My reading of your post is that free software is a good model for journalism. You handle the most obvious objection that if "hobbyist journalists" replace professional ones, accuracy might suffer.
But I think that depth would suffer more than accuracy. Journalists today are (still) notorious for biased journalism, especially since that tends to sell better. But they are also motivated to really dig deep for original news stories. I don't think amateurs would do as well.
But the main flaw as I see it with applying the free software model to journalism is that free software is primarily _not_ developed by hobbyists. It's mostly developed by professionals whose companies need to specialize the software in one way or another. The business case is made (repeatedly) that extending existing open source software one way or another is easier than writing something proprietary from scratch. Or that open-sourcing new software will provide returns as others develop and debug the software further.
I don't see how the same kind of business case can be made for news, because it's not extensible in the same way. With news you have limited opportunities for incremental improvement. One improvement is verification, another is translation, and another is repackaging (search engines and aggregation like slashdot). Verification requires more actual investigative journalism. But it's still not at all like software where one can put in a small amount of effort to extend an existing large codebase to accomplish a fundamentally new task.
Maybe the disconnect is that news is an end product in and of itself, whereas software is a tool that helps people do something else?
I was just musing at what the court case might be like. Probably a jury case. First, select jurors that are easily persuaded by faulty logic. Then present theory after theory to make the jurors themselves feel unable to trust their own common sense. Overwhelm them with details and information. Sadly, propaganda tricks are commonplace (not at all illegal) in the courtroom.
I would hope that common sense would help someone judge which explanations are more likely and which ones are less likely. It's difficult to absolutely rule out the possibility that the scanner incident helped to cause these problems, but not hard to make the case that there are several other _far_ more likely scenarios where the scanner was purely harmless.
Flashing a price scanner in someone's eyes is rude and annoying, but so is waving your hand in front of someone's face, shouting at them, etc. Permanent harm could not have been anticipated. You can't sue someone for being a little bit rude one day.... Can you?
I'm not sure what to think about the real pros and cons of the bill. I'm mostly sad because driving in this morning was listening to a news program with a heavy republican bias. They likened passing the bill to 9/11 and the attack Pearl Harbor. I'm not kidding! Sound bites of politicians arguing were 100% propaganda tricks on both sides, and completely devoid of sound reason.
Media coverage and the strong polarization along party lines says a lot about US government. I'm concerned that my country is tearing itself apart in a massive power struggle between two parties. We don't have two competing ideologies. That would be impossible because ideologies are multi-faceted. We have two warring factions, us versus them. The very same techniques used to teach terrorists to hate their targets and soldiers to hate their enemies, dehumanisation etc, we're using on ourselves to hate republicans and democrats. Listen to 10 minutes of Limbaugh and you'll see exactly what I mean.
I'm disgusted by all this. And right here in this forum you see arguments on both sides that should ideally be settled by observation of fact. But no science can be done on this topic since there is no way to avoid overwhelming political bias. Is the bill a Good Thing? I'm sure I won't know for at least 50 years if ever.
Even though I'm skeptical that relatively low levels of microwave radiation could really be harmful, I thought I should point out that these metallic coatings or similar solutions do not absorb the radiation, they merely reflect it. Since complete coverage (floor, ceiling, windows, doors, etc) isn't realistic, you may easily reduce radiation overall but you might be allowing standing waves in certain locations, concentrating the radiation here and there, like hot spots in a microwave oven. An appartment is much bigger than a microwave oven, the walls are less flat, and there is more absorption etc. But the overall principle still applies to some degree. So I guess if I were worried enough about microwave radiation, still bought the apartment, but applied these reflective paints and such, then I'd also be worried about standing waves. Sniffing these out would be very time-consuming.
I'm not sure asking for evidence is a reasonable objection here. Would you go to a bittorrent search, gnutella, or other p2p network to try to find a version of Windows to install on your own computer? I don't have statistics but I'm pretty sure a significant fraction of other available-to-be-pirated software on p2p nets do contain trojans. A while ago I was trying to install XP on an upgraded gaming computer, going for a clean install since so many drivers had changed and such. I couldn't get my very old XP disk to work, and was looking into warez versions of, say, XP sp3. What I found was not at all reassuring. I eventually learned from a friend about slipstreaming the service pack in to make a new install CD from my old one and SP3. What a relief!
So I can actually appreciate VTBlue's perspective here, and also have some friends from India who say that almost all software they can buy there comes on a CD-R. So whether or not it's free of trojans is actually a point of real concern for many. Or maybe it's just understood that the copy of windows is not going to be "genuine" under most circumstances, I'm not certain, but in that case the question of whether your Windows copy is backdoor and trojan-free is still not addressed by WAT/WGA.
And I still don't see how checking automatically every 90 days is really the right solution given the supposed goal of simply reassuring the product owners. And I don't think that Windows that claims it's not genuine is truly functional in every sense. Other software can refuse to install or work in this case. But still the underlying issue of validating software of unknown origin is worth addressing somehow.
