Yeah, the crime of the modern educational system is that it produces people who know all the answers and have no sense of wonder. That "older than dirt" guy probably looks at a computer and only sees a white box.
You should look at a computer and see the thread of execution hopping between kernel routines and pausing at mutexes. You should see the electrons whooshing through the silicon, underneath an overhanging crystalline gate electrode. You should feel the electric field sucking at you: it's almost strong enough to rip electrons out of the SiO2 dielectric. And back up at higher level, those spin locks should be like an amusement park ride: puke your guts out if you go around in one for more than a few microseconds.:)
Yah, we knew stars became red giants. But that's not the right way to look at it.
Yah, yah, yah. Everyone knows that there is climate change that "just happens". But so what? That doesn't prevent humans from changing the climate also.
The AGW story is that human-generated climate change has been fairly small so far. But it is becoming more and more important, as everyone's economy grows and the population grows. Who would argue with that? We have more cars, more computers, more cheap airline flights. India and China have been growing economically at amazing speeds and they have been burning more coal and whatever. Anyone can see that whatever part of climate change is caused by use humans will be rather bigger in the coming century than it has been in the last.
So, it is easy to imagine that we were relatively unimportant in the last century, we are now having effects comparable to natural fluctuations, and in the next century we will be driving big changes.
Throwing out all the data is a reasonable response, but probably over-cautious. There are twice as many of us humans now as there were in 1960, and we are affecting the environment more. So, perhaps trees are affected by recent changes in the environment; and the rules are different after 1960. I don't know: I'm not an expert on trees, but there are tree experts out there. Somebody knows the right answer (or will know it, soon enough).
Ultimately, throwing out useful data can be just as bad as keeping some contaminated data. This isn't a question that has a simple answer. No one is pretending that trees are perfect thermomenters; it's all a matter of comparing the magnitude of the temperature effect to everything else that can affect the growth of a tree. One cannot make that comparison sensibly without doing some serious research on trees. If you haven't done any, perhaps you shouldn't comment.
The thing is, there's plenty of evidence that we humans affect the environment, and we need data to build models so that we can respond intelligently. Suppose, just suppose, that we really do need to reduce CO2 production 80% by 2050. Wouldn't you want to know about it? That magnitude of reduction really won't happen by the natural operation of free markets, not without a carbon tax or similar.
So, we need to have good climate data, as much as possible, and we need to build models. I'm a firm believer in opening up this critical research to make sure it is as accurate as possible, but perfection may not be attainable. Nonetheless, we will need to operate on whatever is the best achievable, rather than demanding impossible perfection. Demanding perfection is just an way of hiding one's head in the sand.
> And read about their naming conventions for files... they were an IT disaster.
Yes? So? The question is, did they manage to run the right scripts on the right files, eventually. Naming conventions just make that easier or harder. You can still get the right answer with bad names, or get the wrong answer with a brilliant naming scheme.
Geez! Programmers think that science is nothing more than programming...
Personally, I am also disturbed by the slapdash approach that seems evident in HARRY_READ_ME, but the point is that slapdash is not the same as wrong. And, of course, bugs do not necessarily imply useless (as anyone who runs software should know quite well).
> Here is a hint. If he says "Trust me" he ain't no scientist he is a salesman/politician.
Not really. Science progresses because we generally trust each other. If you didn't trust other people's results, you'd spend your entire career repeating experiments that someone else has already done correctly. You'd waste your life.
Of course, you trust, but not 100%. You keep an eye open for things that might indicate that someone is wrong or incompetent. But, really, most of the checking happens when you try to use someone's result as a tool. If they are wrong, it usually becomes pretty obvious pretty fast "Hey! all my nails are bending over. I wonder if this is a good hammer..."
No, I do it all the time, and it is the correct thing to do.
A scientific review is not a trap for fraud or a re-analysis of the data. It is not adversarial (well, it is not intrinsically adversarial). The idea is that you are helping the person write a better paper, in addition to deciding whether it is good enough to publish. And you assume that they have described their work accurately.
