I apologize if I didn't make it clear: Because of the enthusiasm for televised entertainment, penetration of basic CRT TVs with composite or RF ins has reached(at least on the neighborhood level) down to people who live in shacks and have highly intermittent access to electricity siphoned off the nearest utility pole and no running water.
Now that flat panels have gotten cheap(and a combination of shipping costs, consumer tastes, and environmental concerns have largely eliminated the CRT TV from the first world retail channel), you will find that many low income households have a TV(almost definitely capable of composite in, if purchased in the last couple of years, or going forward, probably a cheap and nasty LCD with an HDMI port). They've already gone out and purchased one. It's a sunk cost. This hypothetical board would be capable of exploiting the money already spent. I do think that they pared it down a bit too far in terms of I/O; but between composite and HDMI/DVI, video output doesn't seem too serious an issue(the only issue would be lack of VGA-HDMI is electrically compatible with DVI-D; but does not include the analog signals broken out by a full DVI connector).
Of course, if you live in the middle of nowhere, "receive mail" is only an "archaic form of communication"; because newer ones are also subsidized for you. Rural telecommunications and electrification are also projects that weren't exactly undertaken because the ROI was enticing to Wall Street.
If the objective is efficiency, you might as well tell any rural areas that aren't totally loaded to shove off and learn to enjoy natural solitude, and let any impecunious urban areas enjoy the newfound feeling of community that comes with being cut off.
I'm not sure that that would be such a popular move; but it definitely would decrease the per-capital cost of infrastructure.
The question of interest, I think, is whether the postal service is in the red because they suck, or in the red because their mandate has(with the decline of the letter as a medium) largely shrunk to cover the unprofitable shit work of shipping(Picking up a letter, at your post box, in fucking nowhere, and delivering it to somebody else's postbox in a different fucking nowhere on the other side of the country, for the price of a stamp, is not exactly a lucrative niche...) while FedEx and UPS are free to ignore the low value segments and focus on carrying packages, with an emphasis on larger shippers and aggregated pickups.
In a sense, the real question facing us about the postal service is approximately similar to the real questions behind rural electrification, or telcom access: There are places in the country where providing infrastructure is, per capita, cheap. There are others where providing it is really, really, really expensive. There are areas where the infrastructure customers are relatively wealthy, and ones where they definitely aren't.
We can definitely trust the private sector(as long as they don't gain monopolies or oligopolies) to serve areas where customer willingness to pay is sufficiently high and cost per capita sufficiently low. We then come to the question of what to do about the ones where that isn't the case.
Obviously, this doesn't imply that the postal system is well managed, or that it couldn't do better(and, if improvement is available, it should definitely be undertaken); but, like rural telco and electrification, the fundamental question is not one of wringing out small operational efficiencies; but of whether or not we, as a society, wish to provide a baseline infrastructure to areas where it is not strictly economically justified. Depending on exactly how efficient you are, these areas may be somewhat smaller or somewhat larger; but it will almost always be the case that you could improve financial performance by just writing off your lossy service areas and letting them suck it up.The question is, is that what we want?
At the hardware level, it should be no more or less inscrutable-and-functionally-unfixable than essentially any modern board. At the software level, it probably depends on where the providing entity falls on the "Let them explore" vs. "Lock it down, we'll tell them what they need" spectrum: If you give the user full control over the device, in the spirit of hackerly independent exploration, you'll probably have them show up in fair variety of conditions. If you control more or less tightly, you probably won't.
Barring direct physical destruction of hardware, though, it wouldn't be total rocket surgery to have a rescue bootloader that, say, causes the device to expose its entire internal flash as a USB MSC volume if a particular sequence of inputs is given. Somebody fuck up their stick? Plug it into a computer, enter the rescue sequence, dd if=base_image.img of=/dev/SD_borked_stick Wait a minute or two, unplug, back in business.
