Well, according to a recent planet money episode small business LOVE the sales tax since many collect it but under-report it. Amazon couldn't get away with it, but the local mom&pop restaurant probably does it all the time and it gives them a few percent leg up on big companies that have to report it.
Rich people save money (which just leads to bubbles), and poor people spend almost all of their money. Sales taxes tend to hit the poor. The reason the Forbes 100 are so wealthy is that they don't spend most of their money - indeed it is hard to even spend that kind of money unless you do it wholesale on stuff like jets or whatever.
If anything I'd like to see significant income tax increases on incomes over 100-200k or so, and all capital gains would be taxed at the highest possible rate, with no exceptions.
I think that savings are the thing that actually is messing up the economy - it creates an economy that is about nothing more than increasing the size of large bank accounts. Ordinary people get almost all of their money in wages, and that should be taxed the least as it creates a real economy. Somehow the economy did fine back when the rich were paying 75% or whatever.
I'm not sure I'd compare just portage and apt to really get a sense of the difference. Gentoo is fairly different from other distros in a few regards:
1. Tends to stay close to upstream - very vanilla. Granted, there are quite a few distros like this, but many of the big ones tend to tweak everything.
2, VERY flexible. You can run any version of anything with just about any version of anything else. You can run stable versions of everything, but then run ONE package off of the upstream git repository updating it 3x/day if you want. That ONE package can even be glibc, assuming you don't mind recompiling half your system 3x/day (but if you're a glibc developer maybe that hassle is worth being able to test out a change on a real system). No binary distro can support this to any significant degree, since the library dependencies become unmanageable.
3. USE flags are very useful when you are doing something atypical. I was just chatting with somebody the other day and he was running a CNC router controller on ubuntu (it uses a realtime kernel). The router software pulls in X11 and even OpenOffice as dependencies somehow. Now, this could just be bad packaging, but Gentoo gives you the ability to set USE=-X and then a package that could support X will not link to it, and you can more easily get rid of X. This makes Gentoo very well suited to niche/embedded projects.
4. Rolling releases mean that your software tends to stay up-to-date - change is more gradual (but constant). Of course, you can always limit yourself to only installing security bugs for a period of time, as long as it isn't too long as Gentoo does not indefinitely maintain upgrade paths (months are usually fine, years are usually not).
As you can see from the replies strong personalities are one of the things we are dealing with. I think this is somewhat due to the nature of the distro - it appeals to very-technically-minded people, and often these people can have VERY strong opinions on how things should be done. While issues flare up from time to time I'd say things have been relatively under control for a few years now - there was a big flare-up back around 2008 or so. People sometimes get fed up and quit, although from my perspective often when they do so their contributions don't really stop - they just start their own projects outside the distro and then somebody else brings the improvements back in (sometimes people just need their own space to work in, and results argue for themselves).
No IPO is going to punish a company for screwing employees.
Read the article - the CEO says, well, sure it is a lousy deal for the employees but it is what is best for the company.
Duh - getting people to work for years and then not paying them for that work is ALWAYS in the interest of the company as long as they can get away with it. Shareholders don't care about employees - they care about EPS, and when you revoke shares you increase EPS.
Does it really matter if the price of a stock is $30.15 or $30.22? What real-world-based valuation system has the ability to tell what the value of future earnings is within a couple of cents per share?
Nothing wrong with measuring prices in dollars and cents, but this sounds like arguing over how many millimeters it is from the sun to alpha centauri.
Sometimes I wonder if the solution is to just expand social security and the associated payroll taxes/etc so that employers don't need to offer a pension at all.
The problem with pensions is that employers have incentive to put as little money as possible into them while promising as much money as possible out of them. Promises get employees to stick around for less pay, and underfunding means spending less on compensation as well - either way the company wins. Pension managers typically have incentives to maximize today's value of the fund, but employees really care about the value 30 years from today.
The whole thing seems like a big scheme and I really wonder when this will just become the next bailout target.
Well, I'd say that system-hosing updates used to be more common in the past than they are now, but occasionally they still come up (usually with more warning these days, but not always).