Fraud gets detected sooner or later when people try to replicate the experiment. And, wrong papers get detected that way also.
Reviews are there to remove (not catch!) any visible errors, to make sure that the logic make sense, to make sure that nothing important was forgotten, and to make sure that the experiment was described completely enough so that someone else could replicate it. That's more than enough work for the poor (unpaid) reviewer.
In civilized countries, justice is temperred with mercy and the knowledge that everyone does something stupid occasionally. I'm not entirely sure that there is a civilized country yet, but I look forward to it.
Well, don't forget that there are good _approximate_ solutions for many NP problems.
NP can be a red herring in the real world: we don't need to solve (for example) the traveling salesman problem _exactly_. So long as we get within 1% or so of the optimum solution, that's good enough for most purposes.
(Of course, we don't have good approximations for all NP problems, only some of them, so it isn't a total red herring.)
Because one of the greatest of all human rights is to pack your bags and go somewhere else.
Visas shouldn't be entirely for the convenience of corporations. They should be for people, too. As far as I'm concerned, if you can pull your weight, you should be welcome.
The most important thing you lose in an on-line university is the research-teaching connection.
The thing that most people don't realize as students, is that there are many very attractive ideas that are simply wrong. They are as attractive to the professors as they are to the students (professors are people too). And, some fields (like Linguistics, where I have personal knowledge) it is a permanent struggle to avoid the easy, attractive theories. The real experimental data is often much more complex and messy than anyone wants to believe.
If you don't do research yourself, or don't have colleagues who are doing research, it's all too easy to believe whatever is easy to teach. Chomsky, for instance. While the guy deserves credit for starting the field, most of what he said has since been proven wrong (except where he was careful to hedge his statements). But, his theories live on because people want to believe them.
So, if you cut the teaching-research linkage, I fear that we will gradually lose most of our knowledge and settle into the comfortable routine of teaching a curriculum with all the rough edges smoothed away. Efficient, but (largely) wrong.
"Nothing (except my generally honest nature, and a lack of money) was stopping me from hiring a smarter person to write my papers and do my projects, or even sit in classes and take my tests. "
Indeed. And nothing is stopping you from hiring a world-class football player to have a career in your name. Or from hiring a charismatic politician to run for office for you. I agree.
Yet, for some unknown reason, none of these types of identity fraud happen offen enough to cause much trouble. I wonder why.
Yeah, like History is important. Take a good course and think about it a little bit, and the first lesson is that an awful lot of people in the past were simultaneously (a) completely wrong, and (b) completely convinced they were right.
If you apply that knowledge when you look around you, it changes the way you look at the world quite a bit.
The thing that no one has mentioned here is that universities benefit everyone, not jus their students. Not just their investors (as if they had them!).
It's very simple. Get sick and go to the doctor. Would you rather go to an educated doctor or an uneducated one? So, everyone benefits from the education that the doctor received. Likewise, go up in an elevator. Do you want the building to be designed by a well-educated engineer, or someone who went to a diploma mill? Again, the well-educated engineer is a public good, valuable to everyone.
The same with scientists, programmers, everyone. The point is that life is better when you're surrounded by educated people. So society should pay for that to happen. That's why sensible governments pay for universities out of public money: because the universities benefit the entire public.
Of course, they benefit the students, too. That's true, but it doesn't make the value to the public disappear.
So, if you look at a university as a profit-making enterprise, you'll kill it, and then complain that you can't get well-educated help any more.
And universities do more than just teach, too, but that's another article...
I like to think that my bug reports and feature requests are helpful. Certainly I intend them that way, though I suppose it's easy enough to imagine some maintainers think otherwise.
There isn't a sharp line between feature requests and bug reports. Personally, I take the liberal stance and say that "if it confuses the user, it's a bug", but some differ. Some strange people even write specifications and believe that "if it meets the specifications, it's not a bug." I can't agree with that. If a program misbehaves but meets specifications, that's a bug in the specifications. And, if the specifications are wrong, then you need to change the code anyway, just as much as if the bug were in the code.
