Or, more likely, the guy on the other end will take your name, phone number, address, and a statement of the complaint. Then they'll reassure you that they'll get right on it, and thanks for calling. Click.
Unless you're a TV celebrity, a Congressman, or at least an appellate judge good luck getting them to do a thing for you.
Re:IT is only one facet of healthcare
on
IT and Health Care
·
· Score: 2, Insightful
Your example is a good one, but after having all these arteries clamped and fixed, how many patients then go on and die because some nurse adminsters the wrong drug - or the drug that the records say is the right drug but that was due to some kind of clerical error?
My concern is that for every miracle life saved there are probably 500 lost or otherwise shortened through the medical meat grinder. Quite a bit of pain and suffering too as patients take needlessly long to recover from less critical problems.
Medicine seems to be optimized to handle these kinds of major trauma scenarios and less optimized to handle some poor guy with sepsis who is about 95% likely to recover with prompt and correct treatment and about 50% likely to die if there is much delay in getting them the care they need, but in the meantime there isn't any blood pooling on the floor.
Re:As someone who has worked on it...
on
IT and Health Care
·
· Score: 1
That is half the problem with looking at the cost of an IT solution.
If it is an IT solution it is going to have to be compliant, since exactly what you're doing is completely on display for the world to see.
If it is a manual solution than you can put one thing on paper, and do another thing in practice. Most auditors aren't going to discover this unless you're just blatantly in violation. If they do discover a violation it is easy to pretend that it is an isolated incident.
So, usually what happens is the cost of a compliant IT solution gets compared to the cost of a manual non-compliant solution. That greatly disadvantages the IT solution since it can't cut corners. I've seen that many times when dealing with regulated IT systems (different industry). To have a fair comparison you need to compare the cost and pain of the IT system to the cost and pain of a truly compliant manual process (which usually doesn't actually exist - though everybody claims it does).
I don't pretend to know all the regulations involved, but that website mentions that such a device is suitable for emergency destruction of top secret data.
In an emergency this probably would be a good tradeoff between security and time - you can't take three weeks to do an "emergency" destruction if your security guards are holding off a regiment of troops looking to capture your data (which I think is the actual scenario envisioned - maybe some paratroops drop in on your roof or something or there are rioters outside looking to break in).
However, I think that if a hard drive truly contained top secret data it would probably need to be almost completely incinerated to be secure - preferably to the point of melting the platters and destroying the memory chips. Top secret data potentially would be of interested to a very determined government - a merely bent hard drive could probably be read just fine with something like a tunnelling electron microscope. Sure, it would take quite a bit of determination, but if you're talking about the detailed designs and source code for an F22 or a nuclear bomb or something like that I'm sure somebody would be willing to go through the trouble. Reading the bits off of a bent hard drive has to be easier than building your own from scratch.
If some part of the business is expensive (usually because it requires following regulations or requires the company to be safe) it gets outsourced. The main qualification for the outsourcer is that they are dirt cheap and that they sign off that they do everything by the book. Then when it turns out that they don't do things by the book they get fired (after making profits for 10 years), and then the contract is put out for bid again and the cheapest supplier is again hired.
Meanwhile, all the outsourcing contractors who actually do things in a reputable manner go out of busienss since they can't compete on price with companies that will happily sign the agreements and then not follow them.
The solution - hold companies responsible for the actions of their outside contractors. Then we'll see actual due diligence. It works this way in at least some industries - particularly anything FDA-regulated. If 30 people go blind because J&J is supplied with defective saline to package their contacts in by no-name-salt-co, then J&J will have products pulled from the market and will need to satisfy lots of regulators before being allowed to put them back on the market. As a result, companies like J&J regularly inspect their suppliers to make sure that their lucrative business isn't shut down.
As others have pointed out, AAs have their limitations. However, this really just points to a need for a few more battery standards for modern electronics.
Rather than everybody who comes out with a device inventing a new battery design, why not invent a few more standard cell sizes with standardized voltages? You could even write up charging specifications for them.
If there is a concern that charging specs would stifle new battery designs, then just specify the voltages and minimum capacities. Then design the physical shape so that any battery will plug into any device, but batteries will be keyed to specific models of chargers so that the charging specs can vary by make/model. That isn't actually hard to do - put a pattern of bumps/grooves on the battery, and matching bumps/grooves in the charger, and then a big empty spot on devices so that any pattern will fit.
I'd think that Sega vs Accolade would essentially apply in this case. The supreme court has ruled that if somebody makes a product that requires the violation of a trademark or copyright in order to make it interoperate with another part, then they cannot enforce their IP rights against those who violate them purely to make devices interoperable.