Yes, and maybe humans in some form have been around for a million years. Still it seems reasonable to think that if we found an extraterrestrial species, their evolutionary timeline could differ from ours by as much as 2 or 3 billion years. We've had radio technology for say 100 years. Making the very crude assumption that their technological development matches ours, then any species we find with radio telescopes would be somewhere (at random) between 100 years behind us and 2 billion years ahead of us. Sure it's not a uniform random distribution, but no matter what shape that distribution has, we would still expect any species found to be well in advance of us.
Two reasons why they are more likely to be "more intelligent" than us. One is that the universe is nearly 14 billion years old, and the human race is probably 50 thousand years old. If we find intelligent life, and they had a vaguely similar evolution of intelligence that we did, then it's very unlikely to be less evolved than us.
The other reason is by definition. SETI has so far (I think) been looking for life that has developed the capacity for radio communication. That's an even more recent evolution for us, making it all the more unlikely that this search could find life that's less intelligent than us. I guess radio communication might be naturally evolved somehow by a species that's not very intelligent. And then there's the possibility that in the short span of a few million years, if humans still exist, we'll consider radio communication so primitive that it has no practical uses over whatever else we use for communication instead.
Also I think the major focus of SETI has so far been "just looking" -- there isn't a program that I'm aware of to broadcast to other possible civilizations. We hardly need a program to accomplish that, however, so I think it's a moot argument. People have been broadcasting by radio for just over 100 years, so anyone closer to us than 100 lightyears would be able to pick that up. If we actively broadcast to ET, our signals wouldn't get there any faster than those from 100 years ago. They'd just be stronger in certain directions.
I didn't mean for the above comment to sound so sarcastic, but it comes off that way. I guess part of what I meant to say is that we all need to find ways of coping with fear. Science does not offer absolute truths, but it tends to be an effective way of eliminating wrong ideas and producing useful (not necessarily correct) theories. It's a way of dealing with uncertainty. I believe the scientists that have worked on these issues surrounding the LHC's operation have concluded, in many ways, that the currently-established Standard Model as well as most promising new theories don't predict anything dangerous happening. That's a kind of due diligence which seems reasonable to me. To worry beyond this is impractical since there is nothing to guide one's speculations.
As soon as you discount any arguments such as the GP (cosmic rays can have several magnitudes more energy than LHC protons, and have been harmlessly bombarding our sun and planets for the last 4 billion years) on the basis that our theories are imperfect, you open the door wide to hopeless speculation and paranoia. Yes, new behavior emerges at different scales of complexity, and considering how many large systems there are, only an infinitesimal fraction of all possible phenomenology is known.
Should I fear the implosion of the universe if I eat a peanut butter and jelly sandwich by biting off the corners first? Probably someone else has done that before, but they aren't me and I've certainly not eaten one like that before. My predictive theories of teeth-sandwich interactions are not absolute. A theory of everything is not known and even if it was, computations on the scale of my mouth would be impossible.
Heck, for that matter just sending this message is dangerous since this sequence of words has never been written before. I *think* it's harmless but I'm not absolutely 100% certain. Dive for cover, I'm sending this off!
I think a really good professor will avoid wasting too much time catering to the lowest common denominator. The problem with slowing down too much is that it lowers incentive for the students to work hard, which then causes the professor to go even slower. It's important to get frequent feedback from students while teaching, for their benefit as well as the professor's, and slow down once in a while if necessary, but the overall pace should be constant.
In my limited teaching experience, I found that lecturing was by far the least effective method of teaching, and was the most susceptible to this effect. If the class is spending most of the time doing activities, discussions, group problems, participating in demonstrations, and such, then all students become much more engaged across the board. Students who have fun (on topic) in class are far more motivated to do well, and will learn more in class too.
So while powerpoints can be both good and bad, lecturing of any kind should be kept to a minimum. Retention from lectures tends to be around 10%, whereas a good professor will be using a variety of activities that give closer to 50% retention. That's a big difference!
I had a physics professor who was very proud of the fact that he needed no notes and would fill board after board with derivations. He was very good at math, obviously, and had a good grasp of the subject. But he was not a particularly good teacher because he remained so focused on the algebra during class that very little was taught about how ideas were being represented and transformed. Had he spent more time explaining the setup, motivation, and multiple approaches one might try, he would have taught us much more. Time could have been saved by statements like "this ultimately can be written as..." Once in a while most of the hour would be wasted by "I'm missing a minus sign somewhere." and insisting that this be resolved.
Still, it's very sad when professors are penalized so much by failing students. Classes can be very different with different groups of students. It's very easy for a couple students with a bad attitude to turn a whole class against a novice professor.
I taught for a while at a college where very few professors did any research. I struggled to do a little bit of research but ultimately did not have enough time to devote to it. But I did see some of the reasons why keeping current in the field (by doing research) is important for professors. It is very difficult to divide one's time appropriately into teaching and research.