I think that Gentoo (and everything else) will be much better off when btrfs becomes stable. With btrfs you can just snapshot everything inexpensively, run your update, and if you're hosed you can just switch back to your snapshot. I know when I update Gentoo VMs I make liberal use of snapshots.
Well, don't completely discount being able to set -march to something other than i686 and use -ftree-parallelize-loops or whatever.
Sure, this will break the odd package, but it is easy enough to just fall back to simple flags for a few of them.
Re:Marine version tripped up the whole program
on
The F-35 Story
·
· Score: 1
I agree that having the amphibious assault capability is good. My point wasn't that we haven't done it because Marines aren't useful - rather we haven't done it because amphibious assaults tend to result in a lot of dead bodies. If we ever did do one then chances are we'd do all we can to keep those bodies to a minimum.
Well, the Navy and Air Force not being interested is a political issue - if this is the best solution then you make them get interested if you want the capability.
WWII was a bit of an unusual situation in that the enemy had a serious navy that had to be contended with. The fleet carriers were busy trying to control the entire theater and didn't have time to spend on one beachhead once they finished escorting the invasion fleet. Plus there were other options since short-field aircraft were more practical when that term described almost everything flying.
If the F35 did work I agree it might have value. It just seems like it would make more sense to just dedicate an airframe to this kind of role rather than making something that replaced both the harrier and the F18 at the same time.
I suspect some of the budget issue is the fact that international buyers are going to be disproportionately interested in the Marine role, since nobody else has carriers like the US. The Naval version would be useless to anybody but the US Navy. The F16 replacement would be broadly usable.
I do agree that when it comes to the 3rd world we shouldn't sacrifice the 80% solution for the sake of the 99% solution that nobody can afford.
Actually, that applies in general to medicine - sometimes the "do no harm" philosophy holds us back a little. In the present climate I'm surprised we even have pain-killers, since in terms of outcomes they don't cure any disease and yet they still can have side-effects.
I agree that the German system costs more in time than in money, and in truth a typical college student tends to have an abundance of the former, and they have some incentive to think twice before signing up.
I really was pointing my finger more to the US system where we promise everybody that they can be an astronaut, and frequently we don't even fail them out, but we collect four years of very expensive tuition from them and then tell them they can work in fast food.
Yes and no. You probably need solvents to extract it. Modern techniques use very refined solvents (we like our drugs to be clean), and those which are best suited to the job. If you're willing to deal with lots of waste/etc you could probably find some way to make it using less ideal materials.
If you just want to eat raw mold or whatever then that is pretty easy to manage, but of course allergies/etc will be even more of a problem than they already are.
I think the real issue is that the website seems to be very "green" in nature, and chemical manufacture just doesn't fit in all that well with that, although the fact that they are refining metals also stands out.
Yeah, but chemicals are so, ugh, non-appealing to the kinds of people who send money who fund initiatives like these. Ecology is nice and green. Antibiotics are tolerable even though they're made by evil profiteering pharmaceutical companies, but I suspect that every time somebody looked into them they kept finding these things called "solvents" used in their manufacture and decided that it wouldn't do for fundraising.
I'm not sure what the 3D printer would even be used for - just about everything in that list would probably use almost exclusively metal parts. Is there anything you can do with a 3D printer that you can't do with a CNC mill?
I have no issues with the work that they are doing, but again I think that making a list of "50" machines is more about fundraising (hey, we only need $4M to revolutionize the world). Otherwise, not have 51 or 52?
I think the issue is that the list seems rather arbitrary. They have fancy single-purpose stuff like a machine to extract aluminum from clay, but just a generic induction furnace to handle every other metal that exists.
No society with ONLY these 50 machines and only token imports would be able to exist with what most people would call "modern comforts." Sure, you might be able to make a few of them, but your CNC mill isn't going to work too well when a transistor in your power supply dies. You couldn't even fashion a vacuum tube replacement with that equipment list. You might be able to fashion a crude aluminum electrolytic capacitor if you're willing to wrap it by hand, although you'd have no way to test it to see if it is a suitable replacement for one that dies short of plugging it in and crossing your fingers.
And your tribal elder who manages your machine shop will be 30 years old, since anytime somebody cuts their finger on a piece of metal they face a real risk of death without antibiotics.