Courts generally don't like legal loopholes - at least not the supreme court. Sure, you can tie up soembody in court with a clever legal theory that clearly violates the intent of a supreme court decision, but eventually they'll find against you. I suspect that since there have been a few rulings along these lines now that lower courts aren't going to look kindly on playing games with IP law to stick it to consumers.
Hardware is just a drop in the bucket for something like this. If the project needed 100 servers it couldn't cost more than $1-2m - I doubt Wikipedia has 100 servers. Now, it isn't uncommon to see servers quoted at enormous budgets in internal accounting, but that usually isn't the cost of the hardware but another way of passing on the cost of more people to set it up, maintain it, manage the people who do those things, and the boss's cousin who sits around and plays solitaire all day.
I've seen ERP implementations and it just isn't a pretty sight. I really don't get why they need to be THAT expensive. I have worked on moderately large applications and I have no illusions that something like this can be done for $20k or anything like that, but when tens of millions of dollars are changing hands there is a LOT of money going into overhead of some kind.
Constrained power to the majority means that that majority will have less power , and thus , less rights (as power is required to ensure rights are upholded) . So how can that be protecting the rights of the majority ?
Reasonable restraints on majority power help to protect the rights of EVERYBODY (that is all people under the jurisdiction of the Government). That includes the majority, and all miniorities that are represented.
Basic human rights should not be subject to legislative perogative. We are not granted these rights by government. Instead, the people of the nation give the government power to defend their rights.
Suppose a majority of voters in a referendum vote that women shouldn't be allowed to be employed outside of their homes? Would it be right for government to enforce such a decision?
The only requirement to generate EMPs is a nuclear weapon - the bigger the better.
Now, the higher the altitude the more effective it becomes. You could do quite a bit of damage setting one off near New York City, but if you have an ICBM and can detonate one in space above the target they're going to be in for a world of hurt. Maybe you could even argue that since there were no direct casualties that a nuclear response is not justified. Of course, that assumes that you even have a chance to call a press conference before your nation is a parking lot.
The Iranian's best weapon against the US is economic manipulation of oil. They really don't have any other options. They could easily attack shipping, but the moment they fire a missile all internation opposition to US intervention evaporates and you're going to be under massive attack. Forget EMPs/etc - lauch anything based on nuclear weapons on the US and you're going to be turned into a wasteland. Now, if the Chinese just launched one nuke the US might act differently - since they still would have the potential to do far more damange and the US would prefer to avoid escalating to a total launch. However, a massive strike against Iran would stand a good chance of eliminating its entire nuclear force (at least anything capable of reaching the US) - and at most they really could just nuke another city or two. The US is going to look at the North Koreas / etc of the world and send the message that you can posture all you want but if you like ruling over millions of people it is best to not get them all killed.
The Iranians aren't stupid - they'll posture for the crowds but they aren't going to launch a first strike. They have everything to lose and very little to gain. If they were really convinced that it was all for Allah they wouldn't resort to rigging elections to help Allah out. To them religion is just another tool for manipulation.
Uh, can't vouch for everybody everywhere, but at the fortune 500 company I work for, an assistant manager, a manager, a senior manager, an associate director, a senior director, an executive director, and possibly even a VP would probably go by the general term "manager."
Most people in the real world don't get nearly so caught up in titles, at least not until it comes time for bonuses.
Makes sense to me. Sometimes I wonder if there shouldn't be two grades assigned in a class. One would reflect how quickly the student learned, and the other would reflect what they actually learned in the end.
Most grading systems essentially measure how quickly a student learns. They're given a finite amount of instruction and a finite amount of time and then they're assessed.
Most certification tests (in the working world) tend to work the opposite way - they assess current knowledge (and often have some dimension of evaluating actual experience as well). So, if it takes you 10 years to study for an exam and you pass, then you pass and get the same grade as some guy who didn't study at all and passed.
Both attributes are potentially valuable to know. One measures what you can do, and the other measures how quickly you'd be able to pick up something you hadn't been taught before. In theory one could just take an IQ test to figure that out, but I suspect that per-subject grades would better assess somebody's strengths/weaknesses than some overall assessment.
You won't learn much from an easy professor, and three years after you graduate that easy "A" will be meaningless.
Kind of like the rest of your college education.:) It is only needed to get through the interview screening process.
Don't get me wrong - I don't really believe that college doesn't teach anybody anything. However, for the most part college is designed to prepare you to teach college courses the way that you learned it. If you don't plan on teaching college courses for the rest of your life about 75% of everything you do there will be a waste of time.
