I think you're giving too much credit to the TFA. While it's true that you can essentially interchange between R and Chi-squared with some easy statistics, that doesn't mean that this guy was thinking that.
Everything in TFA indicates he just saw R as being low, thus declared it insignificant. Later on in the TFA, he declares R > 0.10 significant (complete with a citation, which I cannot imagine to be even remotely correct). This assertion makes no sense, as R does not take sample-size into account, where every other significance test on the planet does for obvious reasons.
Honestly, you give the author way too much credit. You're right, he got the conclusion right. But not because of any actual sound statistical logic.
p.s. I really think your guy's argument down below is just a matter of semantics. Statisticians will argue that correlation will never imply causation -- other disciplines (especially the ones in which just about everything can be controlled, ie physics) will go a bit farther, essentially saying "well everything else is controlled for, so we must be proving causation", and then they just accept being wrong as much as their alpha allows.
I've read the Economist religiously for several years. I firmly believe it to be probably the best magazine/newspaper out there. I subscribe despite their sub price being approximately 5 times that of Times or Newsweek or any other magazine out there.
That said, this is the most stupid idea I have ever heard out of them. They actually will compensate you, with a rocking 6-mo web-subscrption to economist.com (street value: roughly $50).
Perhaps the Economist should actually talk to their economists, and ask them what 'Incentive Compatability' means. $50 for a new revolutionary business idea surely isn't incentive compatible. If I were the Economist, I'd be terribly embarassed about this.
While you are correct as asessing it as an oligpoly, you miss one important part -- how it became an oligopoly.
The rights to that spectrum were carefully auctioned off by the FCC in a semi-public auction. The companies who currently own these rights (Sprint, T-Mobile, etc) paid literally hundreds of millions of dollars for their spectrum property rights.
Thus any re-opening up of the spectrum will easily cost in the hundreds of millions of dollars. All this for skype. This seems quite rediculous, there is already an option for new carriers, who are able to buy excess capacity from the current spectrum owners. This is how the companies such as Disney Mobile, Mobile ESPN (before it got axed), Amped, etc, works. This makes a lot more sense, these spectrum owners weren't just gifted with the spectrum, they won it in an auction, presumably because would be the most efficient operators (reflected in their highest bidding prices).
The thing is, this has been the issue for us for literally years, but over time, it has actaully gotten worse, instead of better. Back in 1.2.13 (or whenever you started), dual booting was annoying, but somewhat acceptable -- to get to a unix prompt, that was basically your only choice.
The creation of OS X really makes the dual-boot idea seem quite archaic. If you can get all of the power of unix shells combined with a platform that actaully runs real apps? Suddenly the dual-boot linux solution seems really really lame.
Personally if I were you, I'd look at a mac instead of you can afford it. My friends and I first started using linux like 11 years ago, and over the past two years we've all switched to a mac. For exactly the reasons you describe.
While it is artificial there is a really good reason on why FM licensing is there.
There's only a fixed amount of spectrum out there, and the licensing allows it to be allocated in a fairly efficient manner. If you do not do this, then anybody can blowup anybody else's transmission, and you're left with no reasonable programming (or cellphones for that matter, or satellite tv) at all.
Now while the barriers to creating a ratio station are quite expensive, the fact is that just about everyone would rather have some mediocre programming (what we have currently) over unabashed chaos that would happen otherwise. There are open bands of spectrum where you can do whatever you want with it, so it's not like it's a massive government conspiracy to keep the man down.
I totally agree with your assessment here, but I don't see why anyone on the low end would want it either.
1600s and 2600s are just dirt cheap now, even with WICs. You can't build a comparable x86 (runs on flash, 1u, low power consumption) for the price. And saving 200 bucks by building a crappy x86 really doesn't make any sense when you're paying $500+/mo for your actual T.
I think everyone agrees that you won't get the performance of an asic-based router, so the only thing they have going for them is price. But with sonoma cards going for $500 each, that's like twice the price of a WIC already.
