Is it possible that you are thinking of the 'IBM Ultraport'(Yup, IBM has at least two physically nonstandard USB variants...)?
One wonders if the almost-total invisibility beyond POS systems is because IBM is really nasty about the license fees(from what little I've seen, your basic semi-embedded wintel gets an impressive markup when it becomes a 'Point of Sale System', so it is conceivable that IBM has no real interest in cutting margins in that market) or whether the USB-IF got cold feet because of concerns about consumer confusion...
USB's power delivery capabilities are puny, and only fully standard in theory(y hello thar, laptop drives with 'Y'-cables, and devices that work with some ports but not others, never mind the entire genre of devices that use USB for data/tethering but use USB-shaped AC adapters pushing an amp or more, often at something other than 5volts, when they need to do some serious charging); but they are substantially more consistent than those of powered USB. I wouldn't want to be the chap in charge of choosing what mixture of standard, 5V powered, 12V powered, and 24V powered ports to put onto a motherboard's I/O plate, that's for sure... On the plus side, assuming no nastiness from IBM, the full data bus compatibility with normal USB would make it pretty simple to whip up a 'Powered USB hub' that turns 1 normal USB socket into 4 of whatever flavor your application requires.
Oh, don't get me wrong, I have the greatest respect for the engineers who got a very fast bus working over cables that Joe User can be trusted with, for comparatively cheap. My point was merely that, because the bus is tied to the displayport, rather than just being an external 4x PCIe port, it is assured that most every use case already has a daisy-chain incompatible peripheral in the mix, the video device. Had the two not been combined, that wouldn't have been true.
The fact that getting the two signals into one connector is technically impressive is true; but it's still a handicap for all but terminal minimalists.
The difference is that PCI devices can do DMA, which means that they own you(classy systems with IOMMUs potentially excepted).
There are some cute demos with Firewire(which had the same drawback), of doing neat stuff like drawing arbitrary things on the screen by scribbling over the target's framebuffer, or modifying, live, bits of the kernel or other security-related components to escalate permissions. Also quite handy for debugging, since you can be reading the memory even as the host is puking up its brain; but dangerous...
USB, by contrast, essentially has to ask nicely and trick the host device into doing something stupid(ie. loading a mass storage class driver, mounting the volume, and autorunning haxxor.exe).
I'm not sure about PC Card; but ExpressCard depends: that connector is 1x USB2 + 1xPCIe. Most of the lower speed devices are actually just USB dongles designed to fit inside your laptop. The ones that are actually PCIe peripherals? I hope you brought your IOMMU....
It's the one port that might be technologically capable of ruling them all(use cases that actually use 8x or greater PCIe lanes excepted); but the economics, not of Thunderbolt specificially; but of 'one port to rule them all' generally, have always been problematic:
There are, in broad strokes, three variables that matter for an interface:
1. Cost: What does having a host port of that flavor add to a device's cost? What does being a slave device of that port cost?
2. Speed: What is possible, and what is precluded, based on the bandwidth of this interface?
3. Power: How beefy does a host device need to be to provide spec-approved power to a slave device? What classes of slave device will be impossible without a wall-wart of their own?
It isn't obviously possible to win in all three camps: Cost and speed are usually at each other's throats, which means that you can have a massive ecosystem with no high-end/high-speed products(USB), or an expensive port that you can't even plug a damn mouse into(firewire, before it got cheap). Power is more of an internal contradiction: If you spec the bus to provide considerable power, all sorts of 1-wire-only slave devices are possible; but the cost and bulk of host PSU/battery systems bloat. If you spec the bus to provide minimal power, every peripheral has some janky little wall-wart or y-cable; but even tablets and phones and such can be host devices. If you fudge, and spec a range, you get messy uncertainty.
The 'one interface' problem isn't a technological one; but an economic one.
Architecturally, this is true: with the ugly exception of the fact that a very high speed peripheral bus and a video-out interface were bodged into the same connector for no wildly obvious reason.