Agreed. I majored in chemistry. A common high-school or general chemistry assignment is to memorize the atomic weights of the majority of the elements (as well as number and therefore placement). Somehow I managed to escape ever having to do this. The fact is that I don't feel compromised in the slightest - any value that I'm likely to ever need I memorized anyway just out of necessity, and everything else can be obtained from a periodic table (and there is one of those hanging up about every 20 feet in any lab in the world).
I have been helping somebody with math who really struggles at it. However, if you asked her to go back and take a test from the first month of the year, a test that she might have barely passed at the time would be passed with flying colors today. So, what did that test really mean? It just meant that simply being presented with material and tested on it led to her not learning it, but applying it over months of related work led to her learning it. I've seen the same thing in chemistry classes - those who struggle with basic stoichiometry in the first month or two of a course are able to go back and do it trivially by the time they're doing redox formulas and dynamic equilibrium.
I'm not saying that grading and testing doesn't have a place. However, the fact is that people learn material by applying it - and not just in the form of abstract problems.
Yup - the puritain work ethic at play - if it didn't cause you pain then it wasn't really an accomplishment. A minor should indicate that a person has a given skill - not that they did more work. However, colleges that make money based on the number of courses you take might see that differently...
Even 18 years ago I remember a respected high school calculus teacher telling me that he was concerned that math instruction was all wrong - that too much time was simply spent teaching the meachnics of integration. At the time I didn't really appreciate this, but now I certainly do (though I only minored in math).
Like you pointed out - the why matters far more than the how. Sure, some mechanics of integration are helpful to understand such as the idea behind integration by parts, or the chain rule (really the same thing) - the key fundamental principles that also teach you how things work. However, the fact is that a computer can do almost all of these tasks fairly straightforwardly. Additionally, in the "real world" it is rare to actually need algabreic solutions to calculus problems. Most likely you don't start with an equation and end with an equation. More typically you start with a pile of data and end up with a pile of data - numeric integration techniques can blast through these problems in microsecond.
Let's educate students on the principles and the resources out there, and teach them to solve problems in the most effective manner possible. Being able to take the derivative of a nested set of 35 functions is about as much of an accomplishment as being able to manually recalculate a spreadsheet. It is a lot of work, but it really doesn't require that much intelligence once you know the rules.
Mozilla only requires this if you redistribute the branding. Of course, we source-based distributions only distribute pristine sources and let the user (automatically) do all the patching so these kinds of restrictions really only apply to binary-based distros.:)
Actually, such a law is perfectly compatible with private health care.
However, it is completely incompatible with voluntary private health insurance. It is only compatible with voluntary public health insurance if taxpayers don't mind paying through the nose as a result of it.
And the problem isn't even the law - it is the existance of the technology. Suppose I take a genetic test and find out that I'm unlikely to suffer from any chronic health problems until I'm 70, and then I'm likely to develop diabetes. (Never mind that this is oversimplified - the argument applies equally well if you just get probabilities/etc out of it.)
Now, an insurer can either discriminate against me on the basis of this information or not. If they can discriminate then I get really cheap insurance until I'm 70, and then I'm dropped. Of course, age 70 is when I actually need the insurance. Not very fair, so the public won't stand for this.
Ok, so this law prevents me from being discriminated against. Now I am free to get some catastrophic care insurance package for very little money with a $10k deductible - this covers me if I get hit by a car and need trauma surgery. Then, when I hit age 69, I take out a very expensive and comprehensive policy that will pay even the tiniest expense for diabetes. If everybody does this, the insurance industry goes bankrupt - all their customers will be sick people since the healthy will stay away. Either that, or they have to charge huge premiums so that nobody can get insurance anyway (which the public won't stand for).
The only way you can have insurance in a world with this kind of knowledge is if everybody is forced into the risk pool whether they want to be there or not. That is essentially public health insurance, although there are ways it could be semi-privatized. However, for it to work the government needs to rule with an iron hand and make sure everybody pays in.
Actually, I'm a bit of an aviation enthusiast, and I'm fairly aware of these sorts of things. I'm not trying to say that we don't do a good job trying to turn human pilots into machines by drilling them in 400k different scenarios of things that could go wrong. I'm just suggesting that at least in principle a computer ought to be good at this sort of thing if properly designed.
It's when you reply to a string of earlier messages and place your reply on top, so that whoever reads will have no idea of the context.
What's top posting?
Let's all go into comp.lang.c and start top posting to threads. They LOVE IT when you do that.
Should I do this instead? No, no, no. When trolling a programming forum, make damn sure you post in HTML-formatted text. If you can figure out how to include the tag, you could probably hear their heads explode from halfway around the world.
If not, your best bet is to include code snippets in multiple languages, each using different tab-stops for indentation. Make frequent references in how this would be much easier in Java, unless posting to comp.lang.java, then post on how C# fixed it and is really Java done right.