I went along to some church (Methodist) adult classes on Genesis once. The preacher did a good job, walking the fine line and not interpreting Genesis too literally. But, some of the people there, my neighbours in suburban NJ were not happy with this. A few people were really asking him and pushing him on the reality of the flood and creation in seven days. And how to squeeze dinosaurs into it.
I was shocked. I ended up going into a little speech on radioactive dating and geology, pointing out that there was a lot of evidence for the age of the Earth, and asking if they really wanted to believe in a God who would plant misleading evidence in the rocks. I don't know if I convinced anyone or not. (Though, I must say, no one seemed particularly upset.)
Scientists outside their field are not experts, of course. But, neither are they to be entirely dismissed. Even outside your field, you know how to think logically, you know about statistics, you know that science is a bit of a messy process sometimes, and you know how to evaluate what other scientists write. And, there may be a bit of overlap from one field to another, too.
So, for instance, I am a non-climate scientist. I read the public stuff. I've gone to google scholar and read some of the techical papers. And, while I'd have to spend a decade to really get up to speed and do some of the work, I can understand generally what people are doing.
When I read, I look for failures in logic. I look for places where they make assumptions that are not supported. I look for obvious violations of the physics I learned in school and the bits and pieces I've picked up from reading other papers. I look for signs that people hold their opinions for political reasons rather than because of their data.
And, by and large, it all looks OK. So, I believe their conclusions. That's how *I* do it. No, I do not simply believe what I was taught in school.
Well, most patents are that way. Most patents are things that are cranked out on demand to satisfy some corporation. Many of mine fall into that category.
But, not all. I'm quite proud of a few of them, and there weren't other people doing the same thing at the same time. And I've seen a few inventions (not mine) that are really quite impressive. So, you are certainly over-generalizing.
As for your assertion that "patents are fundamentally based on the assumption that... a particular invention is unique to a particular inventor". Well, I don't know where that came from. I'd rather think that patents are based on the assumption that if you give people incentives to invent and to commercialize their inventions, then you'll get more interesting products.
That last bit may or may not be true -- you certainly get more inventions, but whether you end up with more interesting products or not is a good question. One that should be settled by some kind of research and certainly cannot be settled by bare assertion.
But, what if you say "Excuse me, I have an operating audio recorder."? Then, presumably, if the person continues talking, he is implicitly giving consent to being recorded.
Of course, they may get you for interfering with a police officer/TSA employee/whoever, on the grounds that he cannot fulfill his duties without talking. But I'd guess that's not a felony.
... Booked in advance, a train from London to Edinburgh can be as little as £12.
Well, not all of England lives in London, no matter what Londerners think. (Or, maybe you're just wrong...) You can't get from Oxford to Edinburgh for £12. I looked at the National Rail Journey Planner, and it quotes £154 round trips a month in advance. It looks like there may be fares as cheap as £86, but not lower.
I've lived here for 5 years now, and the only place I've gone for £12 or less is Didcot, which is just 10 miles down the road. Well, maybe Banbury, which is perhaps 20 miles up the road...
Yeah, except that rail isn't cheap for passengers. Here in the UK, you can fly to the South of France for the price of a rail ticket to Scotland. (I.e. On rail, it costs about GBP100 = US$160 to go 350 miles.)
If rail is so efficient for passengers (it presumably *is* for bulk freight) why ain't it cheap?
Certainly rail's fuel costs are small, but what about the carbon costs of all those guys standing around in fluorescent yellow vests?
Anyway, for a shared-use RDBMS, it might be moot, because it usually has more than one query to process at any given time - and so it's easy to load-balance all cores just by assigning queries to them.
Ha! Not so! Load balancing like that means the response time is long. Long response times do eventually add up, especially if you are combining data from many database servers, like in a web mash-up.