Because Firewire was a data-only thing, the probability that a given device would daisy-chain was actually pretty decent in the real world, and you could put the non-cooperative freak on the end of the chain. Thunderbolt more analogous to a port that sneaks firewire into your VGA-out(albeit in a way that makes splitting much more complex than a simple mechanical pinout adapter, is my understanding). Because there are loads of video-only devices in the world, the vast majority don't daisy-chain because video devices aren't expected to.
This is the trouble for Thunderbolt: As with classy firewire devices, most of the "thunderbolt peripherals" daisy-chain just fine. However, your Thunderbolt port is also your only video-out port, and something north of 99% of monitors, TVs, projectors, etc. have never heard of this 'daisy-chain' business.
They improvements TFA describes are the technique of handling collisions by having both sides back off for a randomly chosen period of time, and then send another ship, right?
Honestly, I'd be more worried about the fact that my not-at-all-cheap(and in many environments, not redundant, except at key points, definitely not for individual workstations) switches are booting from a dirt cheap flash card that's had its image loaded with verification so lousy that it missed the viral payload...
I've have a fair number of cheap and nasty flash cards die on me, and that'd be a whole lot more annoying if there were a few grand worth of switch wrapped around the card when it happened(though I can say from personal 'dding-a-working-card-onto-a-CF-card-from-Staples-to-replace-the-boot-medium-of-$3k-worth-of-Alcatel' experience that HP is hardly the only one that does it).
Oh, I'm certainly not going to argue that generating bad PR(along with genuine reductions in the quality delta between Sony and once-inferior competitor brands) has done anything but hurt the company, nor has their NIH approach helped them reduce either their own costs or the customer's total-cost-of-buying-a-sony-thing.
My point is just that, division by division, Sony's departments of Evil are doing alright, while Sony's departments of overpriced-but-not-actually-luxury are getting absolutely hammered. Barring some sort of benevolent visionary, it seems likely that the more-or-less-neutral stuff is going to get 'rightsized' and cut back, while the evil will wax yet fouler.
OLPC will "...build an educational ecosystem around the laptops". Fail.
Depends on how closely, and what exactly, them mean by saying 'around the laptops'. That could mean as little as "Given that the laptops are there, and can be assumed as a resource in the hands of students, we can now plan a series of etexts, electronic assessments, educational websites, etc. That seems reasonable enough, albeit subject to the usual pitfalls of making computers in education work.
Even some level of customization(ie. it'd be pretty painless to whip together a default CSS stylesheet for maximum readability on a known-model screen) isn't necessarily bad.
If they have a bunch of x86/linux/sugar native binaries that'll be dated in a year planned, they should strongly consider thinking again. Especially if they've managed to commit the unforgivable sin of building a 'computer education' curriculum that revolves around "These are the buttons to press in product XYZ that will be obsolete before you graduate..."
Unfortunately, the losses don't seem to be concentrated in their most anti-customer segments(arguably Sony/BMG, Sony Pictures Entertainment, or their gaming division, (with, arguably, their PC division also being included, if only for the sheer incompetence of the crapware bundled with them by default).
Instead, they got Absolutely Fucking Hammered in their "Once reputable; but basically who gives a fuck anymore and Panasonic is cheaper and as good and whoever makes 'Vizio" is cheaper still and I don't notice the difference" segments.
Is it arguable that arrogance is biting them in the ass? Sure. Along with generic failure-to-focus and commodification of what used to be quality-driven markets(with music and 'home theatre' gear, people have either gone hard upmarket to the boutique guys, or are basically buying on price. Sony is neither. Game over.
However, all their truly malicious rather than merely arrogant and feckless, divisions remain viable.
It is true that all consumer OSes of the period were pretty dire; but 'the period' didn't end for MacOS classic until 10.0(March, 2001) or 10.1(September, 2001)(earlier previews exist; but are rough techie stuff). This resulted in the somewhat embarassing period where NT4 and 2000 had things like actual multitasking, memory protection, and ACLs and Classic Mac OS didn't. In fairness, of course, NT4, in particular, was pretty rough, and fully-consumerized NT didn't happen until mid-late 2001.