So where's the benefit? Slower than a Cisco, more expensive than a Cisco, worse support than a Cisco. "It's open source so you can do anything you want"? I don't know about yours, but my IT staff doesn't have the time or the money or the manpower to devote an army of C/asm programmers to go off writing network code when we could just buy an IOS off the shelf that does the same thing.
December 9th 2005 issue of The Economist goes into this very issue. It states specifically why solar power is currently not a viable resource, and uses these exact price comparisons to mention why solar power is only useful in residential contexts.
Unless in specific circumstances, it's rather doubtful that solar panels will actaully lower your overall bill.
Solar power costs something like 18-22 cents/kwh once you amortize the cost of the panels over it's entire lifetime, etc. Commercial power is generally less than this, maxing out around 17cents/kwh in the pacific northwest. In the midwest commercial power costs like 7 cents/kwh.
Solar power is currently [i]extremely[/i] expensive compared to other energy sources. It's main penetration currently is in the residential sector as water heaters, because a) this doesn't require a lot of power, thus not a lot of space (ie on a roof), b) residential electricty rates are much higher than commercial, thus making it almost viable, and c) large tax breaks/credits makes it viable. You have none of these advantages when talking about a commercial data center (though depending on where you live, you *may* have a small tax credit).
I find it extremely doubtful that solar will be cost-effective.
The easiest way to do it is to just not go wireless in the first place. Once you get rid of this criteria, wiring a lecture hall with the cable for a connection really isn't that difficult.
The subjects in which clickers are mainly used (physics, engineering), everyone already has a graphical calculator, and they're generally of either HP or TI variety. Thus you only have two (ok maybe 3, TI-85 line is quite different from 83's), but then you have no mandatory extra cost to the student, since everyone in these disciplines has a suitable calculator already.
No hardware issues, no support issues, you basically just wire a minijack to every seat, and you're set.
I know the physics program at uiuc has experimented with this about 5 years ago, prior to them becoming the new fad. You probably want to check with their physics education group http://www.physics.uiuc.edu/research/per/ about the plus/minuses with it. IIRC they eventaully went with commercial clickers -- I'm pretty sure there's a good reason why, you probably should check with them.
Unlike the majority of these posts that you're going to read from/., these guys actually did the experimentation, are intellectually capable of rolling their own project had they desired, and made a decision based on their experiences. They're very nice people and will probably share their experiences with you, particularly prof. Mats Selen, who afaik headed the project.
You're right that it's not that special, but i think it's more obvious when you realize what's the current technology out there.
It already is possible to create a 'stealth' type of shield by manipulating thin films (width exactly equal to 1/2 the wavelength of a particular light) to create destructive interference so nothing is reflected. Use multiple coats of differing thickness, and you can wipe out quite a bit of the spectrum. The theory behind this is well known and nothing more than undergrad physics. Fab obviously is harder, but the US defense sector is quite able to do it at this current time.
Obviously this technique is notsogood for say, visual invisibility, but works well against radar etc.
The economics are where? Governments are totally ill-equipped to deal with a non-public-good such as wireless internet. They deal with public goods such as parks, where if one guy goes to the park, it doens't really affect anyone else that goes to the park. This is totally not the same as internet, where you have large issues over shared pipes.
There are HUGE economics of scale in the internet business. Your first network engineer costs maybe 60-70k/yr, yet he can handle maybe up to a load of 10-20k users. You lose the 24-7 oversight since it's only 1 guy, and you reek of a small-town business. Not to mention tech support issues. The economies of scale don't really happen signficantly until you're big enough to get a NAP port, and have at least a regional NOC, and maybe an call-center for tech support. Outside of superlarge metropolitan areas, the population base is just not big enough.
Lastly, it should be clear that the economics don't exist. If they had, *companies* would be moving and and attempting this already. Yet they only do this in perhaps the top10 us cities by population density. And notice we haven't heard anything about their profitability. I wonder why.
If a company (with presumably huge increasing returns to scale, and higher efficiency than gov't), can't turn a profit, I'm totally not sure how a gov't agency could provide comparable service, and be a savings to the consumer. I wouldn't be surprised if the cost were 2-3x as much, just spread over the entire city population instead of just the users. Which really seems totally out of whack since you know the amt of ppl that will use it (in cities not named Seattle, Redmond, Champaign, San Jose) will be sub-5%.