That isn't to say that these guys aren't doing useful work, but there will be plenty left to do after 50 machines are done.
I don't think anybody is suggesting that they aren't doing useful work. However, claiming that it only takes 50 tools to create a civilization with modern comforts really isn't true. No doubt those tools are very useful in getting there, but it takes more. For starters, it takes more than those 50 tools to produce the materials necessary to maintain them.
Again, these are 50 nice useful tools, but the number 50 seems a bit arbitrary and I think that it was chosen so that sponsors can be pitched with "hey, just give us a few million dollars and you can turn a bunch of huts into Greenwich Village." Sure, they don't come out and say it, but I think that is the impression the intro gives.
You are of course welcome to disagree. And I don't think that money sent to them is poorly spent, just that this isn't the magical band-aid that will save the 3rd world.
Uh, without drugs you can die from an infected stubbed toe. Believe it or not disease isn't exclusively God's punishment on people who eat too much. Also, you'll find that a good number of villages do have access to basic pharmaceuticals, and a little goes a long way.
I think the gripe is that their website says "The Global Village Construction Set (GVCS) is an open technological platform that allows for the easy fabrication of the 50 different Industrial Machines that it takes to build a small civilization with modern comforts."
Now, I'm all for open sourcing useful stuff like everything on the list. However, I'm under no illusions that it only takes 50 machines to run a society.
I also think the 3D printer is there more to please the hobbyist crowd than anything else. They're great for prototyping, but not really for manufacturing.
And does it strike anybody else as odd that they have about 15 different computer-controlled machines on the list and yet no computer? A CPU might be difficult to manufacture using their pelletizer.
Re:Marine version tripped up the whole program
on
The F-35 Story
·
· Score: 1
Well, you can always stick Marine pilots in FA-18s or whatever, so the training in close air support isn't really a reason to buy a particular plane but just a matter of military politics.
Nobody is going to be fighting a campaign the size of the WWII pacific theater with our current Navy, I suspect that if we care enough about territory that we're willing to lose a few thousand marines in a day in an opposed landing then chances are we'll have every carrier in the fleet backing them up. Once you've landed in a modern war chances are those forces are going to stick around for a long time. Now, in WWII when one island had a secure beachhead the Navy would sail off to start working on the next island or whatever, but island hopping isn't really something I see happening in the future, and I can't think of a lot of situations where you'd otherwise need to have a bunch of little fixed airstrips without carriers nearby.
Maybe the solution is to just give the Marines their own plane or something, if it really is a need. I'm not sure it really even needs to be VTOL so much as STOL, and that can just be something with big unswept wings that is subsonic. You're not going to be using aircraft like this for air superiority.
Re:Marine version tripped up the whole program
on
The F-35 Story
·
· Score: 1
Why does the carrier need to be 500 miles away?
We haven't even done a serious amphibious assault since WWII, though I understand that a few more recent conflicts involved some landings (probably all lightly opposed). If you really were to face an opposed landing, wouldn't you want those carriers covering you? Otherwise those nice close-air-support aircraft that the marines fly are going to be chopped to pieces by air superiority fighters flown by the enemy.
I just am not understanding the threat model here. If you're going up against barbarians with RPGs and spears then helicopters should do the trick, but chances are we'll still be dropping JDAMs on them. If you're going up against China in a war over Taiwan or something then frankly you need everything you have and you're in for quite a fight. I don't see where having a VTOL-capable jet is going to make much difference compared to what you'd get if you just spent that money on Naval or Air Force gear.
Gotta love bureaucracy, where even the secretary of defense is powerless to make a program happen.
The solution isn't to give the bureaucrats what they want, but to tell them that if they don't want to do their jobs properly somebody else will.
Re:Marine infantry says that ...
on
The F-35 Story
·
· Score: 1
I need a new Porche. Trust me, I'm in a much better place to see it in my driveway than you are...
Re:Marine version tripped up the whole program
on
The F-35 Story
·
· Score: 1
Why not just have the Navy fill this niche? Why do the planes need to launch from a particular class of ship? Why do they need to land on improvised fields? Just what kind of war could having this particular plane help win that would otherwise be lost?