Pre OSX MacOS, while it may have gotten raves for friendliness, and was somewhat less bug riddled, was architecturally more or less a toy OS compared to almost anything contemporary. The ecosystem wasn't as large, and the distribution vectors markedly less efficient; but the Mac malware was out there.
I suspect that Derbyshire largely got axed for being blatant and rather tacky, rather than because of anything new learned about his positions.
The National Review aims at being a classy, 'ideas', magazine for conservatives. Having somebody who has just written up a piece that is basically talk-radio or FreeRepublic material with slightly pedantic grammar tacked on just spoils the ambiance.
It is my understanding that one of the more boring elements of 'The Right Stuff'(tm) possessed by astronauts is the ability to subsist for an unlimited time on Mk. IV Standardized Nutritive Paste.
Using them in some sort of Das Boot "man the torpedoes!" role does seem counterproductive(if they even have the hardware for it...); but I have to imagine that carrying a few tactical nuclear missiles of modest range and yield would give you the ability to really fuck up a carrier group's day in relative safety.
I was largely joking with this one; but aiming for a topic that has actually been a bit of a 4th amendment sticky wicket of late:
For technological reasons, not clearly anticipated by the constitutional framers, an increasing percentage of most people's communication and storage are, technically speaking, Google's, or AT&T's, or the like, at any given time. There has been some debate as to whether your gmail, hotmail, etc. account is actually your 'papers, and effects', and subject to 4th amendment protection, or whether they are the host's, fair game whether you like it or not so long as the host feels like complying, which they generally do, especially if offered a cost-recovery fee.
Teams 'living document' and 'intent of the law' naturally tend to assert that the 4th amendment protects anything that would qualify as 'your papers', regardless of the minor matter of the technological means by which they are produced and stored at a given time. Others are not entirely of that opinion.(And this isn't really the first time this has come up: phone tapping that didn't require going inside your house with a set of wiring tools, was, at one point, entirely warrantless, until a court eventually accepted some bootlegger's defense claim that his calls deserved 4th amendment coverage even if they could be tapped at the telco end).
While it scores virtually infinite cartoon-supervillain points(seriously, a massive, ever-expanding labarynthine nuclear-powered ice fortress?), I have to imagine that the cost/benefit got a lot less exciting once the more prosaic 'lots of nuclear submarines sneaking around, also we can use them to attack ships, in a pinch,' strategy became viable.
Incidentally, for anybody who likes our dread overlord Cthulhu, and wishes to be eaten first, this sounds like something ripped straight from A Colder War...
If the Founding Fathers had meant to protect your email from search and seizure, they would have had Ben Franklin invent the OSI 7 layer model and SMTP and then mentioned them in the Fourth Amendment!
I'm sure that having discovered the Secret Conflict at the Heart of the Liberal Agenda is fun and all; but you'll note that the expansions of fun surveillance technology, the making-legal of their use, and the creation and expansion of the entities in charge of them all occur under a "national defense" guise and in the aftermath of assorted defense-related incidents, from the Alien and Sedition acts to the present day.
You'll note that this is an NSA(formed during WWII, post Pearl Harbor) facility being built on a military base to store and sift through the product of post 9/11 surveillance powers exercised largely in concert with and with the assistance of the telcos, retroactively authorized because zOMG Terrorists!.
You'll note also that, in terms of voting patterns, 'National Defense' is broadly popular; both among pro-social-welfare-program(though it threatens the budgets of the programs they like, so support isn't total) and anti-social-welfare-program constituencies(though it threatens their tax cuts, so support isn't total). Each constituency does have a subgroup who oppose it(generally civil libertarians and peaceniks on the one hand, and classical libertarians on the other); but both are comparatively feeble voting groups, especially if something excitingly threatening is going on.
If we were talking about the(also present; but rather shabbier) Department of Education or FDA or such datacenter operations, the fact that you can't have social welfare programs without data gathering would, in fact, be salient. However, we are talking about a DoD spook shop. 'Defense' spending can be, and often is, in terms of empirical voting patterns and popular support, quite neatly decoupled from social welfare spending. It's actually reasonably common under non-communist authoritarians.