Oh, and make sure to quote a multi-page question fully and answer only with one sentence. They love that.
Finally, big sigs with ASCII art and geek code blocks. The bigger the better. True masters have sigs bigger than their actual post.
This does make me think about whether there is a better solution. Here is what we have now:
1. Usenet. Pros - NNTP works great in that it separates content from presentation. Cons - a lot of stuff gets sent all over the place "just in case", spam is a big problem, and archival is redundant (why store an artile in 50,000 places for six months?).
2. Web Forums. Pros - better control of spam and stewardship is clear (site owner controls site), purely on-demand transmission of data. Cons - you're stuck with whatever presentation model the site offers, and you end up having accounts on 500 different sites.
3. Mailing Lists. Pros - mostly on-demand transmission of data, stewardship is clear, separation of content and presentation. Con - needs to be coupled with an archival solution which isn't tightly integrated, administration is per-list but no need for "accounts."
The best of all worlds would have these properties:
1. Spam is controlled. This probably requires some kind of stewardship or at least a web-of-trust of some kind. 2. Data only goes where it is needed. An ISP shouldn't need a full news feed for their 10 customers who follow 5 newsgroups. 3. Data is archived. It need not be archived equally everywhere. 4. Content and presentation are separated. Users use the client of their choosing, which might include a web interface but need not do so.
To be honest, a mailing list combined with a newsreader-like client might be the best of all worlds. Gmane may be a very good example of how to do this. I'd really like to see something more peer-to-peer or distributed, but not at the ISP level like usenet.
I actually got annoyed trying to follow one of my high-volume lists that doesn't allow gmane to archive them. I set up my own innd and am piping the mailing list into my own newsgroup. It seems like this is a mostly-solved problem that just needs a little refinement.
The real mess is a lack of package management on Windows.
On virtually any linux distro I can type one command and have the system check for security updates and provide me a list of all packages that require security updates. Another command will apply those updates. If I'm REALLY brave I can just put it in cron and have it just email me what its doing after the fact (not always wise - some linux distros sometimes break booting with core package upgrades). A different variation on the same process could apply non-security updates as well. Distros like debian actually backport security patches so that you can have very safe updates.
On windows the OS itself is fairly well updated if you configure it correctly. However, the 40 bazillion other pieces of software I use are a mixed bag. Some will auto-update, but using their own update programs with their own configurations and their own update policies. Many don't auto-update at all, but if you look really hard you might find a website (or if you're really lucky an email list) where updates get posted. I'm sure my windows box right now has 5-7 services all running in the background that are just looking for updates to various programs.
Windows really needs a package manager. It could even support installs off of CD, but the installer is a standard component of the OS, and the OS manages updates. The installer could even be extensible (installer creates an enviornment to install into, then program-specific installer does all kinds of magic and dumps files into that environment, then OS deploys files and registry keys and permissions appropriately). Virtually any linux distro would be a vast improvement, and I think there is room for even further improvement.
I'm talking about emergencies where the pilot's life is truly at risk - not ones induced by a flight instructor by pulling back a throttle handle or in a simulator. If there isn't a significant chance that everybody on the plane could die, it isn't really an emergency. There are all kinds of psychological factors that come into play with human pilots in these situations - and there is no way to reproduce them without actually risking the life of the crew. Obviously we can't do that in training.
Simulator time will always be fininte. There is no question that drilling can get pilots to a high level of proficiency, but there is always a limit. At most a human could spend 44562 hours per year in a simulator (24*7*365.25) - a computer program could actually spend more than that (since it can be replicated and tested in parallel).
There is no question that flying is very safe now. However, it pushes the limits of what is possible with humans in command. I suspect that if we really want to see improvements we'll need to take the humans out of the loop. There would also need to be fundamental improvements in maintenance/etc - I'm not sure that pilot error is even the leading cause of crashes (though better pilot reaction might somewhat improve the outcome of equipment failures). I'm not quite sure what to do about bird strikes other than wonder why LGA's runways point in the direction of a bird sanctuary (seems like lots of airports have them nearby - guess the land is cheap).
Also, if the issue was the American legal system, then Airbus would come up with the same solution, being as many of their customers are US airlines
There is no question that Airbus does business in the US. There is no question that somebody could sue them in a US court, and get a delcaration of judgment. Then what? Does airbus have any assets in US jurisdiction that somebody could actually take possession of? With the amount of protectionism in the heavy aircraft industry I doubt the EU is going to be cooperative in efforts to collect.
A delcaration of judgment is only a piece of paper if you can't actually get a hold of any assets...