If such a system had existed. It's a fairly well known fact in academia and science circles that this defense system is nowhere near the capability of actaully shooting a real missile down -- it can barely even shoot down fake ones.
Robert Park (of American Physics Society) has been covering this for a while, and there are many well-documented problems with the system, of which some he mentions in this newsletter.
Funny thing is, if you survey departments outside of mathematics, particularly with regards to the UIUC Netmath program, and their line of mathematica based courses, there's a fairly wide consensus that these courses are crap and the students do not learn anything.
This has been rather well-known in departments like CS and Physics as UIUC. They have been instructing their students for years to expressly avoid these courses, as you don't actaully learn the material in them. As someone who has taken vector calc in mathematica and gotten an A in it, I totally agree with them.
I'm going to slightly disagree. While I agree most of mathematica is waaay too complicated for the level we're looking at, all the other suggestions (other than pen-and-paper) are by far worse.
You don't want to learn some proprietary language that does algebra, then need to learn another one or two come university, just because the one you used in HS is so esoteric and limited no one uses it. Because of this, if you really really have to go the computer route, i suggest using one of Matlab/Mathematica/Maple. It's commonly used in universities, so that you actually are learning a useful skill while not learning real mathematics.
My experience in mathematics is that no one cares which of Matlab/Mathematica/Maple that you use, as long as it gets the job done, and you know how to use it. Very few of the other suggestions are acutally powerful enough that will get you far, even octave (which seems to be the most advanced open-source thing there is) is missing many of the libraries that matlab has.
Let's take the example of a large multinational corporation, and a single individual who pretty much is a peon on the corporate structure.
This single person works because he wants to. He makes himself better off by working, because from it, he gains a wage. He must enjoy it to some extent, or at least he must enjoy it more than sitting at home and not working at all. If he preferred to stay home and not get paid, then he would choose that instead. It doesn't matter that the company makes way more money than he does, he is in some manner content (although possibly unhappy) since because he gets paid, he can pay the bills. In summary, working is better than sitting on your ass at home not getting paid.
Contrast to the IPD game. Each of the sacrificial IPD players could have made themselves better off, but instead chose not to. Depending on the payoff tables, it is most likely that the mean payoff (if they had split the prize 80 ways) would be less than if each person had play tit-for-tat. So with the real world example, this would be analgous for someone to go into work for a big multinational corporation for no wage at all.
Unfortunately the cheating made it defeat the entire purpose of IPD in the first place. The idea behind it (and game theory in general), is that a person will choose actions that will maximize his profit/utility/happiness/etc. With this said, IPD is useful because it can be drawn into games such as repeated-period pricing in a Cournot duopoly - where each firm maximizes their profit.
SO fastforward to what we have today, a strategy where we create 500bazillion smaller firms all for the purpose of going bankraupt, so that one parent firm makes all the money. Huh? Right, that doesn't make any sense.
The *correct* way to think about it, is that if you're going to have said cooperation, the coalitions that cooperate will split the prizes between them, such that if this team had 80 entries, their end utility (ie end profit) would be 1/80th of the one program that won. In this case, then drawn back into the real-world example of a firm, you can see why this isn't a valid strategy, due to the start-up costs of making a bazillion fake firms who all lose money, etc.
So yes, this team won. Yes, it was a pretty ingenius way of winning. But no, it has little if any use to actual game theory.
True that the corn subsidies are BS, but I find your reasoning questionable. The fact of the matter is that farmers are a large voting block, and taking away any subsudies is a sure way to get a good number of people voting against you. The fact that they hold the first caucus is really rather trivial.
The guy you're replying to is close, but not exactly on the money. The latest ethanol subsidies were used mainly as pork to get the democrats in affected states to pass the huge energy bill a few months ago.
The reason Iowa is involved is becasue the senator(s) from there actaully opposed the bill, and wouldn't have voted for it if not for those freebie subsidies for their state.
patents/copyrights encourage investment/innovation because they allow the person who created/invented the concept to reap the profits from it. No profits = no incentive to innovate = no new science.