The textbook guys are really more of a political problem than a technical one. There isn't any particular connection between paper printing and buying from a vendor who retains the copyright and charges accordingly, if one were to purchase a text outright and shop around for people willing to print and bind it, the per unit bids would likely be considerably lower. As you note, there also isn't any magic connection between digital distribution and low prices. If anything, nuking the used and import markets will make the situation worse(though digital distribution does have low fixed costs, which makes small-scale and iteratively developed stuff possible IFF that is supported...)
One has to wonder if Intel is banking on substantial subsidizing going on, or whether they are looking at the higher edge of 'developing'. $300 undercuts iPads if the other specs check out; but isn't very aggressive at all for a future goal by the standards of assorted android tablets of varying quality, available now, or assorted netbooks, packing Intel silicon even, available now.
Given Intel's preferences for nice margins, and ongoing woes at hitting low power targets, along with the horridness of really cheap touchscreens, it isn't a huge surprise that their target would be so undramatic; but it doesn't fill one with optimism for their success in cost-sensitive markets.
As I noted, I'm reasonably certain that the replacement of paper texts by electronics, for most purposes, is likely to be a relatively near-future phenomenon. It depends somewhat on your purpose. Amazon's pilot studies with Kindle textbooks at some American university received tepid reviews, too unwieldy for margin notes/highlighting/quickly finding your place; but the convenience and portability factors of ereader devices are said to make owners of such bigger readers of texts where that isn't a problem. Presumably, it would also be an advantage in locations that have some data connectivity but lousy supply chains. Even a CSD GRPS connection can likely move a book faster than many postal services can.
If broad selection in remote areas, very low fixed costs for a 'print run', or frequent updates are required, electronic text is already a winner. If you want fully papery notetaking capabilities, we seem to be Not Quite There Yet. The jury still seems to be out on readability: LCDs are fast enough but have their detractors, e-ink is more comfortable; but refresh and color aren't so hot. I honestly don't know whether, for standardized texts like newspapers and textbooks, whether the pure economics are there yet or not. Digital has a marginal cost of nearly nothing; but the fixed costs are nontrivial and are spread broadly across end users, where conditions favor hardware damage, rather than centralized at the printer. Electronic gadgets also tend to fail less gracefully than paper. On the other hand, paper has nagging per-unit costs and doesn't favor small runs.
In any case, though, that only addresses replacing books which are, to go by the experience of the developed world, not the primary limiting factor in educational outcomes...
It's hard not to be pessimistic about this scheme. I'm sure that Intel has the engineering muscle and the cash to at least shove some units out the door(if not actually hit their targeted TDPs and battery lives) and the hardware might even be an interesting alternative to some of the present ARM SoC tablets at a similar price point; but that won't really solve the basic problem:
Actually turning computers into educational results, even in the wealthy subsections of wealthy industrialized countries where access to computers has been ubiquitous for a number of years now, has turned out to be difficult. Not necessarily impossible(and certainly a boon for the nonzero-but-hard-to-replicate autodidactic success stories); but definitely not obvious, and generally not happening in places where reasonable amounts of educational success were already being achieved by conventional methods.
It is likely that digital distribution technologies will, at some point in the reasonably near future, firmly undercut print on total price(ie. counting the units needed to read the stuff, and the infrastructure, not just the marginal cost of somebody with a computer and an internet connection snarfing Project Gutenberg), which would be a boon to anybody who has plans for producing material that don't involve paying substantial per-unit license fees; but that only brings computers to parity with print(also, it is fairly likely that sub-$100 e-ink or super-cheapy LCD devices will undercut on price well before fancy tablets do).
Shipping aggressively cheap and robust hardware is certainly a nontrivial engineering challenge, and a necessary condition of any educating-the-poor-with-computers plan; but we already have a test case, wealthy denizens of the developed world, where the hardware and infrastructure exist and we've been able to watch the pedagogical techniques and software in action. The results have not been... overly encouraging...