The fact that a computer is running XP says very little about its capabilities. I can install XP on a brand new top-of-the-line computer. About the only thing it won't do is address more than 2GB of RAM per process (or whatever the 32-bit limit is, assuming we're talking about 32-bit XP). For most jobs XP works just fine, and it uses a lot less RAM than newer OSes. These days the ability to address more RAM is more important than wasting less, but then you have to factor in software compatibility - if your fancy 3D printer or whatever doesn't have Vista+ drivers then you're stuck on XP.
As for why computer science departments don't build their own computers... Most computers are commodity items - you don't need your typical workstation to be cutting-edge. I imagine that most computer science departments that own supercomputers DO build those themselves, that is if the design of the supercomputer itself was what they were pushing the envelope on. If they're only concerned with algorithms they might even buy those from somebody.
When you buy 1000 workstations from Dell you get business-class workstations. That means that they all have the same model hard drives/etc, and that Dell will still sell you an exact replacement for your power supply three years from now. If you test a security patch or whatever on one of them, you can be pretty sure you won't get one or two that mysteriously don't boot. Basically it is just the corporate IT model.
This is no different from an astronomy department buying a telescope out of a catalog. If they're looking to do some kind of cutting edge optical design improvement they'll end up building the thing (no doubt contracting the critical parts as custom jobs). If you are doing research that just involves making some brightness measurements on a bunch of stars or something and they're not 4 billion light years away then it probably makes more sense to just buy something for a few thousand dollars or whatever. Sure, lots of cutting-edge stuff happens on the HST or whatever, but there are also a lot of things that can be learned from commodity equipment that doesn't require you to book your time six months in advance. Disclaimer, I'm not an astronomer.
Bottom line is that any organization has to figure out which elements of what it is doing need to be world class, and which elements need to be good enough. At an undergraduate level I'd think I'd be more concerned with my professor having time to answer my questions and help me out than whether he spends six hours a day mass-assembling PCs.
Would it not make sense to weed out students with some kind of test that doesn't involve a 6-9 months of time and $20k+ in expense?
I agree that the root of the problem is that people try to study things they just aren't cut out for. However, there are better solutions to that than dragging them through months of classes at considerable expense.
Slashdot Mirror
Well, according to a recent planet money episode small business LOVE the sales tax since many collect it but under-report it. Amazon couldn't get away with it, but the local mom&pop restaurant probably does it all the time and it gives them a few percent leg up on big companies that have to report it.
Rich people save money (which just leads to bubbles), and poor people spend almost all of their money. Sales taxes tend to hit the poor. The reason the Forbes 100 are so wealthy is that they don't spend most of their money - indeed it is hard to even spend that kind of money unless you do it wholesale on stuff like jets or whatever.
If anything I'd like to see significant income tax increases on incomes over 100-200k or so, and all capital gains would be taxed at the highest possible rate, with no exceptions.
I think that savings are the thing that actually is messing up the economy - it creates an economy that is about nothing more than increasing the size of large bank accounts. Ordinary people get almost all of their money in wages, and that should be taxed the least as it creates a real economy. Somehow the economy did fine back when the rich were paying 75% or whatever.
I'm not sure I'd compare just portage and apt to really get a sense of the difference. Gentoo is fairly different from other distros in a few regards:
1. Tends to stay close to upstream - very vanilla. Granted, there are quite a few distros like this, but many of the big ones tend to tweak everything.
2, VERY flexible. You can run any version of anything with just about any version of anything else. You can run stable versions of everything, but then run ONE package off of the upstream git repository updating it 3x/day if you want. That ONE package can even be glibc, assuming you don't mind recompiling half your system 3x/day (but if you're a glibc developer maybe that hassle is worth being able to test out a change on a real system). No binary distro can support this to any significant degree, since the library dependencies become unmanageable.
3. USE flags are very useful when you are doing something atypical. I was just chatting with somebody the other day and he was running a CNC router controller on ubuntu (it uses a realtime kernel). The router software pulls in X11 and even OpenOffice as dependencies somehow. Now, this could just be bad packaging, but Gentoo gives you the ability to set USE=-X and then a package that could support X will not link to it, and you can more easily get rid of X. This makes Gentoo very well suited to niche/embedded projects.
