essentially get a hidden charge for their video streaming, one directly to Comcast and the other indirectly to Comcast (through Netflix). The real problem is that the indirect fee also applies to DSL and satellite customers, so you can't even avoid this fee by choosing a Comcast competitor.
I can understand wanting a free market system to avoid tragedy-of-the-commons types of issues with Netflix customers causing other non-streaming subscribers to get worse performance, but this present "solution" is clearly broken and gives Comcast and other last-mile providers a significant economic influence over other companies like Netflix that does not derive from consumer choice.
I think there are at least two underlying problems here. You point out one: corruption in the form of campaign finances and lobbyists. The other is the outrageous effectiveness of propaganda. I don't hear much about potential solutions to this second problem, and it's perhaps more difficult to solve. It might help to have some kind of propaganda analysis/deconstruction as part of a standard curriculum in high school, but that wouldn't be enough. We also need some watchdog-type media coverage that picks apart and shames people using such blatant propaganda techniques. Obviously TFA is just such a thing, as are programs like The Daily Show etc. These are good but not yet spread widely enough across the political spectrum.
Ideally a large fraction of voters need to be able to identify propaganda whenever they see it. And they need to react negatively to it, more than logic demands, overcompensating for the rest who get suckered by it all. Unfortunately, nearly all political campaigning is so densely infused with propaganda that we get accustomed to it. If all advertisers and all politicians do this all the time, there's no person more shamed than any other.
If one cannot be "redeemed," why even serve long sentences? I can only think of four possible goals or outcomes of a penal system. One is revenge, which seems entirely without merit (to put it mildly). The second is deterrent, to reduce crime by fear of punishment. The third is protection, to remove harmful elements from an otherwise-healthy society. The fourth is reformation/rehabilitation, so that individuals can once again participate in society in a positive way. Both of these latter goals get difficult to define when you have to get detailed or specific, but in a broad, general sense they seem reasonable to me.
I don't see the fourth goal being very effectively accomplished by penal systems today. The first two on this list are at odds with the fourth to a large degree. If you torture or drug prisoners to think they've served a 1000-year sentence, there is probably no way for them to return to society after having been driven mad with isolation, boredom, and anger. Capital punishment would better serve the first three goals unless, like the article's author, you feel that death is too good. If redemption is ever not a realistic possibility, I don't see a better alternative than death.
On the other hand, death is irreversible and courts do not always correctly determine guilt. A disturbingly-large large fraction of death-row inmates have been proven innocent.
Just to counterbalance this, I recently switched from doing academic physics research to a programming career. Now I'm a senior software developer with a wide variety of experiences (currently Android internals). I've taken exactly one comp sci course: data structures. That course was helpful but habitually writing, reading, and debugging code has taught me most everything I've needed. Reading code is very important, as is starting programming at an early age and applying it to whatever you're doing. Learning a variety of languages is also important as each one brings its own paradigms. I read a patterns textbook once but didn't see anything that I wasn't already familiar with. I just learned terminology from that.
Still, if one is new to programming I'd have to think that getting a degree one way or another would teach a lot in a short period of time. I had never really thought of programming as a profession but rather as a tool that was necessary to whatever else I wanted to do. Economic pressures turned my hobby into a profession. I'm still just a bit grumpy about that even though I'm having fun.
I would naively have expected that cuts to NIH programs would have low impact relative to programs funded by NASA and DoE. I don't have much knowledge of the NIH programs, but I feel like they are shorter term studies with less specialized infrastructure involved. So a 5% cut in funding there has closer to a 5% reduction in output. But DoE and NASA programs are often 10 to 20-year projects. If you cut funding one year (and they have cut funding 5-10% every year for a long time), you start cancelling programs and losing the money already spent. An oversimplified model would be that if you fund one of these agencies for, say, $5B for 8 years, $2B the 9th year, and then $8B the 10th year you'd get $20B worth of "output" not the $50B you've spent. In year 9 programs get cancelled, equipment mothballed, and people leave the field.
I agree with most of this, but I still use emacs a lot.
Sometimes an IDE can become an impediment if it keeps you from looking at hard-to-find code, makes you in any way lazy about debugging, or helps you read through code too quickly (making happy assumptions). I'm currently working on an Android system and regularly have to jump between application code, Android internal services and components, various JNI or hardware abstraction layers, and kernel code. Sometimes I need to read through standard C library implementations too. So while I like doing some things in eclipse, most of the time eclipse is not configured to let me dig deeply enough or follow calls all the way through the framework. Similarly the interactive debugging utilities are often limited and I resort to reading code and adding print/logging statements. It feels old-fashioned sometimes but on the other hand I have no excuses or reluctance to work hard enough to get a deep understanding of the whole system.
So I'd judge an IDE by its ability to help you quickly navigate around a large and maybe unfamiliar code base. A good programmer's #1 job is probably reading code carefully and accurately. Features like auto-completion and live error checking are much less important. Also keep a tab on the amount of time it takes to keep your IDE configured correctly and able to compile independently from a more automated build system. Don't let it become a timesink, crutch or an excuse!