Reasons the US is falling behind and will fall more in the future is:
large federal spending cuts to the non-health sciences (total spending is up, but that's only because bio/health spending went up so much for 'anti-terrorism' research.)
New immigration redtape preventing a ton of smart people from other countries come here for grad school. The new patriot act etc is seriously turning off some people to the U.S., and keeping many others who want to come from coming, just because of their last name.
A total failure of education reform (ie no-child left behind) which everyone that's not in the executive branch agrees will totally fail in its goal to increase education quality.
The promotion of voodoo research that satisfies political gains rather than credible research that could actaully help society. And the discrediting of credible research by various religious groups, just because they think its wrong.
Somewhat related to the previous one, the general ignorance of many southern religious groups to what science is about. Hint: religion and science can co-exist, I suggest any of Paul Davies books as reading material on this. As the southern christians are a large voting block (and basically the reason for the last election results), much policy is catered towards them.
One thing I find funny is that there were many people that used due diligence to report flaws to the vendor, and said vendor doesn't bother to fix it.
Not to single you out, but it seems a really lot of people bring up this (generally true) generalization, but fail to realize that in most cases Cisco and the like are not one of these types of vendors, definitely not for their upper-level model lines.
My general experience with cisco is if there's some sort of bug/exploit/problem in one of their current IOS's, i call them up and generally they'll get a new ios build to me before the week is over. Cisco doesn't bullshit about serious problems, even if the problem is unique to my situation.
Well with physics/mathematics degrees, you certainly need to sell yourself. There aren't job openings for 'Physicist' or 'Mathematician' outside of Academia/National Labs, however when people are hiring for research/analysis jobs, while not explicitly saying it, they certainly will consider qualified physicists, especially if you give them a good reason why to.
If you can get past the resumes into the interview phase, you normally can blow most everyone else out of the water using your analytical skills.
And i know this from experience, having a degree in physics, while spending several years in networks/system adminstration. My current job? I'm in grad school... in economics. Solely because of my analytical background. A Physics/Math type major certainly gives you the most versitility of any college degree. You just need to sell yourself.
Im a Physics/Economics double major graduating senior, going to gradschool in Economics next year...
I would advise not going for exclusively a physics major, if you're unsure whether that's what you really want to do. Out of all the physics majors i know, very few are in there to actaully do physics research as a career, or many of us start with that intention, then realize how difficult/strange/boring/uninteresting/etc that we think it really is. We have a very large amount of double majors, (Physics/math, physics/finance, physics/chem, some premeds even), where we use the physics courses to teach us how to think, not necessarily for the physics itself.
Unless you really really want to know/study stuff like the boundary conditions of the fields of a conductor in an oscillating magnetic field, I would stay away from physics as a pure major; but if you wanted to do something like a M.S in Physics w/ a PhD in Economics, your analytic skills for something like IndustrialOrganization or GameTheory (maybe even theoretical econometrics) would be awesome.
Put a switch on each floor (unmanaged 16-port switches are less than $80)
No. You want a really spiffy switch. It needs to a) be able to do mac-port mapping, b) be able to remotely enable-disable ports, and c) support rmon/snmp. Maybe you dont need c) if you have netflow configured/running correctly, but a) and b) will save you tons of time (and therefore labor costs) longrun by doing these two things. Unless you want to walk to the place at 3am because some dumbass got rooted and you need to go unplug him because he's pingflooding efnet (it's going to happen, trust me.)
Good idea, but your numbers are wrong. T1s btwn 25 buildings = 24 * (linecost of t1 between them). which is more then $3k already. Frame relay can get it a bit cheaper, but you're not gonna get it under the 1500 you need for profitability.
The idea works if you have large-occupancy buildings, ie 100+, however doing internet connectivity to an apartment correctly is very hard. There's always going to be like 30 people with their kazaa [or thing like it] on at all times, that's going to make the connection unbearable. Then there's going to be the irc dumbass who is getting synflooded for insulting some 1337 h4x0ring group. Then there's.... on and on. Apt. connectivity requires a lot of babying that other clients dont, and that means much lower margins... and a lot of the time it just isnt worth it.