Slashdot Mirror
Is it possible that you are thinking of the 'IBM Ultraport'(Yup, IBM has at least two physically nonstandard USB variants...)?
One wonders if the almost-total invisibility beyond POS systems is because IBM is really nasty about the license fees(from what little I've seen, your basic semi-embedded wintel gets an impressive markup when it becomes a 'Point of Sale System', so it is conceivable that IBM has no real interest in cutting margins in that market) or whether the USB-IF got cold feet because of concerns about consumer confusion...
USB's power delivery capabilities are puny, and only fully standard in theory(y hello thar, laptop drives with 'Y'-cables, and devices that work with some ports but not others, never mind the entire genre of devices that use USB for data/tethering but use USB-shaped AC adapters pushing an amp or more, often at something other than 5volts, when they need to do some serious charging); but they are substantially more consistent than those of powered USB. I wouldn't want to be the chap in charge of choosing what mixture of standard, 5V powered, 12V powered, and 24V powered ports to put onto a motherboard's I/O plate, that's for sure... On the plus side, assuming no nastiness from IBM, the full data bus compatibility with normal USB would make it pretty simple to whip up a 'Powered USB hub' that turns 1 normal USB socket into 4 of whatever flavor your application requires.
Oh, don't get me wrong, I have the greatest respect for the engineers who got a very fast bus working over cables that Joe User can be trusted with, for comparatively cheap. My point was merely that, because the bus is tied to the displayport, rather than just being an external 4x PCIe port, it is assured that most every use case already has a daisy-chain incompatible peripheral in the mix, the video device. Had the two not been combined, that wouldn't have been true.
The fact that getting the two signals into one connector is technically impressive is true; but it's still a handicap for all but terminal minimalists.
The difference is that PCI devices can do DMA, which means that they own you(classy systems with IOMMUs potentially excepted).
There are some cute demos with Firewire(which had the same drawback), of doing neat stuff like drawing arbitrary things on the screen by scribbling over the target's framebuffer, or modifying, live, bits of the kernel or other security-related components to escalate permissions. Also quite handy for debugging, since you can be reading the memory even as the host is puking up its brain; but dangerous...
USB, by contrast, essentially has to ask nicely and trick the host device into doing something stupid(ie. loading a mass storage class driver, mounting the volume, and autorunning haxxor.exe).
I'm not sure about PC Card; but ExpressCard depends: that connector is 1x USB2 + 1xPCIe. Most of the lower speed devices are actually just USB dongles designed to fit inside your laptop. The ones that are actually PCIe peripherals? I hope you brought your IOMMU....
It's the one port that might be technologically capable of ruling them all(use cases that actually use 8x or greater PCIe lanes excepted); but the economics, not of Thunderbolt specificially; but of 'one port to rule them all' generally, have always been problematic:
There are, in broad strokes, three variables that matter for an interface:
1. Cost: What does having a host port of that flavor add to a device's cost? What does being a slave device of that port cost?
2. Speed: What is possible, and what is precluded, based on the bandwidth of this interface?
3. Power: How beefy does a host device need to be to provide spec-approved power to a slave device? What classes of slave device will be impossible without a wall-wart of their own?
It isn't obviously possible to win in all three camps: Cost and speed are usually at each other's throats, which means that you can have a massive ecosystem with no high-end/high-speed products(USB), or an expensive port that you can't even plug a damn mouse into(firewire, before it got cheap). Power is more of an internal contradiction: If you spec the bus to provide considerable power, all sorts of 1-wire-only slave devices are possible; but the cost and bulk of host PSU/battery systems bloat. If you spec the bus to provide minimal power, every peripheral has some janky little wall-wart or y-cable; but even tablets and phones and such can be host devices. If you fudge, and spec a range, you get messy uncertainty.
The 'one interface' problem isn't a technological one; but an economic one.
Architecturally, this is true: with the ugly exception of the fact that a very high speed peripheral bus and a video-out interface were bodged into the same connector for no wildly obvious reason.
