In fairness to Apple, they have been working to improve the situation, and things are better now; but the state of the possible when this program started(~2 years ago) was rather less pleasant. They started tightly wedded to the 'device basically has one user, who has an account directly with Apple and a CC number on file' model; and it has been a rather slow path to getting support for a model where things like 'applications owned by the institution' actually works smoothly.
Apple's first-party support for remote management is still better than Android's; but their grip is tight enough that it is them or nothing, while Android is all over the map; but 3rd parties can actually offer options without the keys to the OS.
Wow, asking you to do the work so that they can deliver a sales pitch is really, really, nervy.
Are you running something in-house(or off the shelf but fairly heavily specialized) enough that they couldn't just put together an equivalent test environment on their end and use that for the sales pitch, or are they actually that lazy and entitled?
We certainly deal with doing the various things required to make what our users choose work; that's sort of what they pay us for; but I wouldn't have imagined a salesweasel demanding that I set up their tech demo for them.
Hasn't the future of 'open' android always looked bleak, more or less by design? At the bottom of the stack, we have SoC vendors who don't give a damn, handset OEMs who don't give a damn and/or actively prefer that older handsets remain as outdated as possible so you'll buy something new, and carriers who have largely the same incentives as handset vendors; but with their own crapware. This ensures that hardware support is spotty and typically weak for anything except whatever the device shipped with(and it's a moot point on the cryptographically locked devices). Markedly worse than the PC world in terms of vendor helpfullness or ability to do much of anything without a BSP or cobbling blobs together from vendors with slightly longer update support windows.
From the top of the stack, the 'free' parts of android are basically Google's hardware abstraction layer for google play services, and getting steadily more so.
I can't digg up the original contract to check; but some of the stories state that they are going to Apple because the deal was to purchase 'iPad+software', as a packaged product, from Apple. By all accounts Pearson was the significant weak link (not a shock, that's pretty typical for them), while Apple's stuff suffered only from the fairly pitiful state of iOS management; but the school district didn't structure the deal as 'Contract #1, buy ipads, Contract #2, buy textbook apps'; it was a package, and their claim is that half the package was rotten and the other half is of little use to them without the underdelivered component.
Given that Apple is reputed to be a brutal and efficient taskmaster of its suppliers, I'd imagine that either the school district will fail, or Apple will gouge it out of Pearson; but to the best of my understanding there is logic behind complaining to Apple, given the terms under which the devices were purchased.
If you are working on the (probably bullshit) theory that these devices improve educational outcomes, there is a civil rights interest in ensuring that students who made poor prenatal choices still have the opportunity to get a decent education.
However, that assumption is under-supported(even if they were free, it's not news that electronic gizmos are good for slacking off with, so they might have a negative effect unless the school actually has a good plan in mind; and since they aren't free, they are being chosen to the exclusion of other possible educational aids), and if it doesn't happen to be true; then there isn't much of a civil rights case for access to toys. If anything, devices for slacking off probably amplify the effects of differing qualities of home life, since parental attention will have a major effect on how much slacking you can get away with.
The superintendent at the time 'resigned' over the controversy; but depending on the outcome of the FBI's ongoing investigation into the circumstances of the bidding process, he may or may not be looking at further consequences.
Pearson is a company that brings a sort of defense contractor vibe to the educational sector. They are huge, superb at landing contracts, excellent at writing contracts that promise somewhat less than they appear to; but not so hot on delivering, much less on time or on budget.
Anyone buying a zillion ipads for school children without realizing that they'll be using them mostly to screw around on the internet within about five minutes is certainly an idiot; and Pearson certainly can't take the blame for that; but their failure to deliver some curriculum slurry and a terrible textbook app or two within the agreed upon time? That's the sort of thing they do.
It would not much fail to surprise me if it wasn't done this way; but something like those seatback location/direction displays require relatively little data transfer(you wouldn't need more than the 4800 baud NMEA spew you'd get from a standard GPS device, and you could likely get away with less) and no responses from the seatback unit; so you could do everything you'd need over an isolated, intrinsically unidirectional, link.
Put the avionics on the emitter side of an optoisolator, blindly blinking out location and heading data, the controller for the seatback entertainment system on the receiver side, listening, and you get an arrangement where there simply isn't anything to attack at the software level(you could probably hose in flight entertainment for the entire aircraft one way or another; but boredom isn't very lethal); and where physical attacks might be possible; but (by choosing an optoisolator and the location of the interface between the critical and noncritical side) can be made quite difficult.