I spent over a month setting up the architure to actually do an apt building correctly. Accounting, bandwidth monitoring, priority queueing, rate limiting, etc. This required a rather large infrastructure upgrade. The cost of that plus my labor costs will put us at break-even with the proposal in two-years time. While not necessarily a bad investment, it's a lot worse than your typical insurance company that just wants email and a webpage with their t1.
Slashdot Mirror
I think you're giving too much credit to the TFA. While it's true that you can essentially interchange between R and Chi-squared with some easy statistics, that doesn't mean that this guy was thinking that.
Everything in TFA indicates he just saw R as being low, thus declared it insignificant. Later on in the TFA, he declares R > 0.10 significant (complete with a citation, which I cannot imagine to be even remotely correct). This assertion makes no sense, as R does not take sample-size into account, where every other significance test on the planet does for obvious reasons.
Honestly, you give the author way too much credit. You're right, he got the conclusion right. But not because of any actual sound statistical logic.
p.s. I really think your guy's argument down below is just a matter of semantics. Statisticians will argue that correlation will never imply causation -- other disciplines (especially the ones in which just about everything can be controlled, ie physics) will go a bit farther, essentially saying "well everything else is controlled for, so we must be proving causation", and then they just accept being wrong as much as their alpha allows.
No kidding.
I've read the Economist religiously for several years. I firmly believe it to be probably the best magazine/newspaper out there. I subscribe despite their sub price being approximately 5 times that of Times or Newsweek or any other magazine out there.
That said, this is the most stupid idea I have ever heard out of them. They actually will compensate you, with a rocking 6-mo web-subscrption to economist.com (street value: roughly $50).
Perhaps the Economist should actually talk to their economists, and ask them what 'Incentive Compatability' means. $50 for a new revolutionary business idea surely isn't incentive compatible. If I were the Economist, I'd be terribly embarassed about this.
While you are correct as asessing it as an oligpoly, you miss one important part -- how it became an oligopoly.
The rights to that spectrum were carefully auctioned off by the FCC in a semi-public auction. The companies who currently own these rights (Sprint, T-Mobile, etc) paid literally hundreds of millions of dollars for their spectrum property rights.
Thus any re-opening up of the spectrum will easily cost in the hundreds of millions of dollars. All this for skype. This seems quite rediculous, there is already an option for new carriers, who are able to buy excess capacity from the current spectrum owners. This is how the companies such as Disney Mobile, Mobile ESPN (before it got axed), Amped, etc, works. This makes a lot more sense, these spectrum owners weren't just gifted with the spectrum, they won it in an auction, presumably because would be the most efficient operators (reflected in their highest bidding prices).
The thing is, this has been the issue for us for literally years, but over time, it has actaully gotten worse, instead of better. Back in 1.2.13 (or whenever you started), dual booting was annoying, but somewhat acceptable -- to get to a unix prompt, that was basically your only choice.
The creation of OS X really makes the dual-boot idea seem quite archaic. If you can get all of the power of unix shells combined with a platform that actaully runs real apps? Suddenly the dual-boot linux solution seems really really lame.
Personally if I were you, I'd look at a mac instead of you can afford it. My friends and I first started using linux like 11 years ago, and over the past two years we've all switched to a mac. For exactly the reasons you describe.
While it is artificial there is a really good reason on why FM licensing is there.
There's only a fixed amount of spectrum out there, and the licensing allows it to be allocated in a fairly efficient manner. If you do not do this, then anybody can blowup anybody else's transmission, and you're left with no reasonable programming (or cellphones for that matter, or satellite tv) at all.
Now while the barriers to creating a ratio station are quite expensive, the fact is that just about everyone would rather have some mediocre programming (what we have currently) over unabashed chaos that would happen otherwise. There are open bands of spectrum where you can do whatever you want with it, so it's not like it's a massive government conspiracy to keep the man down.
I totally agree with your assessment here, but I don't see why anyone on the low end would want it either.