Because Firewire was a data-only thing, the probability that a given device would daisy-chain was actually pretty decent in the real world, and you could put the non-cooperative freak on the end of the chain. Thunderbolt more analogous to a port that sneaks firewire into your VGA-out(albeit in a way that makes splitting much more complex than a simple mechanical pinout adapter, is my understanding). Because there are loads of video-only devices in the world, the vast majority don't daisy-chain because video devices aren't expected to.
This is the trouble for Thunderbolt: As with classy firewire devices, most of the "thunderbolt peripherals" daisy-chain just fine. However, your Thunderbolt port is also your only video-out port, and something north of 99% of monitors, TVs, projectors, etc. have never heard of this 'daisy-chain' business.
They improvements TFA describes are the technique of handling collisions by having both sides back off for a randomly chosen period of time, and then send another ship, right?
Honestly, I'd be more worried about the fact that my not-at-all-cheap(and in many environments, not redundant, except at key points, definitely not for individual workstations) switches are booting from a dirt cheap flash card that's had its image loaded with verification so lousy that it missed the viral payload...
I've have a fair number of cheap and nasty flash cards die on me, and that'd be a whole lot more annoying if there were a few grand worth of switch wrapped around the card when it happened(though I can say from personal 'dding-a-working-card-onto-a-CF-card-from-Staples-to-replace-the-boot-medium-of-$3k-worth-of-Alcatel' experience that HP is hardly the only one that does it).
Oh, I'm certainly not going to argue that generating bad PR(along with genuine reductions in the quality delta between Sony and once-inferior competitor brands) has done anything but hurt the company, nor has their NIH approach helped them reduce either their own costs or the customer's total-cost-of-buying-a-sony-thing.
My point is just that, division by division, Sony's departments of Evil are doing alright, while Sony's departments of overpriced-but-not-actually-luxury are getting absolutely hammered. Barring some sort of benevolent visionary, it seems likely that the more-or-less-neutral stuff is going to get 'rightsized' and cut back, while the evil will wax yet fouler.
OLPC will "...build an educational ecosystem around the laptops". Fail.
Depends on how closely, and what exactly, them mean by saying 'around the laptops'. That could mean as little as "Given that the laptops are there, and can be assumed as a resource in the hands of students, we can now plan a series of etexts, electronic assessments, educational websites, etc. That seems reasonable enough, albeit subject to the usual pitfalls of making computers in education work.
Even some level of customization(ie. it'd be pretty painless to whip together a default CSS stylesheet for maximum readability on a known-model screen) isn't necessarily bad.
If they have a bunch of x86/linux/sugar native binaries that'll be dated in a year planned, they should strongly consider thinking again. Especially if they've managed to commit the unforgivable sin of building a 'computer education' curriculum that revolves around "These are the buttons to press in product XYZ that will be obsolete before you graduate..."
Unfortunately, the losses don't seem to be concentrated in their most anti-customer segments(arguably Sony/BMG, Sony Pictures Entertainment, or their gaming division, (with, arguably, their PC division also being included, if only for the sheer incompetence of the crapware bundled with them by default).
Instead, they got Absolutely Fucking Hammered in their "Once reputable; but basically who gives a fuck anymore and Panasonic is cheaper and as good and whoever makes 'Vizio" is cheaper still and I don't notice the difference" segments.
Is it arguable that arrogance is biting them in the ass? Sure. Along with generic failure-to-focus and commodification of what used to be quality-driven markets(with music and 'home theatre' gear, people have either gone hard upmarket to the boutique guys, or are basically buying on price. Sony is neither. Game over.
However, all their truly malicious rather than merely arrogant and feckless, divisions remain viable.
It is true that all consumer OSes of the period were pretty dire; but 'the period' didn't end for MacOS classic until 10.0(March, 2001) or 10.1(September, 2001)(earlier previews exist; but are rough techie stuff). This resulted in the somewhat embarassing period where NT4 and 2000 had things like actual multitasking, memory protection, and ACLs and Classic Mac OS didn't. In fairness, of course, NT4, in particular, was pretty rough, and fully-consumerized NT didn't happen until mid-late 2001.