I'm too lazy to look up the data, but someone posted earlier the details on the radioactive content. These batteries use a sulfur isotope which I think is a *very* low energy (relative to most radioactive sources) beta emitter. I'm guessing that trace amounts of this stuff would not pose any health hazard to anyone. If you cracked open the battery and drank the whole thing, that couldn't be very smart but I doubt it would have any short-term effects. Long term effects would depend on how the body handles sulfur -- whether it passes through quickly or remains for a long time or even gets routed into vital organs.
There are natural radioactive sources all around us. As far as I know the studies linking low levels of radiation to cancer show no correlation below a certain level. The linear hypothesis that risk of cancer is linearly proportional to exposure appears only to work beyond a certain level that's well above natural levels. In addition to that the kind of radiation (alpha, beta, gamma, neutron, etc.) and energy of the particles makes a big difference, as well as what body parts are irradiated. I'm certainly no expert but my point is that the hazards of a given radioactive source are well-understood by scientists and yet much more complex than simply "It's radioactive! Bad!"
Many years ago an operator of an experimental nuclear reactor told me about an incident. He was accosted by a student who demanded that he shut the reactor down because it was leaking radiation. He could not convince the student that detectors in and around the building showed no signs of leakage. He finally got the student to explain what had upset them so much. They had found an apple core on the ground outside the lab and it had turned ALL BROWN!!!
The announcement says that these nuclear batteries have power densities a million times larger than standard batteries. That can't possibly be right unless it meant energy density instead.
But you forget that DNA are delicate critters, and like all life they need water to survive! Microwaves heat water which would burn some of the DNA causing them to mutate with anger!
Don't forget, too, that lots of viruses travel around the interweb! You don't want those viruses to get into garlic, do you? Even though they are computer viruses, they could mutate (from all the radiation) into a pandemic that wipes out all humans shorter than 5-foot seven. That's a lot of people, so we should never take such a stupid risk.
And don't bother telling me that microwaves cannot affect DNA. They can! Obviously, DNA are small, and so are microwaves or they wouldn't be called "micro". Gamma rays are much less harmful since they are rays and not waves and don't cause such disastrous shaking of the DNA. They are also much bigger since "gamma" is like "mega" which is MUCH bigger than "micro".
But seriously, I wonder since folks like this are completely impervious to logic, might they be convinced if they hear the mad ravings of someone even crazier than themselves?
There are a few things college administrations often tolerate that makes using Linux a little harder for students. One is distributing various files such as from Word or PowerPoint in OOXML format. I guess support for reading those formats is improving, but I don't think it's really good yet. Another problem is critical web-based services that require IE. Blackboard is one such beast by some reports, though I never had much trouble with it. I think Banner may also have occasional problems too. As others have mentioned here, spreadsheets with VB scripting are sometimes taken for granted, etc.
Obviously EndSoftwarePatents needs specific examples of how companies are being hurt. So far I don't see any examples posted here. It depends on what is meant by "harmed". Does this mean they have lost a court case? Perhaps the best example is IBM's own court expenses in the case brought against them by SCO. No, that was alleged copyright infringement, not patents. I guess a proper example is how the vfat filesystem in Linux has to dance around with short and long filenames. That's not on the list yet.
With copyrights, people loose productivity all the time by having to actively avoid looking at certain pieces of code for examples or ideas of good implementation. But with patents it's more of a sinister fear that any idea you come up with might be illegal to distribute, or "speak", because someone else might have patented it. You can't do anything about that other than live in fear, since there's no process in place to automatically avoid patent infringement. Maybe existing patent law could be argued as impinging on free speech! Okay I've rambled enough already.
I can't say you're wrong here that there are a lot of pitfalls and challenges getting Linux to work well. But one take on the responses we all get of "strange, it works great for me" is that these comments are actually helpful if you don't take them personally. In some cases, battery life is a problem when running Linux. In other experiences, it is not. That's the first step we have to take in identifying the problem. The next step is to try to figure out what things are different between these cases, and hence more people chime in with "does not!", "does too!" This can sound like kids bragging on a playground or a technical discussion, depending on your point of view. If you look carefully enough, there are suggestions for improving battery life scattered around. While not everyone will care about the details, at some point someone will probably care and do something about it and distros will evolve to be better in that respect.
I wish the driver situation was better for Linux. Maybe the Linux version of the Microsoft tax is accidentally buying the wrong hardware and having to get something new. My feeling is that on average this tax is nearly equivalent, or slightly less if you do a lot of research before buying. But even good research hasn't kept me from making the wrong choices on occasion.
A good rule of thumb when looking for work is to consider how much an employer is investing in their new hires up front. They can invest in the time spent interviewing, time and money for training, equipment, hiring bonus, moving expenses, etc. These days of course almost any job is better than none, but to whatever extent one has choices to make, I think up-front investment is one of the best indicators of a good employer.