Technologically, these should be no harder(and quite possibly easier) to deal with. Philosophically, the tradeoff between making a computer "theirs" vs. making sure that it is ready for class is only partially a technological problem, and much more a problem of educational philosophy(though the OLPC project did some interesting work on technology to reduce the sharpness of the tradeoff...)
The spec list mentions composite output as well, though it doesn't seem to be broken out on the dev board(to a standard connector). All but the nastiest TVs can handle that. In the case of schools, I assume that the use case would be making the computer lab cheaper(even the cheapest nettops run ~$150 on a good day, and Thin client hardware, presumably because of its Big Serious Corporate provenance, can run rather more than that. Bottom end business-line PCs that can be more or less relied upon to have a standard hardware profile are ~$200 in off-lease refurb, ~$500 new). Even if you are using a bunch of Windows only or x86 only software, these little puppies would be substantially cheaper than most thin-client offerings(and linux-on-ARM is supported by Citrix and VMware, and has support for RDP and X, which pretty much covers all the bases). If you are doing things that are supported natively, you could skip the terminal server and go cheaper still.
Plus(somewhat sad to say), there probably are a reasonable number of families where a TV ranks higher than a computer. A basic nasty "HD" LCD setup will almost always come with HDMI now, and is pretty accessible(even if by rent-to-own or usurious credit card borrowing) for people pretty far down the totem pole. Kiddo might well have better luck sneaking in some computer time when the parents aren't watching soaps or sports than having access to a standard computer....
It might be that it is implementing something described in the "On-The-Go and Embedded Host Supplement to the USB 2.0 Specification", which includes the ability for devices to switch between master and slave roles(which would suit the use of a slave-device type connector on the board alongside the fact that the board is driving a hub loaded with slave devices...) My reading of that spec suggests that the OTG device shouldn't be using that particular plug(they are supposed to use microA/B sockets only); but the history of USB is loaded with mechanical abuses of the specs, so that wouldn't be a major surprise...
Not quite as capable, in certain respects, as the Gumstix line of similarly sized ARM boards; but, on the other hand, you'll be lucky to walk away with change from $200 after getting your main board and an I/O expander if needed if you go that route. I wonder where the cost delta comes from?
One minor nit, this system doesn't appear to have any onboard networking(aside from the USB port which, from the picture of it connected to the B port of a hub, would appear to be one of those 'OTG' master or slave jobbies, which could easily enough act as a USB CDC or RNDIS connection to a host PC(which is kind of a waste for a single user; but a basic cheapy desktop loaded with USB cards could easily act as a gateway/fileserver/host for CPU intensive or x86 only programs over an X tunnel for a classroom full of the things)). I have to wonder if a "Flash drive sized" computer that basically doesn't work unless connected to a powered USB hub and a USB network adapter or CDC host PC might be rather less useful than would be a "pack of playing cards sized" computer that actually has a NIC and at least enough USB ports to support a mouse and keyboard(and ideally one extra for miscellaneous purposes)...
Are you calling all nine of these photogenic-but-irrelevant stock-footage models liars? For shame, cad. And they have "secure phone call" technology! That's, like, CIA shit, man. Totally trustworthy. I, for one, eagerly await the chance to enter all SSNs and CCNs into an improperly secured form when I get an email from the "s0ny h3rbal cust0mer Protection Se4vice" asking me to verify.
I offered to license my patented technique of "Apparatus and method for the preparation of a shit-adhesive wall"; but we couldn't come to an agreement. I bet they are sorry now...
The problem isn't taxing for road usage(in fact, were it not for the culture that you mention, such a proposal would theoretically be equally desireable for conservatives(who tend to like "users fees" because they dislike taxes and dislike the possibility that people might end up being subsidized)) and liberals(who tend to suspect that individual-vehicle transport, especially the petro kind, has been the recipient of massive, if not overt, subsidies for pretty much the entire post WWII period, making possible an entire suburban material culture that cannot exist without those subsidies).