4. Rolling releases mean that your software tends to stay up-to-date - change is more gradual (but constant). Of course, you can always limit yourself to only installing security bugs for a period of time, as long as it isn't too long as Gentoo does not indefinitely maintain upgrade paths (months are usually fine, years are usually not).
As you can see from the replies strong personalities are one of the things we are dealing with. I think this is somewhat due to the nature of the distro - it appeals to very-technically-minded people, and often these people can have VERY strong opinions on how things should be done. While issues flare up from time to time I'd say things have been relatively under control for a few years now - there was a big flare-up back around 2008 or so. People sometimes get fed up and quit, although from my perspective often when they do so their contributions don't really stop - they just start their own projects outside the distro and then somebody else brings the improvements back in (sometimes people just need their own space to work in, and results argue for themselves).
No IPO is going to punish a company for screwing employees.
Read the article - the CEO says, well, sure it is a lousy deal for the employees but it is what is best for the company.
Duh - getting people to work for years and then not paying them for that work is ALWAYS in the interest of the company as long as they can get away with it. Shareholders don't care about employees - they care about EPS, and when you revoke shares you increase EPS.
Does it really matter if the price of a stock is $30.15 or $30.22? What real-world-based valuation system has the ability to tell what the value of future earnings is within a couple of cents per share?
Nothing wrong with measuring prices in dollars and cents, but this sounds like arguing over how many millimeters it is from the sun to alpha centauri.
Yeah, at 5 pounds per transaction minimum I would think that issuing 10,000 trades per second would sink you pretty fast....
Sometimes I wonder if the solution is to just expand social security and the associated payroll taxes/etc so that employers don't need to offer a pension at all.
The problem with pensions is that employers have incentive to put as little money as possible into them while promising as much money as possible out of them. Promises get employees to stick around for less pay, and underfunding means spending less on compensation as well - either way the company wins. Pension managers typically have incentives to maximize today's value of the fund, but employees really care about the value 30 years from today.
The whole thing seems like a big scheme and I really wonder when this will just become the next bailout target.
Well, I'd say that system-hosing updates used to be more common in the past than they are now, but occasionally they still come up (usually with more warning these days, but not always).
I think that Gentoo (and everything else) will be much better off when btrfs becomes stable. With btrfs you can just snapshot everything inexpensively, run your update, and if you're hosed you can just switch back to your snapshot. I know when I update Gentoo VMs I make liberal use of snapshots.
Well, don't completely discount being able to set -march to something other than i686 and use -ftree-parallelize-loops or whatever.
Sure, this will break the odd package, but it is easy enough to just fall back to simple flags for a few of them.
I agree that having the amphibious assault capability is good. My point wasn't that we haven't done it because Marines aren't useful - rather we haven't done it because amphibious assaults tend to result in a lot of dead bodies. If we ever did do one then chances are we'd do all we can to keep those bodies to a minimum.
Well, the Navy and Air Force not being interested is a political issue - if this is the best solution then you make them get interested if you want the capability.
WWII was a bit of an unusual situation in that the enemy had a serious navy that had to be contended with. The fleet carriers were busy trying to control the entire theater and didn't have time to spend on one beachhead once they finished escorting the invasion fleet. Plus there were other options since short-field aircraft were more practical when that term described almost everything flying.
If the F35 did work I agree it might have value. It just seems like it would make more sense to just dedicate an airframe to this kind of role rather than making something that replaced both the harrier and the F18 at the same time.
I suspect some of the budget issue is the fact that international buyers are going to be disproportionately interested in the Marine role, since nobody else has carriers like the US. The Naval version would be useless to anybody but the US Navy. The F16 replacement would be broadly usable.
I do agree that when it comes to the 3rd world we shouldn't sacrifice the 80% solution for the sake of the 99% solution that nobody can afford.
Actually, that applies in general to medicine - sometimes the "do no harm" philosophy holds us back a little. In the present climate I'm surprised we even have pain-killers, since in terms of outcomes they don't cure any disease and yet they still can have side-effects.