(Note that this criteria eliminates all pyramid sales jobs. The pyramid structure would not work if hiring cost anything more than the time of the next guy up in the pyramid.)
So I'm agreeing with the above post: if you are looked over because you had EE CAD experience but not of the right type, then they didn't want to invest in any training and you can consider your own investigative work to have been done for you.
Maybe part of the problem is the expected length of employment? If you only expect to hire an employee for 3 months, then the couple of days it would take a competent engineer to go from GenericCAD to OrCad might be significant "on paper". It's a tiny percentage when compared with differences in talent from one individual to the next, but it might on the surface appear to be more measurable (tangible) for a manager reviewing dozens or even hundreds of resumes. If you hope to hire an employee for a few years at least, then this could not possibly be a meaningful hiring issue.
In management-speak, you might want to work toward having an agile company instead.
You might even consider the reverse bias. A student who learned strictly on OrCad and nothing else *might* have been taught in a procedural way which is very much inferior. I know from teaching physics that most introductory courses are taught with a similar approach: teach the equations, and problem solving becomes a process of searching your equation library for ones that contain all known and the unknown quantities in a given "word problem". (Some algebra might be necessary.) But research shows that students in these courses come out with NO measurable improvement of their conceptual grasp of physics, and don't develop any real-world problem-solving abilities either. The course might be good as a weed-out except the talented students who push themselves to truly understand the subject tend to do worse than those who learn to play the game of plug-n-chug. Those professors who care about teaching, have the time and resources necessary, and pay any attention at all to pedagogy adopt entirely new approaches.
To me this is a similar situation to the CAD-of-the-day issue. If you feel an employee may lack the flexibility to quickly learn the particular software in use in your company, then that's a good reason not to hire them. This also applies to some extent to programming languages. Although there the learning curve can be a little more steep, again it depends strongly on the breadth and skill of the applicant more than the specific languages they have used.
Why was this post modded down? It's very relevant. I don't agree with this hiring criteria but it's a prevalent one. Moderators should only mod down for posts that detract from meaningful debate.
I should add that one of the most common ways to handle this kind of analysis correctly is to divide the data randomly into two samples. Study the first sample carefully to choose the best selection criteria. Then and only then, apply that selection on the second (previously unseen) sample and immediately report the results without making any excuses or adjustments. This isn't exactly foolproof and can be improved upon in many ways, but for a casual study like this it's not bad.
There is a common error made in statistics to take the statistical significance of results like this at face value, without regard to how carefully you have chosen the selection criteria.
How many ways could Bennet have chosen an elite sample? He chose those who scored themselves as Excelent, or 5 on a 5 point scale. He could have chosen 4 or above, 3 or above, in math. He could even have looked for significance in how those with poor math skills tended to call this a crime. He also mentions the more mild correlation with those self-scoring English/composition skills as 5/5. Of all the different ways he could have sliced up the sample, which ways would we have considered interesting? What about all the folks who scored 1s OR 5s, the extremists?
I don't know how many variations Bennet tried before deciding on those who scored 5/5 in math as being the noteworthy population. But I can easily think of about 10 different variations that would seem potentially interesting or worthy of interpretation.
If you slice up data in 10 different ways, you would expect that even if they are all random results with no inherent correlation to anything surveyed, you'll find a 90% significant result I guess about half the time. I'm not being really precise here, but the point is that to a rough approximation, you should multiply 1% (1-99%) by about 10, the number of reasonable ways to choose his sample. That's closer to the real chance that his result was a random fluke. Now 10% is still low, but at this point I would make a weaker claim that this study is not by itself conclusive, but suggests a promising direction in which a larger and more careful study might find an insightful result.
I think Richard Feynman, at the beginning of one of his famous lectures, told the audience the license plate number of the car parked in front of him in the parking lot. He said something along the lines of "What are the chances of that?! The one car with that exact license plate number just happened to be parked there, in front of me!" The point is that this would have been surprising if he had chosen or written down the number before he looked at the other car, but there is a 100% chance that whatever license plate he spotted would have something on it. If Bennet had chosen his 5/5 in math qualification before looking at the data, then the significance would have been correctly computed. But he would more likely have chosen a different criteria which would not have shown as high a correlation.
A similar story was posted on Slashdot, running in the Washington Post I think, about a 99.5% likelihood that the Iranian election results were tampered with. This suffered even worse from this error, with highly arbitrary criteria such as finding a 5 or a 2 in the last two digits of election results. In that case the criteria was chosen from such a large variety of possibilities, I was surprised the "significance" was as low as it was! and to then run an article claiming election fraud (an issue about which people are violently concerned) in the Washington Post based on such erroneous statistics is just tragic.
Slashdot Mirror
My reading of your post is that free software is a good model for journalism. You handle the most obvious objection that if "hobbyist journalists" replace professional ones, accuracy might suffer.