Ideally, if you could do the queries really fast, and work at only one query at a time with all processors, everyone would be happier.
Yeah. Is it supposed to be cool to work too hard or is it supposed to be stupid?
Seems to me like a properly planned and managed launch shouldn't be all that exciting. Or painful, whichever.
It's never like that: no one ever tells you to violate safety rules; it is just made clear that X needs to get done. The rest is left up to the student.
The pressure is often self-applied. Everyone on the academic track knows that you need to publish or perish (sorry!). The thing is, that when you go hunting for your next post-doc or your professorial position, all that matters is results. And there are lots of applicants for permanent positions, so it is crucial to get more results than everyone else. I've gotten more than one letter back from some university apologizing for delays in a job application process because "...we had over 100 applicants" or "...we had over 200 applicants."
In that kind of environment, it is amazing that safety gets any attention at all. But it does, even if perhaps not enough.
And, don't forget that in a research environment, everyone is making up procedures as they go along. Industry has the advantage that you can do something again and again, until you figure out the best way to do it. Researchers often don't have that option. Once you've done something a few times, that's the end of it. Either you graduate or you move into another part of the experiment. Or, the technique becomes obsolete, or it needs to be modified for the needs of some other experiment.
Correct. One of the proper functions of government is to stop people from doing stupid things, like driving 70 in a residential area or selling their future.
Another proper function of government is to protect the small guy -- the dumb programmer who cannot really imagine being unemployed because this is his first job -- from people and corporations with power, money, and institutional memory.
Now, you may reasonably want to add some qualifying adjectives up there. Neither protection can/should be absolute. But, the point is that the playing field is very much not level, if only because the programmer hasn't ever had one of those contracts before while the employer has. I suspect it's one of those things in life that you don't understand at a gut level until it has happened to you. Kind of like having kids or getting married or buying a house.
Slashdot Mirror
Yeah, the crime of the modern educational system is that it produces people who know all the answers and have no sense of wonder. That "older than dirt" guy probably looks at a computer and only sees a white box.
You should look at a computer and see the thread of execution hopping between kernel routines and pausing at mutexes. You should see the electrons whooshing through the silicon, underneath an overhanging crystalline gate electrode. You should feel the electric field sucking at you: it's almost strong enough to rip electrons out of the SiO2 dielectric. And back up at higher level, those spin locks should be like an amusement park ride: puke your guts out if you go around in one for more than a few microseconds. :)
Yah, we knew stars became red giants. But that's not the right way to look at it.
Yah, yah, yah. Everyone knows that there is climate change that "just happens". But so what? That doesn't prevent humans from changing the climate also.
The AGW story is that human-generated climate change has been fairly small so far. But it is becoming more and more important, as everyone's economy grows and the population grows. Who would argue with that? We have more cars, more computers, more cheap airline flights. India and China have been growing economically at amazing speeds and they have been burning more coal and whatever. Anyone can see that whatever part of climate change is caused by use humans will be rather bigger in the coming century than it has been in the last.
So, it is easy to imagine that we were relatively unimportant in the last century, we are now having effects comparable to natural fluctuations, and in the next century we will be driving big changes.
Throwing out all the data is a reasonable response, but probably over-cautious. There are twice as many of us humans now as there were in 1960, and we are affecting the environment more. So, perhaps trees are affected by recent changes in the environment; and the rules are different after 1960. I don't know: I'm not an expert on trees, but there are tree experts out there. Somebody knows the right answer (or will know it, soon enough).
Ultimately, throwing out useful data can be just as bad as keeping some contaminated data. This isn't a question that has a simple answer. No one is pretending that trees are perfect thermomenters; it's all a matter of comparing the magnitude of the temperature effect to everything else that can affect the growth of a tree. One cannot make that comparison sensibly without doing some serious research on trees. If you haven't done any, perhaps you shouldn't comment.