Pre OSX MacOS, while it may have gotten raves for friendliness, and was somewhat less bug riddled, was architecturally more or less a toy OS compared to almost anything contemporary. The ecosystem wasn't as large, and the distribution vectors markedly less efficient; but the Mac malware was out there.
I suspect that Derbyshire largely got axed for being blatant and rather tacky, rather than because of anything new learned about his positions.
The National Review aims at being a classy, 'ideas', magazine for conservatives. Having somebody who has just written up a piece that is basically talk-radio or FreeRepublic material with slightly pedantic grammar tacked on just spoils the ambiance.
It is my understanding that one of the more boring elements of 'The Right Stuff'(tm) possessed by astronauts is the ability to subsist for an unlimited time on Mk. IV Standardized Nutritive Paste.
Using them in some sort of Das Boot "man the torpedoes!" role does seem counterproductive(if they even have the hardware for it...); but I have to imagine that carrying a few tactical nuclear missiles of modest range and yield would give you the ability to really fuck up a carrier group's day in relative safety.
It could be smugglers. There are so many uncharted settlements.
I was largely joking with this one; but aiming for a topic that has actually been a bit of a 4th amendment sticky wicket of late:
For technological reasons, not clearly anticipated by the constitutional framers, an increasing percentage of most people's communication and storage are, technically speaking, Google's, or AT&T's, or the like, at any given time. There has been some debate as to whether your gmail, hotmail, etc. account is actually your 'papers, and effects', and subject to 4th amendment protection, or whether they are the host's, fair game whether you like it or not so long as the host feels like complying, which they generally do, especially if offered a cost-recovery fee.
Teams 'living document' and 'intent of the law' naturally tend to assert that the 4th amendment protects anything that would qualify as 'your papers', regardless of the minor matter of the technological means by which they are produced and stored at a given time. Others are not entirely of that opinion.(And this isn't really the first time this has come up: phone tapping that didn't require going inside your house with a set of wiring tools, was, at one point, entirely warrantless, until a court eventually accepted some bootlegger's defense claim that his calls deserved 4th amendment coverage even if they could be tapped at the telco end).
While it scores virtually infinite cartoon-supervillain points(seriously, a massive, ever-expanding labarynthine nuclear-powered ice fortress?), I have to imagine that the cost/benefit got a lot less exciting once the more prosaic 'lots of nuclear submarines sneaking around, also we can use them to attack ships, in a pinch,' strategy became viable.
Incidentally, for anybody who likes our dread overlord Cthulhu, and wishes to be eaten first, this sounds like something ripped straight from A Colder War...
If the Founding Fathers had meant to protect your email from search and seizure, they would have had Ben Franklin invent the OSI 7 layer model and SMTP and then mentioned them in the Fourth Amendment!
I'm sure that having discovered the Secret Conflict at the Heart of the Liberal Agenda is fun and all; but you'll note that the expansions of fun surveillance technology, the making-legal of their use, and the creation and expansion of the entities in charge of them all occur under a "national defense" guise and in the aftermath of assorted defense-related incidents, from the Alien and Sedition acts to the present day.
You'll note that this is an NSA(formed during WWII, post Pearl Harbor) facility being built on a military base to store and sift through the product of post 9/11 surveillance powers exercised largely in concert with and with the assistance of the telcos, retroactively authorized because zOMG Terrorists!.
You'll note also that, in terms of voting patterns, 'National Defense' is broadly popular; both among pro-social-welfare-program(though it threatens the budgets of the programs they like, so support isn't total) and anti-social-welfare-program constituencies(though it threatens their tax cuts, so support isn't total). Each constituency does have a subgroup who oppose it(generally civil libertarians and peaceniks on the one hand, and classical libertarians on the other); but both are comparatively feeble voting groups, especially if something excitingly threatening is going on.
If we were talking about the(also present; but rather shabbier) Department of Education or FDA or such datacenter operations, the fact that you can't have social welfare programs without data gathering would, in fact, be salient. However, we are talking about a DoD spook shop. 'Defense' spending can be, and often is, in terms of empirical voting patterns and popular support, quite neatly decoupled from social welfare spending. It's actually reasonably common under non-communist authoritarians.