1600s and 2600s are just dirt cheap now, even with WICs. You can't build a comparable x86 (runs on flash, 1u, low power consumption) for the price. And saving 200 bucks by building a crappy x86 really doesn't make any sense when you're paying $500+/mo for your actual T.
I think everyone agrees that you won't get the performance of an asic-based router, so the only thing they have going for them is price. But with sonoma cards going for $500 each, that's like twice the price of a WIC already.
So where's the benefit? Slower than a Cisco, more expensive than a Cisco, worse support than a Cisco. "It's open source so you can do anything you want"? I don't know about yours, but my IT staff doesn't have the time or the money or the manpower to devote an army of C/asm programmers to go off writing network code when we could just buy an IOS off the shelf that does the same thing.
December 9th 2005 issue of The Economist goes into this very issue. It states specifically why solar power is currently not a viable resource, and uses these exact price comparisons to mention why solar power is only useful in residential contexts.
Unfortunately the article is not online.
Unless in specific circumstances, it's rather doubtful that solar panels will actaully lower your overall bill.
Solar power costs something like 18-22 cents/kwh once you amortize the cost of the panels over it's entire lifetime, etc. Commercial power is generally less than this, maxing out around 17cents/kwh in the pacific northwest. In the midwest commercial power costs like 7 cents/kwh.
Solar power is currently [i]extremely[/i] expensive compared to other energy sources. It's main penetration currently is in the residential sector as water heaters, because a) this doesn't require a lot of power, thus not a lot of space (ie on a roof), b) residential electricty rates are much higher than commercial, thus making it almost viable, and c) large tax breaks/credits makes it viable. You have none of these advantages when talking about a commercial data center (though depending on where you live, you *may* have a small tax credit).
I find it extremely doubtful that solar will be cost-effective.
The easiest way to do it is to just not go wireless in the first place. Once you get rid of this criteria, wiring a lecture hall with the cable for a connection really isn't that difficult.
/., these guys actually did the experimentation, are intellectually capable of rolling their own project had they desired, and made a decision based on their experiences. They're very nice people and will probably share their experiences with you, particularly prof. Mats Selen, who afaik headed the project.
The subjects in which clickers are mainly used (physics, engineering), everyone already has a graphical calculator, and they're generally of either HP or TI variety. Thus you only have two (ok maybe 3, TI-85 line is quite different from 83's), but then you have no mandatory extra cost to the student, since everyone in these disciplines has a suitable calculator already.
No hardware issues, no support issues, you basically just wire a minijack to every seat, and you're set.
I know the physics program at uiuc has experimented with this about 5 years ago, prior to them becoming the new fad. You probably want to check with their physics education group http://www.physics.uiuc.edu/research/per/
about the plus/minuses with it. IIRC they eventaully went with commercial clickers -- I'm pretty sure there's a good reason why, you probably should check with them.
Unlike the majority of these posts that you're going to read from
Except that xxx.lanl.gov is ALSO known as lanl.arxiv.org (and has been for the past uh, 4 years or so), which gets around any regex filtering.
And if the ip is banned, you could just use www.arxiv.org, which is the current real site, lanl just hosts a mirror.
But it's funner to invent anecdotal problems that just don't exist, i'm sure.
Which also is totally not a symptom of DNS timeouts either. You need a response from a webserver to get a 404.
The article just seems poorly written, I wouldn't go out and assume that telstra just decided to throw 500 new dns servers at it.
You're right that it's not that special, but i think it's more obvious when you realize what's the current technology out there.
It already is possible to create a 'stealth' type of shield by manipulating thin films (width exactly equal to 1/2 the wavelength of a particular light) to create destructive interference so nothing is reflected. Use multiple coats of differing thickness, and you can wipe out quite a bit of the spectrum. The theory behind this is well known and nothing more than undergrad physics. Fab obviously is harder, but the US defense sector is quite able to do it at this current time.
Obviously this technique is notsogood for say, visual invisibility, but works well against radar etc.
The economics are where? Governments are totally ill-equipped to deal with a non-public-good such as wireless internet. They deal with public goods such as parks, where if one guy goes to the park, it doens't really affect anyone else that goes to the park. This is totally not the same as internet, where you have large issues over shared pipes.