The thing is, there's plenty of evidence that we humans affect the environment, and we need data to build models so that we can respond intelligently. Suppose, just suppose, that we really do need to reduce CO2 production 80% by 2050. Wouldn't you want to know about it? That magnitude of reduction really won't happen by the natural operation of free markets, not without a carbon tax or similar.
So, we need to have good climate data, as much as possible, and we need to build models. I'm a firm believer in opening up this critical research to make sure it is as accurate as possible, but perfection may not be attainable. Nonetheless, we will need to operate on whatever is the best achievable, rather than demanding impossible perfection. Demanding perfection is just an way of hiding one's head in the sand.
> And read about their naming conventions for files... they were an IT disaster.
Yes? So? The question is, did they manage to run the right scripts on the right files, eventually. Naming conventions just make that easier or harder. You can still get the right answer with bad names, or get the wrong answer with a brilliant naming scheme.
Geez! Programmers think that science is nothing more than programming...
Personally, I am also disturbed by the slapdash approach that seems evident in HARRY_READ_ME, but the point is that slapdash is not the same as wrong. And, of course, bugs do not necessarily imply useless (as anyone who runs software should know quite well).
> Here is a hint. If he says "Trust me" he ain't no scientist he is a salesman/politician.
Not really. Science progresses because we generally trust each other. If you didn't trust other people's results, you'd spend your entire career repeating experiments that someone else has already done correctly. You'd waste your life.
Of course, you trust, but not 100%. You keep an eye open for things that might indicate that someone is wrong or incompetent. But, really, most of the checking happens when you try to use someone's result as a tool. If they are wrong, it usually becomes pretty obvious pretty fast "Hey! all my nails are bending over. I wonder if this is a good hammer..."
No, I do it all the time, and it is the correct thing to do.
A scientific review is not a trap for fraud or a re-analysis of the data. It is not adversarial (well, it is not intrinsically adversarial). The idea is that you are helping the person write a better paper, in addition to deciding whether it is good enough to publish. And you assume that they have described their work accurately.
Fraud gets detected sooner or later when people try to replicate the experiment. And, wrong papers get detected that way also.
Reviews are there to remove (not catch!) any visible errors, to make sure that the logic make sense, to make sure that nothing important was forgotten, and to make sure that the experiment was described completely enough so that someone else could replicate it. That's more than enough work for the poor (unpaid) reviewer.
In civilized countries, justice is temperred with mercy and the knowledge that everyone does something stupid occasionally. I'm not entirely sure that there is a civilized country yet, but I look forward to it.
Well, don't forget that there are good _approximate_ solutions for many NP problems.
NP can be a red herring in the real world: we don't need to solve (for example) the traveling salesman problem _exactly_. So long as we get within 1% or so of the optimum solution, that's good enough for most purposes.
(Of course, we don't have good approximations for all NP problems, only some of them, so it isn't a total red herring.)
Because one of the greatest of all human rights is to pack your bags and go somewhere else.
Visas shouldn't be entirely for the convenience of corporations. They should be for people, too.
As far as I'm concerned, if you can pull your weight, you should be welcome.
The most important thing you lose in an on-line university is the research-teaching connection.
The thing that most people don't realize as students, is that there are many very attractive ideas that are simply wrong. They are as attractive to the professors as they are to the students (professors are people too). And, some fields (like Linguistics, where I have personal knowledge) it is a permanent struggle to avoid the easy, attractive theories. The real experimental data is often much more complex and messy than anyone wants to believe.
If you don't do research yourself, or don't have colleagues who are doing research, it's all too easy to believe whatever is easy to teach. Chomsky, for instance. While the guy deserves credit for starting the field, most of what he said has since been proven wrong (except where he was careful to hedge his statements). But, his theories live on because people want to believe them.
So, if you cut the teaching-research linkage, I fear that we will gradually lose most of our knowledge and settle into the comfortable routine of teaching a curriculum with all the rough edges smoothed away. Efficient, but (largely) wrong.