I agree that the German system costs more in time than in money, and in truth a typical college student tends to have an abundance of the former, and they have some incentive to think twice before signing up.
I really was pointing my finger more to the US system where we promise everybody that they can be an astronaut, and frequently we don't even fail them out, but we collect four years of very expensive tuition from them and then tell them they can work in fast food.
Yes and no. You probably need solvents to extract it. Modern techniques use very refined solvents (we like our drugs to be clean), and those which are best suited to the job. If you're willing to deal with lots of waste/etc you could probably find some way to make it using less ideal materials.
If you just want to eat raw mold or whatever then that is pretty easy to manage, but of course allergies/etc will be even more of a problem than they already are.
I think the real issue is that the website seems to be very "green" in nature, and chemical manufacture just doesn't fit in all that well with that, although the fact that they are refining metals also stands out.
Yeah, but chemicals are so, ugh, non-appealing to the kinds of people who send money who fund initiatives like these. Ecology is nice and green. Antibiotics are tolerable even though they're made by evil profiteering pharmaceutical companies, but I suspect that every time somebody looked into them they kept finding these things called "solvents" used in their manufacture and decided that it wouldn't do for fundraising.
I'm not sure what the 3D printer would even be used for - just about everything in that list would probably use almost exclusively metal parts. Is there anything you can do with a 3D printer that you can't do with a CNC mill?
I have no issues with the work that they are doing, but again I think that making a list of "50" machines is more about fundraising (hey, we only need $4M to revolutionize the world). Otherwise, not have 51 or 52?
I think the issue is that the list seems rather arbitrary. They have fancy single-purpose stuff like a machine to extract aluminum from clay, but just a generic induction furnace to handle every other metal that exists.
No society with ONLY these 50 machines and only token imports would be able to exist with what most people would call "modern comforts." Sure, you might be able to make a few of them, but your CNC mill isn't going to work too well when a transistor in your power supply dies. You couldn't even fashion a vacuum tube replacement with that equipment list. You might be able to fashion a crude aluminum electrolytic capacitor if you're willing to wrap it by hand, although you'd have no way to test it to see if it is a suitable replacement for one that dies short of plugging it in and crossing your fingers.
And your tribal elder who manages your machine shop will be 30 years old, since anytime somebody cuts their finger on a piece of metal they face a real risk of death without antibiotics.
That isn't to say that these guys aren't doing useful work, but there will be plenty left to do after 50 machines are done.
I don't think anybody is suggesting that they aren't doing useful work. However, claiming that it only takes 50 tools to create a civilization with modern comforts really isn't true. No doubt those tools are very useful in getting there, but it takes more. For starters, it takes more than those 50 tools to produce the materials necessary to maintain them.
Again, these are 50 nice useful tools, but the number 50 seems a bit arbitrary and I think that it was chosen so that sponsors can be pitched with "hey, just give us a few million dollars and you can turn a bunch of huts into Greenwich Village." Sure, they don't come out and say it, but I think that is the impression the intro gives.
You are of course welcome to disagree. And I don't think that money sent to them is poorly spent, just that this isn't the magical band-aid that will save the 3rd world.
Uh, without drugs you can die from an infected stubbed toe. Believe it or not disease isn't exclusively God's punishment on people who eat too much. Also, you'll find that a good number of villages do have access to basic pharmaceuticals, and a little goes a long way.
I think the gripe is that their website says "The Global Village Construction Set (GVCS) is an open technological platform that allows for the easy fabrication of the 50 different Industrial Machines that it takes to build a small civilization with modern comforts."
Now, I'm all for open sourcing useful stuff like everything on the list. However, I'm under no illusions that it only takes 50 machines to run a society.
I also think the 3D printer is there more to please the hobbyist crowd than anything else. They're great for prototyping, but not really for manufacturing.
And does it strike anybody else as odd that they have about 15 different computer-controlled machines on the list and yet no computer? A CPU might be difficult to manufacture using their pelletizer.
Well, you can always stick Marine pilots in FA-18s or whatever, so the training in close air support isn't really a reason to buy a particular plane but just a matter of military politics.