There are HUGE economics of scale in the internet business. Your first network engineer costs maybe 60-70k/yr, yet he can handle maybe up to a load of 10-20k users. You lose the 24-7 oversight since it's only 1 guy, and you reek of a small-town business. Not to mention tech support issues. The economies of scale don't really happen signficantly until you're big enough to get a NAP port, and have at least a regional NOC, and maybe an call-center for tech support. Outside of superlarge metropolitan areas, the population base is just not big enough.
Lastly, it should be clear that the economics don't exist. If they had, *companies* would be moving and and attempting this already. Yet they only do this in perhaps the top10 us cities by population density. And notice we haven't heard anything about their profitability. I wonder why.
If a company (with presumably huge increasing returns to scale, and higher efficiency than gov't), can't turn a profit, I'm totally not sure how a gov't agency could provide comparable service, and be a savings to the consumer. I wouldn't be surprised if the cost were 2-3x as much, just spread over the entire city population instead of just the users. Which really seems totally out of whack since you know the amt of ppl that will use it (in cities not named Seattle, Redmond, Champaign, San Jose) will be sub-5%.
If such a system had existed. It's a fairly well known fact in academia and science circles that this defense system is nowhere near the capability of actaully shooting a real missile down -- it can barely even shoot down fake ones.
Robert Park (of American Physics Society) has been covering this for a while, and there are many well-documented problems with the system, of which some he mentions in this newsletter.
Funny thing is, if you survey departments outside of mathematics, particularly with regards to the UIUC Netmath program, and their line of mathematica based courses, there's a fairly wide consensus that these courses are crap and the students do not learn anything.
This has been rather well-known in departments like CS and Physics as UIUC. They have been instructing their students for years to expressly avoid these courses, as you don't actaully learn the material in them. As someone who has taken vector calc in mathematica and gotten an A in it, I totally agree with them.
I'm going to slightly disagree. While I agree most of mathematica is waaay too complicated for the level we're looking at, all the other suggestions (other than pen-and-paper) are by far worse.
You don't want to learn some proprietary language that does algebra, then need to learn another one or two come university, just because the one you used in HS is so esoteric and limited no one uses it. Because of this, if you really really have to go the computer route, i suggest using one of Matlab/Mathematica/Maple. It's commonly used in universities, so that you actually are learning a useful skill while not learning real mathematics.
My experience in mathematics is that no one cares which of Matlab/Mathematica/Maple that you use, as long as it gets the job done, and you know how to use it. Very few of the other suggestions are acutally powerful enough that will get you far, even octave (which seems to be the most advanced open-source thing there is) is missing many of the libraries that matlab has.
Let's take the example of a large multinational corporation, and a single individual who pretty much is a peon on the corporate structure.
This single person works because he wants to. He makes himself better off by working, because from it, he gains a wage. He must enjoy it to some extent, or at least he must enjoy it more than sitting at home and not working at all. If he preferred to stay home and not get paid, then he would choose that instead. It doesn't matter that the company makes way more money than he does, he is in some manner content (although possibly unhappy) since because he gets paid, he can pay the bills. In summary, working is better than sitting on your ass at home not getting paid.
Contrast to the IPD game. Each of the sacrificial IPD players could have made themselves better off, but instead chose not to. Depending on the payoff tables, it is most likely that the mean payoff (if they had split the prize 80 ways) would be less than if each person had play tit-for-tat. So with the real world example, this would be analgous for someone to go into work for a big multinational corporation for no wage at all.
Unfortunately the cheating made it defeat the entire purpose of IPD in the first place. The idea behind it (and game theory in general), is that a person will choose actions that will maximize his profit/utility/happiness/etc. With this said, IPD is useful because it can be drawn into games such as repeated-period pricing in a Cournot duopoly - where each firm maximizes their profit.
SO fastforward to what we have today, a strategy where we create 500bazillion smaller firms all for the purpose of going bankraupt, so that one parent firm makes all the money. Huh? Right, that doesn't make any sense.