The problem with mileage taxes is with administration, enforcement, and mission creep. Unlike, say, fuel taxes(which have the fairly convenient advantage of approximately taxing a composite of vehicle size and vehicle miles traveled, which is a good rough estimate of vehicle road 'consumption', without ever having the tax man leave the shop), mileage taxes require, at bare minimum, the tax man inspecting every vehicle's odometer(or at least a sufficiently large sample that most people report theirs honestly on the 1040v.2). If mission creep or ulterior interests come into play, you could pretty easily end up with GPS black boxes, or other less tweakable(you don't even need to crack the odometer, just ensure that the rotations to miles conversion it is using is based on slightly smaller wheels than you are using) methods.
Employing an overhead-heavy, potentially very invasive, taxation strategy when a simple retail sale tax one would work nearly as well strikes me as a serious problem. The notion that, while many roads are basically natural monopolies, and thus cannot be run on free market lines, one can attempt to make one's payment for road use approximately proportional to their use of roads seems entirely sensible.
Arguably, "PCI" the standard/set of requirements is bullshit either way: If a set of requirements designed to force security allows egregious mistakes to be made and/or egregious violators to slip through, it pretty much sucks.
I suspect that anybody who does a competent, good faith, implementation of PCI is at least part of the way toward a secure operation; but PCI isn't intended as polite good advice...
On additional consideration, I'm thinking that there might actually be a more subtle advantage for intel here:
Atom, as a product, has always suffered from two major defects:
It is too high power to profitably compete against ARM(except likely in the embedded/industrial space, which isn't wildly power constrained, except in that fans are bad for reliability in hostile environments, where the fact that you can get an x86 board for $80, or a rugged x86 for under $1k has probably murdered quite a few planned ports to embedded boards based on other architectures...), which cuts Intel out of the smartphone/tablet/STB widget/etc. market.
On the other hand, it is sufficiently cheap, and for many applications sufficiently powerful(especially if Intel were to drop the artificial gimping of screen size and shipping memory) that it threatens the margins of the lower-end of the core 2-derivative business. Thus, intel faces a perverse incentive to keep its performance down, while not being able to push its power consumption down far enough.
Fabbing the "premium" ARM SoC, though, might solve this problem: ARM's existing low-power chops, combined with intel's process superiority should allow them to produce an ARM part for which people will pay a fair premium, thus keeping margins on the part where intel wants them(more or less). And, and here is the icing on the cake, Pretty much no matter how good the ARM SoC they build is, it will take absolute ages to displace their bread-and-butter legacy wintel application-running market.
The world is full of(mostly horrid) legacy software that will never be non-x86. However, since much of it is older, Intel can't afford to make the Atom too good or it will cannibalize the market for core-2s and i3s. ARM, though, by virtue of binary incompatibility, can never cannibalize that market, (and is already cannibalizing the 'light browsing and casual games' market, whether intel likes it or not). Thus, it might actually be a very interesting play for Intel to pare down the low-end x86s as much as competition with AMD will allow, and move its offerings toward two distinct camps of "the highest margin x86 parts we can sell" and "simply the best ARM devices available". That would secure the legacy x86 software base from cannibalization by cheap, low-margin parts, since even the nicest ARM won't be a substitute, while also putting intel in the position to at least make some money from the mobile sector...
An ID that is poor as shit; but legally resident, or with a reasonably clean background, is likely worth something to somebody whose ID is no wealthier; but lacks one or both of those features...
The bottom few rungs of the totem pole are Seriously Ugly.
They call it "The intelligence community". Don't you think that, like any community, it probably has a creepy, lecherous old man somewhere who would patriotically volunteer to work days, nights, and weekends, with nothing but Ramen and lotion for sustenance, to crack the secret of what diabolical plans of terror could be hiding within the depths of 4.6GB of "Bad Burqa Babes: Ankle exposed! vol. 1-237" and 3.2GB of what looks alarmingly like palestinian-occupied-territories-produced amateur femdom/malesub with titles such as "Mossad Lesbians: Haraam Jewesses XXXposed"?