But I think that depth would suffer more than accuracy. Journalists today are (still) notorious for biased journalism, especially since that tends to sell better. But they are also motivated to really dig deep for original news stories. I don't think amateurs would do as well.
But the main flaw as I see it with applying the free software model to journalism is that free software is primarily _not_ developed by hobbyists. It's mostly developed by professionals whose companies need to specialize the software in one way or another. The business case is made (repeatedly) that extending existing open source software one way or another is easier than writing something proprietary from scratch. Or that open-sourcing new software will provide returns as others develop and debug the software further.
I don't see how the same kind of business case can be made for news, because it's not extensible in the same way. With news you have limited opportunities for incremental improvement. One improvement is verification, another is translation, and another is repackaging (search engines and aggregation like slashdot). Verification requires more actual investigative journalism. But it's still not at all like software where one can put in a small amount of effort to extend an existing large codebase to accomplish a fundamentally new task.
Maybe the disconnect is that news is an end product in and of itself, whereas software is a tool that helps people do something else?
So I agree with GP. Sorry.
I was wondering if someone was going to point this out, and there you go! That's the Achilles' heel.
I like that idea! But the air inside might become too dry and cause problems with static electricity. I don't see any other problems.
I was just musing at what the court case might be like. Probably a jury case. First, select jurors that are easily persuaded by faulty logic. Then present theory after theory to make the jurors themselves feel unable to trust their own common sense. Overwhelm them with details and information. Sadly, propaganda tricks are commonplace (not at all illegal) in the courtroom.
I would hope that common sense would help someone judge which explanations are more likely and which ones are less likely. It's difficult to absolutely rule out the possibility that the scanner incident helped to cause these problems, but not hard to make the case that there are several other _far_ more likely scenarios where the scanner was purely harmless.
Flashing a price scanner in someone's eyes is rude and annoying, but so is waving your hand in front of someone's face, shouting at them, etc. Permanent harm could not have been anticipated. You can't sue someone for being a little bit rude one day. ... Can you?
I'm not sure what to think about the real pros and cons of the bill. I'm mostly sad because driving in this morning was listening to a news program with a heavy republican bias. They likened passing the bill to 9/11 and the attack Pearl Harbor. I'm not kidding! Sound bites of politicians arguing were 100% propaganda tricks on both sides, and completely devoid of sound reason.
Media coverage and the strong polarization along party lines says a lot about US government. I'm concerned that my country is tearing itself apart in a massive power struggle between two parties. We don't have two competing ideologies. That would be impossible because ideologies are multi-faceted. We have two warring factions, us versus them. The very same techniques used to teach terrorists to hate their targets and soldiers to hate their enemies, dehumanisation etc, we're using on ourselves to hate republicans and democrats. Listen to 10 minutes of Limbaugh and you'll see exactly what I mean.
I'm disgusted by all this. And right here in this forum you see arguments on both sides that should ideally be settled by observation of fact. But no science can be done on this topic since there is no way to avoid overwhelming political bias. Is the bill a Good Thing? I'm sure I won't know for at least 50 years if ever.
Even though I'm skeptical that relatively low levels of microwave radiation could really be harmful, I thought I should point out that these metallic coatings or similar solutions do not absorb the radiation, they merely reflect it. Since complete coverage (floor, ceiling, windows, doors, etc) isn't realistic, you may easily reduce radiation overall but you might be allowing standing waves in certain locations, concentrating the radiation here and there, like hot spots in a microwave oven. An appartment is much bigger than a microwave oven, the walls are less flat, and there is more absorption etc. But the overall principle still applies to some degree. So I guess if I were worried enough about microwave radiation, still bought the apartment, but applied these reflective paints and such, then I'd also be worried about standing waves. Sniffing these out would be very time-consuming.
I'm not sure asking for evidence is a reasonable objection here. Would you go to a bittorrent search, gnutella, or other p2p network to try to find a version of Windows to install on your own computer? I don't have statistics but I'm pretty sure a significant fraction of other available-to-be-pirated software on p2p nets do contain trojans. A while ago I was trying to install XP on an upgraded gaming computer, going for a clean install since so many drivers had changed and such. I couldn't get my very old XP disk to work, and was looking into warez versions of, say, XP sp3. What I found was not at all reassuring. I eventually learned from a friend about slipstreaming the service pack in to make a new install CD from my old one and SP3. What a relief!
So I can actually appreciate VTBlue's perspective here, and also have some friends from India who say that almost all software they can buy there comes on a CD-R. So whether or not it's free of trojans is actually a point of real concern for many. Or maybe it's just understood that the copy of windows is not going to be "genuine" under most circumstances, I'm not certain, but in that case the question of whether your Windows copy is backdoor and trojan-free is still not addressed by WAT/WGA.