That, plus iterative improvement by using the mk2 rock you produced by banging mk1 rocks together to shape a mk3 rock, and so on; is pretty much the truth.
Ants are awesome; but really a different flavor: they(along with termites) manage to get extremely impressive results from emergent behavior among swarms of really, really, dumb individuals. Very cool, particularly fascinating if you are trying to get good results from multiple agents without tacking on an unweildy command and control system.
The sort of tool use in TFA is interesting because it suggests fairly advanced cognition(and sometimes communication and transmission of learned techniques). Ants aren't so hot at that.
Oh no, his behavior was a flagrant violation of the rules; and they can selectively destroy his ability to interact with 8x8 black and white tiled surfaces using a cold war mycotoxin for all I care.
If anything, my intended thesis(that even a relatively weak or computationally limited computer could be a substantial aid to a reasonably skilled human) suggests that even modest machine assistance is quite dramatic cheating in terms of allowing you to beat people above your skill level. Being reasonably good just serves to make the computer's job markedly easier in this case, which then allows it to make your job markedly easier, which defeats the point of involving humans at all.
I don't know what the state of the art is(and it would presumably vary a bit depending on whether the phone was running the analysis or just acting as a nice UI for a remote server); but it's possible that he was using the computer to augment competent-or-better human play; but not replace it entirely.
Chess has a large enough search space that full scale brute force and ignorance is quite a challenge; but more constrained states(like the actual state of the board partway through a game, or after the use of one of the standard openings) have correspondingly smaller search spaces. A computer program could also assist in checking proposed moves for 'something dumb you would usually recognize; but sometimes miss under stress'. Again, by being one move 'in the future'(since you are just testing, without commitment) you reduce the difficulty of the analyzing the game; but get potentially useful error checking.
I don't know how thoroughly, in terms of human level of play, the game has been beaten with various amounts of hardware at your disposal; but it is definitely beaten enough that a good player with access to a machine is likely to play a better game, possibly a markedly better one.
Well, if you have a taste for brutal irony, newer revisions of the eDP spec include the option to use lossy compression(but it's, um, 'visually lossless', we swear!) to reduce the amount of data you need to send to the screen, and the power costs of the link... Where better than the suckers overpaying for resolution to introduce such a feature?
This is true; but if you've bothered to produce a small, very high resolution, display anyway; you'd have to really not care in order to be unwilling to sell some of the product you already make to somebody willing to buy it.
That likely excludes 'quantity: 1' orders from random hobbyists; but if you can hit the minimum order quantity, your money is a lot more important than your intended application.
The curious quirk, in this case, is that you probably won't even be able to get a 'theoretically better; but not perceptably so and definitely not worth the price' product; you'll almost inevitably get a worse one.
Any phone/tablet SoC with claims to being remotely high end is already some mixture of thermally constrained and deliberately crippled to save the device's battery life. If you demand their full performance, they'll throttle within minutes; and if they somehow had the thermal headroom to avoid that, they'd flatten the battery in a some egregiously short time.
Assuming reasonably equal tech(ie. not a 1920x1080 phone from two years ago against a phone from next year with this screen) the higher resolution device will have worse battery life(or a visibly larger battery) and be at a greater risk of annoying frame rate/responsiveness issues in any applications that try to do complex GPU work at native resolution. Some amount of this is accepted, since visible giant eyeball-slashing pixels suck; but the returns on graphical prettiness diminish, while the power and thermal costs just keep on scaling...
At least audiophile nonsense is generally good at what it does, if you ignore the price tag and the nonsense; this will be actively worse than a similar device based on a slightly less ambitious screen.
Speaking of updating outdated regulations... Is it time to give some thought to the amount of precious, precious, spectrum we dedicate to broadcasting low quality audio using extraordinarily archaic techniques? Sure, I appreciate being able to tune in to talk radio with nothing but a chunk of germanium and the patience to poke it until it agrees to start rectifying; but I need a better reason than that to operate a dinosaur preserve.
What principled justification would there be for excluding 'religious' and 'talk' stations from payment? One would think that any 'religious' station would either be a nonprofit or deserve to pay like any other business; and 'talk' is huge business, and presumably not a terribly heavy consumer of music.