They practice what they call "Communism with Chinese characteristics". Westerners know it as "Crony capitalism" or "what we are busy running headlong into".
The situation is a bit trickier(as I said, this feature is a bit.. oversold); but it does support wireless connections as well. For corporate network purposes, there are ways to provision auth keys to protected memory(so that the embedded management processor can strike up a wireless connection even if the HDD is currently on fire). Off site, it depends on the host OS to strike up a wireless connection(since anything from no auth, to WPA-PSK, to random captive-portal stuff could be at work); but is then capable of initiating a VPN connection back to HQ.
I've actually been a little surprised at how painful it is to use, as an IT management feature(apparently all Intel's good software people work on compilers or chip design tools...); and we've actually largely abandoned bothering with it(on occasion, it can wake a machine when WoL won't; but most of the bells and whistles are really rough around the edges) but its capabilities are a bit creepy. My understanding, as well, is that even more control excitement is Coming Real Soon for systems that have an embedded cellular broadband card: remote control/lock/HDD encryption key purge, at the BIOS level, any time the machine has some power and is near a tower.
From the conversations we've recorded, I can state with 99.35% accuracy, with an N value in the millions, that 9 out of 10 onstar owners are definitely using the preinstalled hardware, and whatever firmware update we prefer.
XOXOXO- GM
You had plywood? I had to steal napkins from the dining hall, chew them into pulp and then place a layer of pulp underneath my hard, threadbare pallet, so that my uneasy tossing and turning would press it against the concrete floor and produce enough medium-density fiberboard to support my 8086.
I can't speak for all colleges; but it isn't uncommon to find that, once laptop ownership reaches a certain critical mass, the condition of public PC labs tends to fall right off the bottom of the IT department's priority list. Sometimes formally cancelled, sometimes just allowed to rot.
As for laptop vs. desktop, once you factor in the monitor and peripherals, the delta between a cheap and nasty laptop and a cheap and nasty desktop is pretty small, and being able to work with your study group in the library is pretty handy...
Bad news, I'm afraid. I just sent an HTCPCP PROPFIND request to your coffee maker and recieved a response code: "418 I'm a teapot". On the plus side, that is fully standard compliant. On the minus side, you'd better get ready to like earl grey.
I suspect that, with the downright spooky performance you can wring out of decent array mic designs(not in audiophile terms; but in 'picking perfectly audible human voices out of the background' terms) and the impressive size and relatively low price of MEMS mics, the days of unplugging the mic might be over soon in many cases(cases in both senses, in this case...)
30 seconds of googling pulls up a cute little 3x4mm package, 1mm high, SMT, pick and place compatible, mic and ADC in one package, under 650 microamps operational for the lot. If they didn't still cost $2/ea by the reel, those little bastards(or perhaps some superior competitor, I wasn't exactly comparison shopping) would be sprouting like barnacles on just about anything turing complete and intended for consumer use...
Unless there are additional details not disclosed on their kinda sleazy looking; but unabashed website, the hardware component seems like it is either a mistake(perhaps referring to some sort of anti-theft tag?) or an additional feature specific to this major chain.
If you look at the company's pricing information, it only quotes software licensing fees and prices for additional/replacement/updated install media. No mention of hardware components, much less the sort of model-specific inventory mess that any deep integration would require. Obviously, the information available on the seller's site for that offering doesn't preclude a custom offering for a large customer, with more robust features; but it also isn't as though lying about the existence of super-tough hardware security in order to reduce the risk that your clueless customer tries to have their nephew who "knows computers" install a cracked copy of XP on a "bricked" machine would be a terribly unlikely strategy...
Slashdot Mirror
I apologize if I didn't make it clear: Because of the enthusiasm for televised entertainment, penetration of basic CRT TVs with composite or RF ins has reached(at least on the neighborhood level) down to people who live in shacks and have highly intermittent access to electricity siphoned off the nearest utility pole and no running water.