And I still don't see how checking automatically every 90 days is really the right solution given the supposed goal of simply reassuring the product owners. And I don't think that Windows that claims it's not genuine is truly functional in every sense. Other software can refuse to install or work in this case. But still the underlying issue of validating software of unknown origin is worth addressing somehow.
Yes, and maybe humans in some form have been around for a million years. Still it seems reasonable to think that if we found an extraterrestrial species, their evolutionary timeline could differ from ours by as much as 2 or 3 billion years. We've had radio technology for say 100 years. Making the very crude assumption that their technological development matches ours, then any species we find with radio telescopes would be somewhere (at random) between 100 years behind us and 2 billion years ahead of us. Sure it's not a uniform random distribution, but no matter what shape that distribution has, we would still expect any species found to be well in advance of us.
Two reasons why they are more likely to be "more intelligent" than us. One is that the universe is nearly 14 billion years old, and the human race is probably 50 thousand years old. If we find intelligent life, and they had a vaguely similar evolution of intelligence that we did, then it's very unlikely to be less evolved than us.
The other reason is by definition. SETI has so far (I think) been looking for life that has developed the capacity for radio communication. That's an even more recent evolution for us, making it all the more unlikely that this search could find life that's less intelligent than us. I guess radio communication might be naturally evolved somehow by a species that's not very intelligent. And then there's the possibility that in the short span of a few million years, if humans still exist, we'll consider radio communication so primitive that it has no practical uses over whatever else we use for communication instead.
Also I think the major focus of SETI has so far been "just looking" -- there isn't a program that I'm aware of to broadcast to other possible civilizations. We hardly need a program to accomplish that, however, so I think it's a moot argument. People have been broadcasting by radio for just over 100 years, so anyone closer to us than 100 lightyears would be able to pick that up. If we actively broadcast to ET, our signals wouldn't get there any faster than those from 100 years ago. They'd just be stronger in certain directions.
I didn't mean for the above comment to sound so sarcastic, but it comes off that way. I guess part of what I meant to say is that we all need to find ways of coping with fear. Science does not offer absolute truths, but it tends to be an effective way of eliminating wrong ideas and producing useful (not necessarily correct) theories. It's a way of dealing with uncertainty. I believe the scientists that have worked on these issues surrounding the LHC's operation have concluded, in many ways, that the currently-established Standard Model as well as most promising new theories don't predict anything dangerous happening. That's a kind of due diligence which seems reasonable to me. To worry beyond this is impractical since there is nothing to guide one's speculations.
As soon as you discount any arguments such as the GP (cosmic rays can have several magnitudes more energy than LHC protons, and have been harmlessly bombarding our sun and planets for the last 4 billion years) on the basis that our theories are imperfect, you open the door wide to hopeless speculation and paranoia. Yes, new behavior emerges at different scales of complexity, and considering how many large systems there are, only an infinitesimal fraction of all possible phenomenology is known.
Should I fear the implosion of the universe if I eat a peanut butter and jelly sandwich by biting off the corners first? Probably someone else has done that before, but they aren't me and I've certainly not eaten one like that before. My predictive theories of teeth-sandwich interactions are not absolute. A theory of everything is not known and even if it was, computations on the scale of my mouth would be impossible.
Heck, for that matter just sending this message is dangerous since this sequence of words has never been written before. I *think* it's harmless but I'm not absolutely 100% certain. Dive for cover, I'm sending this off!
I think a really good professor will avoid wasting too much time catering to the lowest common denominator. The problem with slowing down too much is that it lowers incentive for the students to work hard, which then causes the professor to go even slower. It's important to get frequent feedback from students while teaching, for their benefit as well as the professor's, and slow down once in a while if necessary, but the overall pace should be constant.
In my limited teaching experience, I found that lecturing was by far the least effective method of teaching, and was the most susceptible to this effect. If the class is spending most of the time doing activities, discussions, group problems, participating in demonstrations, and such, then all students become much more engaged across the board. Students who have fun (on topic) in class are far more motivated to do well, and will learn more in class too.
So while powerpoints can be both good and bad, lecturing of any kind should be kept to a minimum. Retention from lectures tends to be around 10%, whereas a good professor will be using a variety of activities that give closer to 50% retention. That's a big difference!
I had a physics professor who was very proud of the fact that he needed no notes and would fill board after board with derivations. He was very good at math, obviously, and had a good grasp of the subject. But he was not a particularly good teacher because he remained so focused on the algebra during class that very little was taught about how ideas were being represented and transformed. Had he spent more time explaining the setup, motivation, and multiple approaches one might try, he would have taught us much more. Time could have been saved by statements like "this ultimately can be written as ..." Once in a while most of the hour would be wasted by "I'm missing a minus sign somewhere." and insisting that this be resolved.
Still, it's very sad when professors are penalized so much by failing students. Classes can be very different with different groups of students. It's very easy for a couple students with a bad attitude to turn a whole class against a novice professor.