The *correct* way to think about it, is that if you're going to have said cooperation, the coalitions that cooperate will split the prizes between them, such that if this team had 80 entries, their end utility (ie end profit) would be 1/80th of the one program that won. In this case, then drawn back into the real-world example of a firm, you can see why this isn't a valid strategy, due to the start-up costs of making a bazillion fake firms who all lose money, etc.
So yes, this team won. Yes, it was a pretty ingenius way of winning. But no, it has little if any use to actual game theory.
The guy you're replying to is close, but not exactly on the money. The latest ethanol subsidies were used mainly as pork to get the democrats in affected states to pass the huge energy bill a few months ago.
The reason Iowa is involved is becasue the senator(s) from there actaully opposed the bill, and wouldn't have voted for it if not for those freebie subsidies for their state.
patents/copyrights encourage investment/innovation because they allow the person who created/invented the concept to reap the profits from it. No profits = no incentive to innovate = no new science.
Reasons the US is falling behind and will fall more in the future is:
My general experience with cisco is if there's some sort of bug/exploit/problem in one of their current IOS's, i call them up and generally they'll get a new ios build to me before the week is over. Cisco doesn't bullshit about serious problems, even if the problem is unique to my situation.
Well with physics/mathematics degrees, you certainly need to sell yourself. There aren't job openings for 'Physicist' or 'Mathematician' outside of Academia/National Labs, however when people are hiring for research/analysis jobs, while not explicitly saying it, they certainly will consider qualified physicists, especially if you give them a good reason why to.
If you can get past the resumes into the interview phase, you normally can blow most everyone else out of the water using your analytical skills.
And i know this from experience, having a degree in physics, while spending several years in networks/system adminstration. My current job? I'm in grad school... in economics. Solely because of my analytical background. A Physics/Math type major certainly gives you the most versitility of any college degree. You just need to sell yourself.
Im a Physics/Economics double major graduating senior, going to gradschool in Economics next year...
I would advise not going for exclusively a physics major, if you're unsure whether that's what you really want to do. Out of all the physics majors i know, very few are in there to actaully do physics research as a career, or many of us start with that intention, then realize how difficult/strange/boring/uninteresting/etc that we think it really is. We have a very large amount of double majors, (Physics/math, physics/finance, physics/chem, some premeds even), where we use the physics courses to teach us how to think, not necessarily for the physics itself.
Unless you really really want to know/study stuff like the boundary conditions of the fields of a conductor in an oscillating magnetic field, I would stay away from physics as a pure major; but if you wanted to do something like a M.S in Physics w/ a PhD in Economics, your analytic skills for something like IndustrialOrganization or GameTheory (maybe even theoretical econometrics) would be awesome.
No. You want a really spiffy switch. It needs to a) be able to do mac-port mapping, b) be able to remotely enable-disable ports, and c) support rmon/snmp. Maybe you dont need c) if you have netflow configured/running correctly, but a) and b) will save you tons of time (and therefore labor costs) longrun by doing these two things. Unless you want to walk to the place at 3am because some dumbass got rooted and you need to go unplug him because he's pingflooding efnet (it's going to happen, trust me.)
Good idea, but your numbers are wrong. T1s btwn 25 buildings = 24 * (linecost of t1 between them). which is more then $3k already. Frame relay can get it a bit cheaper, but you're not gonna get it under the 1500 you need for profitability.
.... on and on. Apt. connectivity requires a lot of babying that other clients dont, and that means much lower margins... and a lot of the time it just isnt worth it.
The idea works if you have large-occupancy buildings, ie 100+, however doing internet connectivity to an apartment correctly is very hard. There's always going to be like 30 people with their kazaa [or thing like it] on at all times, that's going to make the connection unbearable. Then there's going to be the irc dumbass who is getting synflooded for insulting some 1337 h4x0ring group. Then there's
I spent over a month setting up the architure to actually do an apt building correctly. Accounting, bandwidth monitoring, priority queueing, rate limiting, etc. This required a rather large infrastructure upgrade. The cost of that plus my labor costs will put us at break-even with the proposal in two-years time. While not necessarily a bad investment, it's a lot worse than your typical insurance company that just wants email and a webpage with their t1.