Now that flat panels have gotten cheap(and a combination of shipping costs, consumer tastes, and environmental concerns have largely eliminated the CRT TV from the first world retail channel), you will find that many low income households have a TV(almost definitely capable of composite in, if purchased in the last couple of years, or going forward, probably a cheap and nasty LCD with an HDMI port). They've already gone out and purchased one. It's a sunk cost. This hypothetical board would be capable of exploiting the money already spent. I do think that they pared it down a bit too far in terms of I/O; but between composite and HDMI/DVI, video output doesn't seem too serious an issue(the only issue would be lack of VGA-HDMI is electrically compatible with DVI-D; but does not include the analog signals broken out by a full DVI connector).
I hate to rain on your journal article; but you may have just discovered some of the coal we accidentally left down there when we finished mining...
Of course, if you live in the middle of nowhere, "receive mail" is only an "archaic form of communication"; because newer ones are also subsidized for you. Rural telecommunications and electrification are also projects that weren't exactly undertaken because the ROI was enticing to Wall Street.
If the objective is efficiency, you might as well tell any rural areas that aren't totally loaded to shove off and learn to enjoy natural solitude, and let any impecunious urban areas enjoy the newfound feeling of community that comes with being cut off.
I'm not sure that that would be such a popular move; but it definitely would decrease the per-capital cost of infrastructure.
The question of interest, I think, is whether the postal service is in the red because they suck, or in the red because their mandate has(with the decline of the letter as a medium) largely shrunk to cover the unprofitable shit work of shipping(Picking up a letter, at your post box, in fucking nowhere, and delivering it to somebody else's postbox in a different fucking nowhere on the other side of the country, for the price of a stamp, is not exactly a lucrative niche...) while FedEx and UPS are free to ignore the low value segments and focus on carrying packages, with an emphasis on larger shippers and aggregated pickups.
In a sense, the real question facing us about the postal service is approximately similar to the real questions behind rural electrification, or telcom access: There are places in the country where providing infrastructure is, per capita, cheap. There are others where providing it is really, really, really expensive. There are areas where the infrastructure customers are relatively wealthy, and ones where they definitely aren't.
We can definitely trust the private sector(as long as they don't gain monopolies or oligopolies) to serve areas where customer willingness to pay is sufficiently high and cost per capita sufficiently low. We then come to the question of what to do about the ones where that isn't the case.
Obviously, this doesn't imply that the postal system is well managed, or that it couldn't do better(and, if improvement is available, it should definitely be undertaken); but, like rural telco and electrification, the fundamental question is not one of wringing out small operational efficiencies; but of whether or not we, as a society, wish to provide a baseline infrastructure to areas where it is not strictly economically justified. Depending on exactly how efficient you are, these areas may be somewhat smaller or somewhat larger; but it will almost always be the case that you could improve financial performance by just writing off your lossy service areas and letting them suck it up.The question is, is that what we want?
At the hardware level, it should be no more or less inscrutable-and-functionally-unfixable than essentially any modern board. At the software level, it probably depends on where the providing entity falls on the "Let them explore" vs. "Lock it down, we'll tell them what they need" spectrum: If you give the user full control over the device, in the spirit of hackerly independent exploration, you'll probably have them show up in fair variety of conditions. If you control more or less tightly, you probably won't.
Barring direct physical destruction of hardware, though, it wouldn't be total rocket surgery to have a rescue bootloader that, say, causes the device to expose its entire internal flash as a USB MSC volume if a particular sequence of inputs is given. Somebody fuck up their stick? Plug it into a computer, enter the rescue sequence, dd if=base_image.img of=/dev/SD_borked_stick Wait a minute or two, unplug, back in business.