Slashdot Mirror
Or, more likely, the guy on the other end will take your name, phone number, address, and a statement of the complaint. Then they'll reassure you that they'll get right on it, and thanks for calling. Click.
Unless you're a TV celebrity, a Congressman, or at least an appellate judge good luck getting them to do a thing for you.
Your example is a good one, but after having all these arteries clamped and fixed, how many patients then go on and die because some nurse adminsters the wrong drug - or the drug that the records say is the right drug but that was due to some kind of clerical error?
My concern is that for every miracle life saved there are probably 500 lost or otherwise shortened through the medical meat grinder. Quite a bit of pain and suffering too as patients take needlessly long to recover from less critical problems.
Medicine seems to be optimized to handle these kinds of major trauma scenarios and less optimized to handle some poor guy with sepsis who is about 95% likely to recover with prompt and correct treatment and about 50% likely to die if there is much delay in getting them the care they need, but in the meantime there isn't any blood pooling on the floor.
That is half the problem with looking at the cost of an IT solution.
If it is an IT solution it is going to have to be compliant, since exactly what you're doing is completely on display for the world to see.
If it is a manual solution than you can put one thing on paper, and do another thing in practice. Most auditors aren't going to discover this unless you're just blatantly in violation. If they do discover a violation it is easy to pretend that it is an isolated incident.
So, usually what happens is the cost of a compliant IT solution gets compared to the cost of a manual non-compliant solution. That greatly disadvantages the IT solution since it can't cut corners. I've seen that many times when dealing with regulated IT systems (different industry). To have a fair comparison you need to compare the cost and pain of the IT system to the cost and pain of a truly compliant manual process (which usually doesn't actually exist - though everybody claims it does).
I don't pretend to know all the regulations involved, but that website mentions that such a device is suitable for emergency destruction of top secret data.
In an emergency this probably would be a good tradeoff between security and time - you can't take three weeks to do an "emergency" destruction if your security guards are holding off a regiment of troops looking to capture your data (which I think is the actual scenario envisioned - maybe some paratroops drop in on your roof or something or there are rioters outside looking to break in).
However, I think that if a hard drive truly contained top secret data it would probably need to be almost completely incinerated to be secure - preferably to the point of melting the platters and destroying the memory chips. Top secret data potentially would be of interested to a very determined government - a merely bent hard drive could probably be read just fine with something like a tunnelling electron microscope. Sure, it would take quite a bit of determination, but if you're talking about the detailed designs and source code for an F22 or a nuclear bomb or something like that I'm sure somebody would be willing to go through the trouble. Reading the bits off of a bent hard drive has to be easier than building your own from scratch.
Gotta love modern business.
If some part of the business is expensive (usually because it requires following regulations or requires the company to be safe) it gets outsourced. The main qualification for the outsourcer is that they are dirt cheap and that they sign off that they do everything by the book. Then when it turns out that they don't do things by the book they get fired (after making profits for 10 years), and then the contract is put out for bid again and the cheapest supplier is again hired.
Meanwhile, all the outsourcing contractors who actually do things in a reputable manner go out of busienss since they can't compete on price with companies that will happily sign the agreements and then not follow them.
The solution - hold companies responsible for the actions of their outside contractors. Then we'll see actual due diligence. It works this way in at least some industries - particularly anything FDA-regulated. If 30 people go blind because J&J is supplied with defective saline to package their contacts in by no-name-salt-co, then J&J will have products pulled from the market and will need to satisfy lots of regulators before being allowed to put them back on the market. As a result, companies like J&J regularly inspect their suppliers to make sure that their lucrative business isn't shut down.
As others have pointed out, AAs have their limitations. However, this really just points to a need for a few more battery standards for modern electronics.
Rather than everybody who comes out with a device inventing a new battery design, why not invent a few more standard cell sizes with standardized voltages? You could even write up charging specifications for them.
If there is a concern that charging specs would stifle new battery designs, then just specify the voltages and minimum capacities. Then design the physical shape so that any battery will plug into any device, but batteries will be keyed to specific models of chargers so that the charging specs can vary by make/model. That isn't actually hard to do - put a pattern of bumps/grooves on the battery, and matching bumps/grooves in the charger, and then a big empty spot on devices so that any pattern will fit.
I'd think that Sega vs Accolade would essentially apply in this case. The supreme court has ruled that if somebody makes a product that requires the violation of a trademark or copyright in order to make it interoperate with another part, then they cannot enforce their IP rights against those who violate them purely to make devices interoperable.
Courts generally don't like legal loopholes - at least not the supreme court. Sure, you can tie up soembody in court with a clever legal theory that clearly violates the intent of a supreme court decision, but eventually they'll find against you. I suspect that since there have been a few rulings along these lines now that lower courts aren't going to look kindly on playing games with IP law to stick it to consumers.