The textbook guys are really more of a political problem than a technical one. There isn't any particular connection between paper printing and buying from a vendor who retains the copyright and charges accordingly, if one were to purchase a text outright and shop around for people willing to print and bind it, the per unit bids would likely be considerably lower. As you note, there also isn't any magic connection between digital distribution and low prices. If anything, nuking the used and import markets will make the situation worse(though digital distribution does have low fixed costs, which makes small-scale and iteratively developed stuff possible IFF that is supported...)
One has to wonder if Intel is banking on substantial subsidizing going on, or whether they are looking at the higher edge of 'developing'. $300 undercuts iPads if the other specs check out; but isn't very aggressive at all for a future goal by the standards of assorted android tablets of varying quality, available now, or assorted netbooks, packing Intel silicon even, available now.
Given Intel's preferences for nice margins, and ongoing woes at hitting low power targets, along with the horridness of really cheap touchscreens, it isn't a huge surprise that their target would be so undramatic; but it doesn't fill one with optimism for their success in cost-sensitive markets.
As I noted, I'm reasonably certain that the replacement of paper texts by electronics, for most purposes, is likely to be a relatively near-future phenomenon. It depends somewhat on your purpose. Amazon's pilot studies with Kindle textbooks at some American university received tepid reviews, too unwieldy for margin notes/highlighting/quickly finding your place; but the convenience and portability factors of ereader devices are said to make owners of such bigger readers of texts where that isn't a problem. Presumably, it would also be an advantage in locations that have some data connectivity but lousy supply chains. Even a CSD GRPS connection can likely move a book faster than many postal services can.
If broad selection in remote areas, very low fixed costs for a 'print run', or frequent updates are required, electronic text is already a winner. If you want fully papery notetaking capabilities, we seem to be Not Quite There Yet. The jury still seems to be out on readability: LCDs are fast enough but have their detractors, e-ink is more comfortable; but refresh and color aren't so hot. I honestly don't know whether, for standardized texts like newspapers and textbooks, whether the pure economics are there yet or not. Digital has a marginal cost of nearly nothing; but the fixed costs are nontrivial and are spread broadly across end users, where conditions favor hardware damage, rather than centralized at the printer. Electronic gadgets also tend to fail less gracefully than paper. On the other hand, paper has nagging per-unit costs and doesn't favor small runs.
In any case, though, that only addresses replacing books which are, to go by the experience of the developed world, not the primary limiting factor in educational outcomes...
It's hard not to be pessimistic about this scheme. I'm sure that Intel has the engineering muscle and the cash to at least shove some units out the door(if not actually hit their targeted TDPs and battery lives) and the hardware might even be an interesting alternative to some of the present ARM SoC tablets at a similar price point; but that won't really solve the basic problem:
Actually turning computers into educational results, even in the wealthy subsections of wealthy industrialized countries where access to computers has been ubiquitous for a number of years now, has turned out to be difficult. Not necessarily impossible(and certainly a boon for the nonzero-but-hard-to-replicate autodidactic success stories); but definitely not obvious, and generally not happening in places where reasonable amounts of educational success were already being achieved by conventional methods.
It is likely that digital distribution technologies will, at some point in the reasonably near future, firmly undercut print on total price(ie. counting the units needed to read the stuff, and the infrastructure, not just the marginal cost of somebody with a computer and an internet connection snarfing Project Gutenberg), which would be a boon to anybody who has plans for producing material that don't involve paying substantial per-unit license fees; but that only brings computers to parity with print(also, it is fairly likely that sub-$100 e-ink or super-cheapy LCD devices will undercut on price well before fancy tablets do).
Shipping aggressively cheap and robust hardware is certainly a nontrivial engineering challenge, and a necessary condition of any educating-the-poor-with-computers plan; but we already have a test case, wealthy denizens of the developed world, where the hardware and infrastructure exist and we've been able to watch the pedagogical techniques and software in action. The results have not been... overly encouraging...