"Nothing (except my generally honest nature, and a lack of money) was stopping me from hiring a smarter person to write my papers and do my projects, or even sit in classes and take my tests. "
Indeed. And nothing is stopping you from hiring a world-class football player to have a career in your name. Or from hiring a charismatic politician to run for office for you. I agree.
Yet, for some unknown reason, none of these types of identity fraud happen offen enough to cause much trouble. I wonder why.
Yeah, like History is important. Take a good course and think about it a little bit, and the first lesson is that an awful lot of people in the past were simultaneously (a) completely wrong, and (b) completely convinced they were right.
If you apply that knowledge when you look around you, it changes the way you look at the world quite a bit.
Tell me that again in 20 years.
The thing is, you may be the world's expert in the 18-year-old NotSoHeavyD3, but you're not yet the world's expert in the 45-year-old NotSoHeavyD3.
Those two versions of you may have different opinions.
The thing that no one has mentioned here is that universities benefit everyone, not jus their students. Not just their investors (as if they had them!).
It's very simple. Get sick and go to the doctor. Would you rather go to an educated doctor or an uneducated one? So, everyone benefits from the education that the doctor received. Likewise, go up in an elevator. Do you want the building to be designed by a well-educated engineer, or someone who went to a diploma mill? Again, the well-educated engineer is a public good, valuable to everyone.
The same with scientists, programmers, everyone. The point is that life is better when you're surrounded by educated people. So society should pay for that to happen. That's why sensible governments pay for universities out of public money: because the universities benefit the entire public.
Of course, they benefit the students, too. That's true, but it doesn't make the value to the public disappear.
So, if you look at a university as a profit-making enterprise, you'll kill it, and then complain that you can't get well-educated help any more.
And universities do more than just teach, too, but that's another article...
I like to think that my bug reports and feature requests are helpful. Certainly I intend them that way, though I suppose it's easy enough to imagine some maintainers think otherwise.
There isn't a sharp line between feature requests and bug reports. Personally, I take the liberal stance and say that "if it confuses the user, it's a bug", but some differ. Some strange people even write specifications and believe that "if it meets the specifications, it's not a bug." I can't agree with that. If a program misbehaves but meets specifications, that's a bug in the specifications. And, if the specifications are wrong, then you need to change the code anyway, just as much as if the bug were in the code.
Even New Jersey.
I went along to some church (Methodist) adult classes on Genesis once. The preacher did a good job, walking the fine line and not interpreting Genesis too literally. But, some of the people there, my neighbours in suburban NJ were not happy with this. A few people were really asking him and pushing him on the reality of the flood and creation in seven days. And how to squeeze dinosaurs into it.
I was shocked. I ended up going into a little speech on radioactive dating and geology, pointing out that there was a lot of evidence for the age of the Earth, and asking if they really wanted to believe in a God who would plant misleading evidence in the rocks. I don't know if I convinced anyone or not. (Though, I must say, no one seemed particularly upset.)
Scientists outside their field are not experts, of course. But, neither are they to be entirely dismissed. Even outside your field, you know how to think logically, you know about statistics, you know that science is a bit of a messy process sometimes, and you know how to evaluate what other scientists write. And, there may be a bit of overlap from one field to another, too.
So, for instance, I am a non-climate scientist. I read the public stuff. I've gone to google scholar and read some of the techical papers. And, while I'd have to spend a decade to really get up to speed and do some of the work, I can understand generally what people are doing.
When I read, I look for failures in logic. I look for places where they make assumptions that are not supported. I look for obvious violations of the physics I learned in school and the bits and pieces I've picked up from reading other papers. I look for signs that people hold their opinions for political reasons rather than because of their data.
And, by and large, it all looks OK. So, I believe their conclusions. That's how *I* do it. No, I do not simply believe what I was taught in school.
Well, most patents are that way. Most patents are things that are cranked out on demand to satisfy some corporation. Many of mine fall into that category.