Technologically, these should be no harder(and quite possibly easier) to deal with. Philosophically, the tradeoff between making a computer "theirs" vs. making sure that it is ready for class is only partially a technological problem, and much more a problem of educational philosophy(though the OLPC project did some interesting work on technology to reduce the sharpness of the tradeoff...)
The spec list mentions composite output as well, though it doesn't seem to be broken out on the dev board(to a standard connector). All but the nastiest TVs can handle that. In the case of schools, I assume that the use case would be making the computer lab cheaper(even the cheapest nettops run ~$150 on a good day, and Thin client hardware, presumably because of its Big Serious Corporate provenance, can run rather more than that. Bottom end business-line PCs that can be more or less relied upon to have a standard hardware profile are ~$200 in off-lease refurb, ~$500 new). Even if you are using a bunch of Windows only or x86 only software, these little puppies would be substantially cheaper than most thin-client offerings(and linux-on-ARM is supported by Citrix and VMware, and has support for RDP and X, which pretty much covers all the bases). If you are doing things that are supported natively, you could skip the terminal server and go cheaper still.
Plus(somewhat sad to say), there probably are a reasonable number of families where a TV ranks higher than a computer. A basic nasty "HD" LCD setup will almost always come with HDMI now, and is pretty accessible(even if by rent-to-own or usurious credit card borrowing) for people pretty far down the totem pole. Kiddo might well have better luck sneaking in some computer time when the parents aren't watching soaps or sports than having access to a standard computer....
It might be that it is implementing something described in the "On-The-Go and Embedded Host Supplement to the USB 2.0 Specification", which includes the ability for devices to switch between master and slave roles(which would suit the use of a slave-device type connector on the board alongside the fact that the board is driving a hub loaded with slave devices...) My reading of that spec suggests that the OTG device shouldn't be using that particular plug(they are supposed to use microA/B sockets only); but the history of USB is loaded with mechanical abuses of the specs, so that wouldn't be a major surprise...
Not quite as capable, in certain respects, as the Gumstix line of similarly sized ARM boards; but, on the other hand, you'll be lucky to walk away with change from $200 after getting your main board and an I/O expander if needed if you go that route. I wonder where the cost delta comes from?
One minor nit, this system doesn't appear to have any onboard networking(aside from the USB port which, from the picture of it connected to the B port of a hub, would appear to be one of those 'OTG' master or slave jobbies, which could easily enough act as a USB CDC or RNDIS connection to a host PC(which is kind of a waste for a single user; but a basic cheapy desktop loaded with USB cards could easily act as a gateway/fileserver/host for CPU intensive or x86 only programs over an X tunnel for a classroom full of the things)). I have to wonder if a "Flash drive sized" computer that basically doesn't work unless connected to a powered USB hub and a USB network adapter or CDC host PC might be rather less useful than would be a "pack of playing cards sized" computer that actually has a NIC and at least enough USB ports to support a mouse and keyboard(and ideally one extra for miscellaneous purposes)...
Are you calling all nine of these photogenic-but-irrelevant stock-footage models liars? For shame, cad. And they have "secure phone call" technology! That's, like, CIA shit, man. Totally trustworthy. I, for one, eagerly await the chance to enter all SSNs and CCNs into an improperly secured form when I get an email from the "s0ny h3rbal cust0mer Protection Se4vice" asking me to verify.
I offered to license my patented technique of "Apparatus and method for the preparation of a shit-adhesive wall"; but we couldn't come to an agreement. I bet they are sorry now...
The Immigration and Customs enforcement guys were folded into the DHS during the second Bush years.
The problem isn't taxing for road usage(in fact, were it not for the culture that you mention, such a proposal would theoretically be equally desireable for conservatives(who tend to like "users fees" because they dislike taxes and dislike the possibility that people might end up being subsidized)) and liberals(who tend to suspect that individual-vehicle transport, especially the petro kind, has been the recipient of massive, if not overt, subsidies for pretty much the entire post WWII period, making possible an entire suburban material culture that cannot exist without those subsidies).