Hardware is just a drop in the bucket for something like this. If the project needed 100 servers it couldn't cost more than $1-2m - I doubt Wikipedia has 100 servers. Now, it isn't uncommon to see servers quoted at enormous budgets in internal accounting, but that usually isn't the cost of the hardware but another way of passing on the cost of more people to set it up, maintain it, manage the people who do those things, and the boss's cousin who sits around and plays solitaire all day.
I've seen ERP implementations and it just isn't a pretty sight. I really don't get why they need to be THAT expensive. I have worked on moderately large applications and I have no illusions that something like this can be done for $20k or anything like that, but when tens of millions of dollars are changing hands there is a LOT of money going into overhead of some kind.
Constrained power to the majority means that that majority will have less power , and thus , less rights (as power is required to ensure rights are upholded) . So how can that be protecting the rights of the majority ?
Reasonable restraints on majority power help to protect the rights of EVERYBODY (that is all people under the jurisdiction of the Government). That includes the majority, and all miniorities that are represented.
Basic human rights should not be subject to legislative perogative. We are not granted these rights by government. Instead, the people of the nation give the government power to defend their rights.
Suppose a majority of voters in a referendum vote that women shouldn't be allowed to be employed outside of their homes? Would it be right for government to enforce such a decision?
The only requirement to generate EMPs is a nuclear weapon - the bigger the better.
Now, the higher the altitude the more effective it becomes. You could do quite a bit of damage setting one off near New York City, but if you have an ICBM and can detonate one in space above the target they're going to be in for a world of hurt. Maybe you could even argue that since there were no direct casualties that a nuclear response is not justified. Of course, that assumes that you even have a chance to call a press conference before your nation is a parking lot.
The Iranian's best weapon against the US is economic manipulation of oil. They really don't have any other options. They could easily attack shipping, but the moment they fire a missile all internation opposition to US intervention evaporates and you're going to be under massive attack. Forget EMPs/etc - lauch anything based on nuclear weapons on the US and you're going to be turned into a wasteland. Now, if the Chinese just launched one nuke the US might act differently - since they still would have the potential to do far more damange and the US would prefer to avoid escalating to a total launch. However, a massive strike against Iran would stand a good chance of eliminating its entire nuclear force (at least anything capable of reaching the US) - and at most they really could just nuke another city or two. The US is going to look at the North Koreas / etc of the world and send the message that you can posture all you want but if you like ruling over millions of people it is best to not get them all killed.
The Iranians aren't stupid - they'll posture for the crowds but they aren't going to launch a first strike. They have everything to lose and very little to gain. If they were really convinced that it was all for Allah they wouldn't resort to rigging elections to help Allah out. To them religion is just another tool for manipulation.
Uh, can't vouch for everybody everywhere, but at the fortune 500 company I work for, an assistant manager, a manager, a senior manager, an associate director, a senior director, an executive director, and possibly even a VP would probably go by the general term "manager."
Most people in the real world don't get nearly so caught up in titles, at least not until it comes time for bonuses.
Makes sense to me. Sometimes I wonder if there shouldn't be two grades assigned in a class. One would reflect how quickly the student learned, and the other would reflect what they actually learned in the end.
Most grading systems essentially measure how quickly a student learns. They're given a finite amount of instruction and a finite amount of time and then they're assessed.
Most certification tests (in the working world) tend to work the opposite way - they assess current knowledge (and often have some dimension of evaluating actual experience as well). So, if it takes you 10 years to study for an exam and you pass, then you pass and get the same grade as some guy who didn't study at all and passed.
Both attributes are potentially valuable to know. One measures what you can do, and the other measures how quickly you'd be able to pick up something you hadn't been taught before. In theory one could just take an IQ test to figure that out, but I suspect that per-subject grades would better assess somebody's strengths/weaknesses than some overall assessment.
You won't learn much from an easy professor, and three years after you graduate that easy "A" will be meaningless.
Kind of like the rest of your college education. :) It is only needed to get through the interview screening process.
Don't get me wrong - I don't really believe that college doesn't teach anybody anything. However, for the most part college is designed to prepare you to teach college courses the way that you learned it. If you don't plan on teaching college courses for the rest of your life about 75% of everything you do there will be a waste of time.
Agreed. I majored in chemistry. A common high-school or general chemistry assignment is to memorize the atomic weights of the majority of the elements (as well as number and therefore placement). Somehow I managed to escape ever having to do this. The fact is that I don't feel compromised in the slightest - any value that I'm likely to ever need I memorized anyway just out of necessity, and everything else can be obtained from a periodic table (and there is one of those hanging up about every 20 feet in any lab in the world).