I taught for a while at a college where very few professors did any research. I struggled to do a little bit of research but ultimately did not have enough time to devote to it. But I did see some of the reasons why keeping current in the field (by doing research) is important for professors. It is very difficult to divide one's time appropriately into teaching and research.
I'm too lazy to look up the data, but someone posted earlier the details on the radioactive content. These batteries use a sulfur isotope which I think is a *very* low energy (relative to most radioactive sources) beta emitter. I'm guessing that trace amounts of this stuff would not pose any health hazard to anyone. If you cracked open the battery and drank the whole thing, that couldn't be very smart but I doubt it would have any short-term effects. Long term effects would depend on how the body handles sulfur -- whether it passes through quickly or remains for a long time or even gets routed into vital organs.
There are natural radioactive sources all around us. As far as I know the studies linking low levels of radiation to cancer show no correlation below a certain level. The linear hypothesis that risk of cancer is linearly proportional to exposure appears only to work beyond a certain level that's well above natural levels. In addition to that the kind of radiation (alpha, beta, gamma, neutron, etc.) and energy of the particles makes a big difference, as well as what body parts are irradiated. I'm certainly no expert but my point is that the hazards of a given radioactive source are well-understood by scientists and yet much more complex than simply "It's radioactive! Bad!"
Many years ago an operator of an experimental nuclear reactor told me about an incident. He was accosted by a student who demanded that he shut the reactor down because it was leaking radiation. He could not convince the student that detectors in and around the building showed no signs of leakage. He finally got the student to explain what had upset them so much. They had found an apple core on the ground outside the lab and it had turned ALL BROWN!!!
The announcement says that these nuclear batteries have power densities a million times larger than standard batteries. That can't possibly be right unless it meant energy density instead.
Cool stuff even so!
But you forget that DNA are delicate critters, and like all life they need water to survive! Microwaves heat water which would burn some of the DNA causing them to mutate with anger!
Don't forget, too, that lots of viruses travel around the interweb! You don't want those viruses to get into garlic, do you? Even though they are computer viruses, they could mutate (from all the radiation) into a pandemic that wipes out all humans shorter than 5-foot seven. That's a lot of people, so we should never take such a stupid risk.
And don't bother telling me that microwaves cannot affect DNA. They can! Obviously, DNA are small, and so are microwaves or they wouldn't be called "micro". Gamma rays are much less harmful since they are rays and not waves and don't cause such disastrous shaking of the DNA. They are also much bigger since "gamma" is like "mega" which is MUCH bigger than "micro".
But seriously, I wonder since folks like this are completely impervious to logic, might they be convinced if they hear the mad ravings of someone even crazier than themselves?
There are a few things college administrations often tolerate that makes using Linux a little harder for students. One is distributing various files such as from Word or PowerPoint in OOXML format. I guess support for reading those formats is improving, but I don't think it's really good yet. Another problem is critical web-based services that require IE. Blackboard is one such beast by some reports, though I never had much trouble with it. I think Banner may also have occasional problems too. As others have mentioned here, spreadsheets with VB scripting are sometimes taken for granted, etc.
Obviously EndSoftwarePatents needs specific examples of how companies are being hurt. So far I don't see any examples posted here. It depends on what is meant by "harmed". Does this mean they have lost a court case? Perhaps the best example is IBM's own court expenses in the case brought against them by SCO. No, that was alleged copyright infringement, not patents. I guess a proper example is how the vfat filesystem in Linux has to dance around with short and long filenames. That's not on the list yet.
With copyrights, people loose productivity all the time by having to actively avoid looking at certain pieces of code for examples or ideas of good implementation. But with patents it's more of a sinister fear that any idea you come up with might be illegal to distribute, or "speak", because someone else might have patented it. You can't do anything about that other than live in fear, since there's no process in place to automatically avoid patent infringement. Maybe existing patent law could be argued as impinging on free speech! Okay I've rambled enough already.
I can't say you're wrong here that there are a lot of pitfalls and challenges getting Linux to work well. But one take on the responses we all get of "strange, it works great for me" is that these comments are actually helpful if you don't take them personally. In some cases, battery life is a problem when running Linux. In other experiences, it is not. That's the first step we have to take in identifying the problem. The next step is to try to figure out what things are different between these cases, and hence more people chime in with "does not!", "does too!" This can sound like kids bragging on a playground or a technical discussion, depending on your point of view. If you look carefully enough, there are suggestions for improving battery life scattered around. While not everyone will care about the details, at some point someone will probably care and do something about it and distros will evolve to be better in that respect.
I wish the driver situation was better for Linux. Maybe the Linux version of the Microsoft tax is accidentally buying the wrong hardware and having to get something new. My feeling is that on average this tax is nearly equivalent, or slightly less if you do a lot of research before buying. But even good research hasn't kept me from making the wrong choices on occasion.