The problem with mileage taxes is with administration, enforcement, and mission creep. Unlike, say, fuel taxes(which have the fairly convenient advantage of approximately taxing a composite of vehicle size and vehicle miles traveled, which is a good rough estimate of vehicle road 'consumption', without ever having the tax man leave the shop), mileage taxes require, at bare minimum, the tax man inspecting every vehicle's odometer(or at least a sufficiently large sample that most people report theirs honestly on the 1040v.2). If mission creep or ulterior interests come into play, you could pretty easily end up with GPS black boxes, or other less tweakable(you don't even need to crack the odometer, just ensure that the rotations to miles conversion it is using is based on slightly smaller wheels than you are using) methods.
Employing an overhead-heavy, potentially very invasive, taxation strategy when a simple retail sale tax one would work nearly as well strikes me as a serious problem. The notion that, while many roads are basically natural monopolies, and thus cannot be run on free market lines, one can attempt to make one's payment for road use approximately proportional to their use of roads seems entirely sensible.
Arguably, "PCI" the standard/set of requirements is bullshit either way: If a set of requirements designed to force security allows egregious mistakes to be made and/or egregious violators to slip through, it pretty much sucks.
I suspect that anybody who does a competent, good faith, implementation of PCI is at least part of the way toward a secure operation; but PCI isn't intended as polite good advice...
On additional consideration, I'm thinking that there might actually be a more subtle advantage for intel here:
Atom, as a product, has always suffered from two major defects:
It is too high power to profitably compete against ARM(except likely in the embedded/industrial space, which isn't wildly power constrained, except in that fans are bad for reliability in hostile environments, where the fact that you can get an x86 board for $80, or a rugged x86 for under $1k has probably murdered quite a few planned ports to embedded boards based on other architectures...), which cuts Intel out of the smartphone/tablet/STB widget/etc. market.
On the other hand, it is sufficiently cheap, and for many applications sufficiently powerful(especially if Intel were to drop the artificial gimping of screen size and shipping memory) that it threatens the margins of the lower-end of the core 2-derivative business. Thus, intel faces a perverse incentive to keep its performance down, while not being able to push its power consumption down far enough.
Fabbing the "premium" ARM SoC, though, might solve this problem: ARM's existing low-power chops, combined with intel's process superiority should allow them to produce an ARM part for which people will pay a fair premium, thus keeping margins on the part where intel wants them(more or less). And, and here is the icing on the cake, Pretty much no matter how good the ARM SoC they build is, it will take absolute ages to displace their bread-and-butter legacy wintel application-running market.
The world is full of(mostly horrid) legacy software that will never be non-x86. However, since much of it is older, Intel can't afford to make the Atom too good or it will cannibalize the market for core-2s and i3s. ARM, though, by virtue of binary incompatibility, can never cannibalize that market, (and is already cannibalizing the 'light browsing and casual games' market, whether intel likes it or not). Thus, it might actually be a very interesting play for Intel to pare down the low-end x86s as much as competition with AMD will allow, and move its offerings toward two distinct camps of "the highest margin x86 parts we can sell" and "simply the best ARM devices available". That would secure the legacy x86 software base from cannibalization by cheap, low-margin parts, since even the nicest ARM won't be a substitute, while also putting intel in the position to at least make some money from the mobile sector...
Time will tell, I suppose...
An ID that is poor as shit; but legally resident, or with a reasonably clean background, is likely worth something to somebody whose ID is no wealthier; but lacks one or both of those features...
The bottom few rungs of the totem pole are Seriously Ugly.
They call it "The intelligence community". Don't you think that, like any community, it probably has a creepy, lecherous old man somewhere who would patriotically volunteer to work days, nights, and weekends, with nothing but Ramen and lotion for sustenance, to crack the secret of what diabolical plans of terror could be hiding within the depths of 4.6GB of "Bad Burqa Babes: Ankle exposed! vol. 1-237" and 3.2GB of what looks alarmingly like palestinian-occupied-territories-produced amateur femdom/malesub with titles such as "Mossad Lesbians: Haraam Jewesses XXXposed"?