I read "...and improve their understanding of math concepts" with a lot of skepticism. I think that schools love to teach computation skills because they are easy to teach and because success there is very easy to measure. But this skill is relatively unimportant compared with what I would consider "math concepts": How you apply mathematical abstractions to real-world situations (beyond making correct change at a cash register). How you break down a hard problem into less-hard pieces. How to visualize quantitative relationships, develop and use algebraic systems, and so on. These are rarely taught in schools because they are relatively difficult to teach and difficult to measure gains. So computation skills are taught instead, regardless of the fact that cheap computers are billions or trillions of times faster than any human.
Can electric current apply to this kind of conceptual learning? If so, it would have application to nearly all kinds of education, not just math.
I think one of the major problems with standardized testing is that the scores are over-used. Scores are used for evaluating students, teachers, and schools. Each group has incentives attached: scholarships and admissions for students, salaries and job opportunities for teachers, and higher funding levels for high-performing schools. If these incentive systems were decoupled instead somehow, then meaningful analyses of scores could be made.
I agree that assessment is important, but it's a very difficult problem that we have not solved. Is it better to use a broken approach than have no assessment at all? I personally think that testimonials and self-evaluations are currently more meaningful than test scores in general. Standardized tests usually measure only those outcomes that are easy to teach and test but are unimportant in real life.
Mathematics is an easy example of this. My 6th-grade daughter gets homework sheets that are mostly repetitive exercises to boost speed. A recent example was computing the circumferences and surface areas of circles. Calculators were allowed in this case, so it was an exercise in key punching. Then we went to a pizzeria a few days later and she was happy to compute the circumference of a pizza (without prompting -- her idea). So I asked her "about how many bites would it take to eat a whole pizza?" She didn't know how to work that out, thinking that she was not taught how to solve that kind of question. I wasn't giving her one of the needed inputs, the size of a bite. And also her understanding of surface area was not strong. The problem here is fairly universal: Math students are taught the abstract computation skills but not synthesis of this knowledge. It is very easy to test how quickly students can compute things and even how fast they can look up formulas that match the given knowns with the required unknown, but these are meaningless skills without also learning how to bridge from the abstract to the real, and how to form connections between the principles they learn. Students should be given harder problems that they have to puzzle over, even starting in grade school. It's okay if they don't get the right answer or if they get stumped, as long as they were able to begin to think productively about them. This builds genuinely-useful math skills and hopefully a certain discipline of thought that is not usually innate. This ability to apply math to real-world situations, and by algebra to think of math as a language, is much harder to assess than timed multiplication tests and so it is generally avoided. If a teacher were to spend time on this kind of development in place of the repetitive speed drills, their students would perform worse on standardized tests. The teacher's students would suffer, as would the teacher and their whole school to some degree.
So, is standardized testing better than nothing? I think that yes, it is, but on the other hand almost any other form of assessment would be better still even if it were not standardized.
Mathematics, especially in this context, is just a language for expressing ideas. So I think that in some sense it is possible that some particular string theory really does describe what's going on. I don't know if it's very likely, but the idea of a "final theory" is that it is, in some sense, a complete and accurate description of our universe's mechanics. (The holographic principle is nice because it states that two seemingly very different theories can actually be equivalent.)
I think there are two main caveats here. One is that you never really know when you're "done", and have a theory that is indeed final. We know now that we are not done today because of inconsistencies, but science also does not have even the capability of perfect validation. The other is that a microscopic, reductionist description of physics is not useful or even the correct language for describing more macroscopic effects since basically "more is different." Chemists don't use quantum field theory because it's just not helpful.
And while it's great (even important) to consider philosophical ramifications of theoretical work like this, we have to remember that it's all still conjecture and it will probably always be conjecture. The philosophical spin-offs, so to speak, should never be taken as a way of either supporting or condemning the theory.
I think I'm basically summarizing some of what Wienberg describes in his book Dreams of a Final Theory (1993). That seems old but I highly recommend it!
That depends on what you mean by "tangible benefits." One argument I've heard for practical, what's-in-it-for-me-today benefits is that the technology produces spin-offs such as techniques to mass-produce rare-earth magnets, the world-wide web, etc. But that's honestly a weak argument because there's a lot of research going on that has similar chances to produce spin-off tech.
For particle physics, the feeling is that we are on the verge of some kind of revolution! Admittedly it's been that way for a few decades now, but the current working theory (the standard model) has a number of deep problems (thanks wikipedia!). Most new theories, and there are a whole lot of them, predict new phenomena just at the edge of our experimental reach. Part of that is because well-meaning theorists prefer to propose theories that are either presently or soon-to-be testable. But part of it is because the experimental frontier has advanced to energies at roughly the electroweak unification point and lots of theories have interesting behavior to predict at this point, broadly speaking.