A good rule of thumb when looking for work is to consider how much an employer is investing in their new hires up front. They can invest in the time spent interviewing, time and money for training, equipment, hiring bonus, moving expenses, etc. These days of course almost any job is better than none, but to whatever extent one has choices to make, I think up-front investment is one of the best indicators of a good employer.
(Note that this criteria eliminates all pyramid sales jobs. The pyramid structure would not work if hiring cost anything more than the time of the next guy up in the pyramid.)
So I'm agreeing with the above post: if you are looked over because you had EE CAD experience but not of the right type, then they didn't want to invest in any training and you can consider your own investigative work to have been done for you.
Maybe part of the problem is the expected length of employment? If you only expect to hire an employee for 3 months, then the couple of days it would take a competent engineer to go from GenericCAD to OrCad might be significant "on paper". It's a tiny percentage when compared with differences in talent from one individual to the next, but it might on the surface appear to be more measurable (tangible) for a manager reviewing dozens or even hundreds of resumes. If you hope to hire an employee for a few years at least, then this could not possibly be a meaningful hiring issue.
In management-speak, you might want to work toward having an agile company instead.
You might even consider the reverse bias. A student who learned strictly on OrCad and nothing else *might* have been taught in a procedural way which is very much inferior. I know from teaching physics that most introductory courses are taught with a similar approach: teach the equations, and problem solving becomes a process of searching your equation library for ones that contain all known and the unknown quantities in a given "word problem". (Some algebra might be necessary.) But research shows that students in these courses come out with NO measurable improvement of their conceptual grasp of physics, and don't develop any real-world problem-solving abilities either. The course might be good as a weed-out except the talented students who push themselves to truly understand the subject tend to do worse than those who learn to play the game of plug-n-chug. Those professors who care about teaching, have the time and resources necessary, and pay any attention at all to pedagogy adopt entirely new approaches.
To me this is a similar situation to the CAD-of-the-day issue. If you feel an employee may lack the flexibility to quickly learn the particular software in use in your company, then that's a good reason not to hire them. This also applies to some extent to programming languages. Although there the learning curve can be a little more steep, again it depends strongly on the breadth and skill of the applicant more than the specific languages they have used.
Why was this post modded down? It's very relevant. I don't agree with this hiring criteria but it's a prevalent one. Moderators should only mod down for posts that detract from meaningful debate.
I should add that one of the most common ways to handle this kind of analysis correctly is to divide the data randomly into two samples. Study the first sample carefully to choose the best selection criteria. Then and only then, apply that selection on the second (previously unseen) sample and immediately report the results without making any excuses or adjustments. This isn't exactly foolproof and can be improved upon in many ways, but for a casual study like this it's not bad.
There is a common error made in statistics to take the statistical significance of results like this at face value, without regard to how carefully you have chosen the selection criteria.
How many ways could Bennet have chosen an elite sample? He chose those who scored themselves as Excelent, or 5 on a 5 point scale. He could have chosen 4 or above, 3 or above, in math. He could even have looked for significance in how those with poor math skills tended to call this a crime. He also mentions the more mild correlation with those self-scoring English/composition skills as 5/5. Of all the different ways he could have sliced up the sample, which ways would we have considered interesting? What about all the folks who scored 1s OR 5s, the extremists?
I don't know how many variations Bennet tried before deciding on those who scored 5/5 in math as being the noteworthy population. But I can easily think of about 10 different variations that would seem potentially interesting or worthy of interpretation.
If you slice up data in 10 different ways, you would expect that even if they are all random results with no inherent correlation to anything surveyed, you'll find a 90% significant result I guess about half the time. I'm not being really precise here, but the point is that to a rough approximation, you should multiply 1% (1-99%) by about 10, the number of reasonable ways to choose his sample. That's closer to the real chance that his result was a random fluke. Now 10% is still low, but at this point I would make a weaker claim that this study is not by itself conclusive, but suggests a promising direction in which a larger and more careful study might find an insightful result.
I think Richard Feynman, at the beginning of one of his famous lectures, told the audience the license plate number of the car parked in front of him in the parking lot. He said something along the lines of "What are the chances of that?! The one car with that exact license plate number just happened to be parked there, in front of me!" The point is that this would have been surprising if he had chosen or written down the number before he looked at the other car, but there is a 100% chance that whatever license plate he spotted would have something on it. If Bennet had chosen his 5/5 in math qualification before looking at the data, then the significance would have been correctly computed. But he would more likely have chosen a different criteria which would not have shown as high a correlation.
A similar story was posted on Slashdot, running in the Washington Post I think, about a 99.5% likelihood that the Iranian election results were tampered with. This suffered even worse from this error, with highly arbitrary criteria such as finding a 5 or a 2 in the last two digits of election results. In that case the criteria was chosen from such a large variety of possibilities, I was surprised the "significance" was as low as it was! and to then run an article claiming election fraud (an issue about which people are violently concerned) in the Washington Post based on such erroneous statistics is just tragic.