But, not all. I'm quite proud of a few of them, and there weren't other people doing the same thing at the same time. And I've seen a few inventions (not mine) that are really quite impressive. So, you are certainly over-generalizing.
As for your assertion that "patents are fundamentally based on the assumption that ... a particular invention is unique to a particular inventor". Well, I don't know where that came from. I'd rather think that patents are based on the assumption that if you give people incentives to invent and to commercialize their inventions, then you'll get more interesting products.
That last bit may or may not be true -- you certainly get more inventions, but whether you end up with more interesting products or not is a good question. One that should be settled by some kind of research and certainly cannot be settled by bare assertion.
But, what if you say "Excuse me, I have an operating audio recorder."? Then, presumably, if the person continues talking, he is implicitly giving consent to being recorded.
Of course, they may get you for interfering with a police officer/TSA employee/whoever, on the grounds that he cannot fulfill his duties without talking. But I'd guess that's not a felony.
... Booked in advance, a train from London to Edinburgh can be as little as £12.
Well, not all of England lives in London, no matter what Londerners think. (Or, maybe you're just wrong...) You can't get from Oxford to Edinburgh for £12. I looked at the National Rail Journey Planner, and it quotes £154 round trips a month in advance. It looks like there may be fares as cheap as £86, but not lower.
I've lived here for 5 years now, and the only place I've gone for £12 or less is Didcot, which is just 10 miles down the road. Well, maybe Banbury, which is perhaps 20 miles up the road...
Yeah, except that rail isn't cheap for passengers. Here in the UK, you can fly to the South of France for the price of a rail ticket to Scotland. (I.e. On rail, it costs about GBP100 = US$160 to go 350 miles.)
If rail is so efficient for passengers (it presumably *is* for bulk freight) why ain't it cheap?
Certainly rail's fuel costs are small, but what about the carbon costs of all those guys standing around in fluorescent yellow vests?
Anyway, for a shared-use RDBMS, it might be moot, because it usually has more than one query to process at any given time - and so it's easy to load-balance all cores just by assigning queries to them.
Ha! Not so! Load balancing like that means the response time is long. Long response times do eventually add up, especially if you are combining data from many database servers, like in a web mash-up.
Ideally, if you could do the queries really fast, and work at only one query at a time with all processors, everyone would be happier.
Yeah. Is it supposed to be cool to work too hard or is it supposed to be stupid? Seems to me like a properly planned and managed launch shouldn't be all that exciting. Or painful, whichever.
It's never like that: no one ever tells you to violate safety rules; it is just made clear that X needs to get done. The rest is left up to the student.
The pressure is often self-applied. Everyone on the academic track knows that you need to publish or perish (sorry!). The thing is, that when you go hunting for your next post-doc or your professorial position, all that matters is results. And there are lots of applicants for permanent positions, so it is crucial to get more results than everyone else. I've gotten more than one letter back from some university apologizing for delays in a job application process because "...we had over 100 applicants" or "...we had over 200 applicants."
In that kind of environment, it is amazing that safety gets any attention at all. But it does, even if perhaps not enough.
And, don't forget that in a research environment, everyone is making up procedures as they go along. Industry has the advantage that you can do something again and again, until you figure out the best way to do it. Researchers often don't have that option. Once you've done something a few times, that's the end of it. Either you graduate or you move into another part of the experiment. Or, the technique becomes obsolete, or it needs to be modified for the needs of some other experiment.
Correct. One of the proper functions of government is to stop people from doing stupid things, like driving 70 in a residential area or selling their future.
Another proper function of government is to protect the small guy -- the dumb programmer who cannot really imagine being unemployed because this is his first job -- from people and corporations with power, money, and institutional memory.
Now, you may reasonably want to add some qualifying adjectives up there. Neither protection can/should be absolute. But, the point is that the playing field is very much not level, if only because the programmer hasn't ever had one of those contracts before while the employer has. I suspect it's one of those things in life that you don't understand at a gut level until it has happened to you. Kind of like having kids or getting married or buying a house.