They practice what they call "Communism with Chinese characteristics". Westerners know it as "Crony capitalism" or "what we are busy running headlong into".
The situation is a bit trickier(as I said, this feature is a bit.. oversold); but it does support wireless connections as well. For corporate network purposes, there are ways to provision auth keys to protected memory(so that the embedded management processor can strike up a wireless connection even if the HDD is currently on fire). Off site, it depends on the host OS to strike up a wireless connection(since anything from no auth, to WPA-PSK, to random captive-portal stuff could be at work); but is then capable of initiating a VPN connection back to HQ.
I've actually been a little surprised at how painful it is to use, as an IT management feature(apparently all Intel's good software people work on compilers or chip design tools...); and we've actually largely abandoned bothering with it(on occasion, it can wake a machine when WoL won't; but most of the bells and whistles are really rough around the edges) but its capabilities are a bit creepy. My understanding, as well, is that even more control excitement is Coming Real Soon for systems that have an embedded cellular broadband card: remote control/lock/HDD encryption key purge, at the BIOS level, any time the machine has some power and is near a tower.
From the conversations we've recorded, I can state with 99.35% accuracy, with an N value in the millions, that 9 out of 10 onstar owners are definitely using the preinstalled hardware, and whatever firmware update we prefer. XOXOXO- GM
You had plywood? I had to steal napkins from the dining hall, chew them into pulp and then place a layer of pulp underneath my hard, threadbare pallet, so that my uneasy tossing and turning would press it against the concrete floor and produce enough medium-density fiberboard to support my 8086.
I can't speak for all colleges; but it isn't uncommon to find that, once laptop ownership reaches a certain critical mass, the condition of public PC labs tends to fall right off the bottom of the IT department's priority list. Sometimes formally cancelled, sometimes just allowed to rot.
As for laptop vs. desktop, once you factor in the monitor and peripherals, the delta between a cheap and nasty laptop and a cheap and nasty desktop is pretty small, and being able to work with your study group in the library is pretty handy...
Bad news, I'm afraid. I just sent an HTCPCP PROPFIND request to your coffee maker and recieved a response code: "418 I'm a teapot". On the plus side, that is fully standard compliant. On the minus side, you'd better get ready to like earl grey.
I suspect that, with the downright spooky performance you can wring out of decent array mic designs(not in audiophile terms; but in 'picking perfectly audible human voices out of the background' terms) and the impressive size and relatively low price of MEMS mics, the days of unplugging the mic might be over soon in many cases(cases in both senses, in this case...)
30 seconds of googling pulls up a cute little 3x4mm package, 1mm high, SMT, pick and place compatible, mic and ADC in one package, under 650 microamps operational for the lot. If they didn't still cost $2/ea by the reel, those little bastards(or perhaps some superior competitor, I wasn't exactly comparison shopping) would be sprouting like barnacles on just about anything turing complete and intended for consumer use...
UVC only came about a decade late; but it has been quite helpful. Before UVC, we had One Man....
Unless there are additional details not disclosed on their kinda sleazy looking; but unabashed website, the hardware component seems like it is either a mistake(perhaps referring to some sort of anti-theft tag?) or an additional feature specific to this major chain.
If you look at the company's pricing information, it only quotes software licensing fees and prices for additional/replacement/updated install media. No mention of hardware components, much less the sort of model-specific inventory mess that any deep integration would require. Obviously, the information available on the seller's site for that offering doesn't preclude a custom offering for a large customer, with more robust features; but it also isn't as though lying about the existence of super-tough hardware security in order to reduce the risk that your clueless customer tries to have their nephew who "knows computers" install a cracked copy of XP on a "bricked" machine would be a terribly unlikely strategy...