I have been helping somebody with math who really struggles at it. However, if you asked her to go back and take a test from the first month of the year, a test that she might have barely passed at the time would be passed with flying colors today. So, what did that test really mean? It just meant that simply being presented with material and tested on it led to her not learning it, but applying it over months of related work led to her learning it. I've seen the same thing in chemistry classes - those who struggle with basic stoichiometry in the first month or two of a course are able to go back and do it trivially by the time they're doing redox formulas and dynamic equilibrium.
I'm not saying that grading and testing doesn't have a place. However, the fact is that people learn material by applying it - and not just in the form of abstract problems.
Yup - the puritain work ethic at play - if it didn't cause you pain then it wasn't really an accomplishment. A minor should indicate that a person has a given skill - not that they did more work. However, colleges that make money based on the number of courses you take might see that differently...
Even 18 years ago I remember a respected high school calculus teacher telling me that he was concerned that math instruction was all wrong - that too much time was simply spent teaching the meachnics of integration. At the time I didn't really appreciate this, but now I certainly do (though I only minored in math).
Like you pointed out - the why matters far more than the how. Sure, some mechanics of integration are helpful to understand such as the idea behind integration by parts, or the chain rule (really the same thing) - the key fundamental principles that also teach you how things work. However, the fact is that a computer can do almost all of these tasks fairly straightforwardly. Additionally, in the "real world" it is rare to actually need algabreic solutions to calculus problems. Most likely you don't start with an equation and end with an equation. More typically you start with a pile of data and end up with a pile of data - numeric integration techniques can blast through these problems in microsecond.
Let's educate students on the principles and the resources out there, and teach them to solve problems in the most effective manner possible. Being able to take the derivative of a nested set of 35 functions is about as much of an accomplishment as being able to manually recalculate a spreadsheet. It is a lot of work, but it really doesn't require that much intelligence once you know the rules.
Mozilla only requires this if you redistribute the branding. Of course, we source-based distributions only distribute pristine sources and let the user (automatically) do all the patching so these kinds of restrictions really only apply to binary-based distros. :)
Actually, such a law is perfectly compatible with private health care.
However, it is completely incompatible with voluntary private health insurance. It is only compatible with voluntary public health insurance if taxpayers don't mind paying through the nose as a result of it.
And the problem isn't even the law - it is the existance of the technology. Suppose I take a genetic test and find out that I'm unlikely to suffer from any chronic health problems until I'm 70, and then I'm likely to develop diabetes. (Never mind that this is oversimplified - the argument applies equally well if you just get probabilities/etc out of it.)
Now, an insurer can either discriminate against me on the basis of this information or not. If they can discriminate then I get really cheap insurance until I'm 70, and then I'm dropped. Of course, age 70 is when I actually need the insurance. Not very fair, so the public won't stand for this.
Ok, so this law prevents me from being discriminated against. Now I am free to get some catastrophic care insurance package for very little money with a $10k deductible - this covers me if I get hit by a car and need trauma surgery. Then, when I hit age 69, I take out a very expensive and comprehensive policy that will pay even the tiniest expense for diabetes. If everybody does this, the insurance industry goes bankrupt - all their customers will be sick people since the healthy will stay away. Either that, or they have to charge huge premiums so that nobody can get insurance anyway (which the public won't stand for).
The only way you can have insurance in a world with this kind of knowledge is if everybody is forced into the risk pool whether they want to be there or not. That is essentially public health insurance, although there are ways it could be semi-privatized. However, for it to work the government needs to rule with an iron hand and make sure everybody pays in.
Actually, I'm a bit of an aviation enthusiast, and I'm fairly aware of these sorts of things. I'm not trying to say that we don't do a good job trying to turn human pilots into machines by drilling them in 400k different scenarios of things that could go wrong. I'm just suggesting that at least in principle a computer ought to be good at this sort of thing if properly designed.
J00 R 50 RI9H7, d00D!!!!
--- Original Message ----
Bunch:
Of:
Headers:
It's when you reply to a string of earlier messages and place your reply on top, so that whoever reads will have no idea of the context.
What's top posting?
Let's all go into comp.lang.c and start top posting to threads. They LOVE IT when you do that.
Should I do this instead?
No, no, no. When trolling a programming forum, make damn sure you post in HTML-formatted text. If you can figure out how to include the tag, you could probably hear their heads explode from halfway around the world.
If not, your best bet is to include code snippets in multiple languages, each using different tab-stops for indentation. Make frequent references in how this would be much easier in Java, unless posting to comp.lang.java, then post on how C# fixed it and is really Java done right.