Nobody is going to be fighting a campaign the size of the WWII pacific theater with our current Navy, I suspect that if we care enough about territory that we're willing to lose a few thousand marines in a day in an opposed landing then chances are we'll have every carrier in the fleet backing them up. Once you've landed in a modern war chances are those forces are going to stick around for a long time. Now, in WWII when one island had a secure beachhead the Navy would sail off to start working on the next island or whatever, but island hopping isn't really something I see happening in the future, and I can't think of a lot of situations where you'd otherwise need to have a bunch of little fixed airstrips without carriers nearby.
Maybe the solution is to just give the Marines their own plane or something, if it really is a need. I'm not sure it really even needs to be VTOL so much as STOL, and that can just be something with big unswept wings that is subsonic. You're not going to be using aircraft like this for air superiority.
Why does the carrier need to be 500 miles away?
We haven't even done a serious amphibious assault since WWII, though I understand that a few more recent conflicts involved some landings (probably all lightly opposed). If you really were to face an opposed landing, wouldn't you want those carriers covering you? Otherwise those nice close-air-support aircraft that the marines fly are going to be chopped to pieces by air superiority fighters flown by the enemy.
I just am not understanding the threat model here. If you're going up against barbarians with RPGs and spears then helicopters should do the trick, but chances are we'll still be dropping JDAMs on them. If you're going up against China in a war over Taiwan or something then frankly you need everything you have and you're in for quite a fight. I don't see where having a VTOL-capable jet is going to make much difference compared to what you'd get if you just spent that money on Naval or Air Force gear.
Gotta love bureaucracy, where even the secretary of defense is powerless to make a program happen.
The solution isn't to give the bureaucrats what they want, but to tell them that if they don't want to do their jobs properly somebody else will.
I need a new Porche. Trust me, I'm in a much better place to see it in my driveway than you are...
Why not just have the Navy fill this niche? Why do the planes need to launch from a particular class of ship? Why do they need to land on improvised fields? Just what kind of war could having this particular plane help win that would otherwise be lost?
The fact that a computer is running XP says very little about its capabilities. I can install XP on a brand new top-of-the-line computer. About the only thing it won't do is address more than 2GB of RAM per process (or whatever the 32-bit limit is, assuming we're talking about 32-bit XP). For most jobs XP works just fine, and it uses a lot less RAM than newer OSes. These days the ability to address more RAM is more important than wasting less, but then you have to factor in software compatibility - if your fancy 3D printer or whatever doesn't have Vista+ drivers then you're stuck on XP.
As for why computer science departments don't build their own computers... Most computers are commodity items - you don't need your typical workstation to be cutting-edge. I imagine that most computer science departments that own supercomputers DO build those themselves, that is if the design of the supercomputer itself was what they were pushing the envelope on. If they're only concerned with algorithms they might even buy those from somebody.
When you buy 1000 workstations from Dell you get business-class workstations. That means that they all have the same model hard drives/etc, and that Dell will still sell you an exact replacement for your power supply three years from now. If you test a security patch or whatever on one of them, you can be pretty sure you won't get one or two that mysteriously don't boot. Basically it is just the corporate IT model.
This is no different from an astronomy department buying a telescope out of a catalog. If they're looking to do some kind of cutting edge optical design improvement they'll end up building the thing (no doubt contracting the critical parts as custom jobs). If you are doing research that just involves making some brightness measurements on a bunch of stars or something and they're not 4 billion light years away then it probably makes more sense to just buy something for a few thousand dollars or whatever. Sure, lots of cutting-edge stuff happens on the HST or whatever, but there are also a lot of things that can be learned from commodity equipment that doesn't require you to book your time six months in advance. Disclaimer, I'm not an astronomer.
Bottom line is that any organization has to figure out which elements of what it is doing need to be world class, and which elements need to be good enough. At an undergraduate level I'd think I'd be more concerned with my professor having time to answer my questions and help me out than whether he spends six hours a day mass-assembling PCs.
Would it not make sense to weed out students with some kind of test that doesn't involve a 6-9 months of time and $20k+ in expense?
I agree that the root of the problem is that people try to study things they just aren't cut out for. However, there are better solutions to that than dragging them through months of classes at considerable expense.