So it's not just a more-is-better kind of effort that won't stop until we build solar-system-sized accelerators. There really is a sense that a major shift, possibly even a philosophically-challenging development, is nearly within our grasp, within our lifetimes. This is not a "practical" argument for basic science, but only history can tell us what has had short and long-term practical benefits. History does tell us that this sort of pursuit has in the past been enormously beneficial. Maybe we are in a whole new era where new physics will be completely impractical, but that would honestly be surprising if true.
You hit numerical problems if you calculate it that way. Wikipedia gives a series expansion that works well for large values of gamma:
v (in units of c) = 1 - 1/2 \gamma^(-2)
v = c (1 - 1.8e-10), or 0.99999999982 c
I don't think in these cases you have multiple labs bidding for the job. You have multiple countries wanting to host the lab, but that's a different story.
The biggest problem for high energy physics is establishing multi-year funding. The US government cannot promise anything beyond a single year of funding. If say $8 B has been spent over 10 years and one year congress says "but I promised to cut spending", then that's the end of the road for that lab. This happened for the SSC in 1993, but also a lot of times since then on lower-profile, some $500 M experiments that were, yes, in construction already.
Now say 15 years later the $10B has been spent, but its not quite done, another $2B would let you finish the project. Do you really throw away $10B to save 2B? There is no fraud, just a mis-estimation of the costs of building a beyond state-of-the-art machine and slightly larger technical problems than were expected.
Most of the cases I'm familiar with, including the SSC, were not actually budget overruns even though they were politicized that way. If you're a politician who wants to (a) publicly demonstrate how fiscally conservative you are and (b) not actually cut spending on items that might affect the bulk of your constituency, then you cut big science every time. Even if the budget grows on the whole, you've made a statement and some headlines.
I also wanted to mention the failed SSC in Texas, cancelled in 1993. That would have been running at double the LHC's energy about a decade earlier. In 1993 congress seats were won by senators promising budget cuts, and Big Science had a large target painted on its back. Killing the SSC was a big-profile way of appearing to reduce spending while at the same time not damaging something that many people understood or cared about.
Since that time, the US has proved time and time again that they are incapable of sustaining funding for a long-term science project. All of the high-energy accelerators in the US are operationally shut down, and almost no proposals in the past 20 years or so have survived all the way to producing results before getting scrapped by some budget shortfall in a particular fiscal year. The LHC survives because the US is not such a major (or critical) contributor.
Well, many of these tunnels, including the one the LHC uses, have been refurbished multiple times already. Cern's main ring was built to be somewhat future-proof, but that was a long time ago. A google search came up with The history of CERN, which dates the groundbreaking to 1954.
In accelerators you have two basic designs: linear and circular(ish). In linear accelerators each boosting element (RF cavity or whatnot) gets one chance to give the beam particles a kick, so the energy is limited to how hard you kick (limited by technology) and how many elements / how long (limited by budget).
In circular accelerators you are limited by synchrotron radiation. At some point the energy pumped into the beam matches the energy lost via synchrotron radiation. To move in a circle you have to accelerate inwardly, and an accelerating charged particle radiates light. At particle accelerator energies, this radiation is in the x-ray spectrum. You can reduce the loss by using a larger ring -- a smaller curvature requires less centripetal acceleration and hence less radiation loss. You can also of course build stronger boosting elements, but the radiation also heats the beamline and surrounding superconducting magnets, so it's not "that simple."
The other thing to vary is the kind of particle accelerated. Electrons have a very small mass and lose a larger fraction of their momentum to synchrotron radiation. SLAC and KEK are linear accelerators that use electrons. (Cornell's CESR is a ring that accelerates electrons too, but at lower energies compared to these others.) Protons are the other obvious choice, which is what Fermilab and CERN's LHC (after the upgrade) are accelerating. Being much more massive, the protons slough off less of their momentum to synchrotron radiation and can be accelerated to higher energies given the same size ring. The disadvantage of protons is that the energy of the proton is shared among its three quarks (and gluons I think) whereas the electron is truly singular as far as can be told.
I've been out of touch lately but as of at least 8 years ago three proposals were being discussed: VLHC -- big ring accelerating protons. Next Linear Collider (NLC) -- long linear accelerator for electrons. Muon collider -- a smaller ring (actually with straight sections like a track&field track) that produces and accelerates muons. Muons are just like electrons only 200 times more massive and is unstable with a half-life of 2 microseconds. The muon collider was thought to be an ideal Higgs factory, but with a lot of design challenges. One of the main challenges is to not only accelerate the muons before they decay, but also collimate, or "cool", the beam very fast as well so that you can create as many head-on collisions as possible.
So the news that the VLHC design is currently in favor is interesting, but this is hardly the first time the issue has been discussed and I doubt it will be the last. Several years ago the NLC design seemed most favorable, but this would, by its length, be limited to a specific design energy and probably be built to produce Higgs, Higgs, and more Higgs. It seems to me like a VLHC would have more discovery potential for more massive Higgs particles, signs of supersymmetry, or whatever else might exist.
I hear there's petrol in Krokodil. That's pretty addictive and dangerous.