Oh, and make sure to quote a multi-page question fully and answer only with one sentence. They love that.
Finally, big sigs with ASCII art and geek code blocks. The bigger the better. True masters have sigs bigger than their actual post.
This does make me think about whether there is a better solution. Here is what we have now:
1. Usenet. Pros - NNTP works great in that it separates content from presentation. Cons - a lot of stuff gets sent all over the place "just in case", spam is a big problem, and archival is redundant (why store an artile in 50,000 places for six months?).
2. Web Forums. Pros - better control of spam and stewardship is clear (site owner controls site), purely on-demand transmission of data. Cons - you're stuck with whatever presentation model the site offers, and you end up having accounts on 500 different sites.
3. Mailing Lists. Pros - mostly on-demand transmission of data, stewardship is clear, separation of content and presentation. Con - needs to be coupled with an archival solution which isn't tightly integrated, administration is per-list but no need for "accounts."
The best of all worlds would have these properties:
1. Spam is controlled. This probably requires some kind of stewardship or at least a web-of-trust of some kind.
2. Data only goes where it is needed. An ISP shouldn't need a full news feed for their 10 customers who follow 5 newsgroups.
3. Data is archived. It need not be archived equally everywhere.
4. Content and presentation are separated. Users use the client of their choosing, which might include a web interface but need not do so.
To be honest, a mailing list combined with a newsreader-like client might be the best of all worlds. Gmane may be a very good example of how to do this. I'd really like to see something more peer-to-peer or distributed, but not at the ISP level like usenet.
I actually got annoyed trying to follow one of my high-volume lists that doesn't allow gmane to archive them. I set up my own innd and am piping the mailing list into my own newsgroup. It seems like this is a mostly-solved problem that just needs a little refinement.
Never thought of it that way... Finger: the first social networking application. :)
The real mess is a lack of package management on Windows.
On virtually any linux distro I can type one command and have the system check for security updates and provide me a list of all packages that require security updates. Another command will apply those updates. If I'm REALLY brave I can just put it in cron and have it just email me what its doing after the fact (not always wise - some linux distros sometimes break booting with core package upgrades). A different variation on the same process could apply non-security updates as well. Distros like debian actually backport security patches so that you can have very safe updates.
On windows the OS itself is fairly well updated if you configure it correctly. However, the 40 bazillion other pieces of software I use are a mixed bag. Some will auto-update, but using their own update programs with their own configurations and their own update policies. Many don't auto-update at all, but if you look really hard you might find a website (or if you're really lucky an email list) where updates get posted. I'm sure my windows box right now has 5-7 services all running in the background that are just looking for updates to various programs.
Windows really needs a package manager. It could even support installs off of CD, but the installer is a standard component of the OS, and the OS manages updates. The installer could even be extensible (installer creates an enviornment to install into, then program-specific installer does all kinds of magic and dumps files into that environment, then OS deploys files and registry keys and permissions appropriately). Virtually any linux distro would be a vast improvement, and I think there is room for even further improvement.
I'm talking about emergencies where the pilot's life is truly at risk - not ones induced by a flight instructor by pulling back a throttle handle or in a simulator. If there isn't a significant chance that everybody on the plane could die, it isn't really an emergency. There are all kinds of psychological factors that come into play with human pilots in these situations - and there is no way to reproduce them without actually risking the life of the crew. Obviously we can't do that in training.
Simulator time will always be fininte. There is no question that drilling can get pilots to a high level of proficiency, but there is always a limit. At most a human could spend 44562 hours per year in a simulator (24*7*365.25) - a computer program could actually spend more than that (since it can be replicated and tested in parallel).
There is no question that flying is very safe now. However, it pushes the limits of what is possible with humans in command. I suspect that if we really want to see improvements we'll need to take the humans out of the loop. There would also need to be fundamental improvements in maintenance/etc - I'm not sure that pilot error is even the leading cause of crashes (though better pilot reaction might somewhat improve the outcome of equipment failures). I'm not quite sure what to do about bird strikes other than wonder why LGA's runways point in the direction of a bird sanctuary (seems like lots of airports have them nearby - guess the land is cheap).
Also, if the issue was the American legal system, then Airbus would come up with the same solution, being as many of their customers are US airlines
There is no question that Airbus does business in the US. There is no question that somebody could sue them in a US court, and get a delcaration of judgment. Then what? Does airbus have any assets in US jurisdiction that somebody could actually take possession of? With the amount of protectionism in the heavy aircraft industry I doubt the EU is going to be cooperative in efforts to collect.
A delcaration of judgment is only a piece of paper if you can't actually get a hold of any assets...