I can take a few guesses about the pragmatic political considerations for those exemptions; but they aren't exactly complementary.
If you plan on having the medium survive your house burning down, it'll either have to be something really exotic(CNCed cuneiform tablets?) or something boring inside a sufficiently fireproof safe (which can get costly; but are a well recognized product category).
If it gets to the point where the fire and/or water are in contact with your storage medium, luck might save you; but the odds are lousy enough that it doesn't really qualify as a plan.
You really should consider off-site storage. This doesn't have to mean 'in the cloud', anything that gets updated very infrequently can be dumped to some backup medium and shoved in a safe deposit box.
ark is helpful, if not always detailed enough; but it make Intel model numbers non-cryptic in much the same way that DNS makes IP addresses human readable.
So, when ICANN floated the gTLD idea, everyone told them that it was pointless bullshit that would only end in trademark wrangling, shakedowns, and vast swaths of slum domains used for little more than scamming.
They decided to go ahead anyway.
Now they are shocked, hurt, and betrayed that someone would be using one of the new TLDs for less than upstanding purposes. What utter fools.
Intel model numbering has often been a bit cryptic, and worse more recently as they've spawned new product lines and taken advantage of their lead over AMD by market-segmenting with incredible precision, producing parts that differ by a single feature enabled or disabled, or have the same clock speed but different 'turbo' speeds, or any number of similar permutations.
As though that isn't enough fun, now even expert level knowledge of the model numbers won't tell you how fast it is because the OEM can gimp it to suit their chassis design. It's a good thing that basically all modern CPUs are really fast, or this would be downright depressing.
There's no market for video of 'banal interaction between officer, person, goes boringly', except possibly as stock footage; but there would definitely be an appetite for footage of people getting shot by cops under justified circumstances. The broadcast media are a little squeamish about showing kills, so they'd be a little more circumspect; but online? hell yeah. It's not as though Russian dash-cam footage is newsworthy in the slightest, and that's a youtube phenomenon. Lovingly curated collections of perp snuff film would be like candy.
If you were running a comparison based on media hits, you'd have to compensate for the fact that justified shootings would tend to be boring local news, while especially lurid unjustified ones would be national; but online video availability and local news crime blotter coverage would reflect the availability of video in either case.
Slashdot Mirror
In fairness to Apple, they have been working to improve the situation, and things are better now; but the state of the possible when this program started(~2 years ago) was rather less pleasant. They started tightly wedded to the 'device basically has one user, who has an account directly with Apple and a CC number on file' model; and it has been a rather slow path to getting support for a model where things like 'applications owned by the institution' actually works smoothly.
Apple's first-party support for remote management is still better than Android's; but their grip is tight enough that it is them or nothing, while Android is all over the map; but 3rd parties can actually offer options without the keys to the OS.
Wow, asking you to do the work so that they can deliver a sales pitch is really, really, nervy.
Are you running something in-house(or off the shelf but fairly heavily specialized) enough that they couldn't just put together an equivalent test environment on their end and use that for the sales pitch, or are they actually that lazy and entitled?
We certainly deal with doing the various things required to make what our users choose work; that's sort of what they pay us for; but I wouldn't have imagined a salesweasel demanding that I set up their tech demo for them.
Hasn't the future of 'open' android always looked bleak, more or less by design? At the bottom of the stack, we have SoC vendors who don't give a damn, handset OEMs who don't give a damn and/or actively prefer that older handsets remain as outdated as possible so you'll buy something new, and carriers who have largely the same incentives as handset vendors; but with their own crapware. This ensures that hardware support is spotty and typically weak for anything except whatever the device shipped with(and it's a moot point on the cryptographically locked devices). Markedly worse than the PC world in terms of vendor helpfullness or ability to do much of anything without a BSP or cobbling blobs together from vendors with slightly longer update support windows.
From the top of the stack, the 'free' parts of android are basically Google's hardware abstraction layer for google play services, and getting steadily more so.
I can't digg up the original contract to check; but some of the stories state that they are going to Apple because the deal was to purchase 'iPad+software', as a packaged product, from Apple. By all accounts Pearson was the significant weak link (not a shock, that's pretty typical for them), while Apple's stuff suffered only from the fairly pitiful state of iOS management; but the school district didn't structure the deal as 'Contract #1, buy ipads, Contract #2, buy textbook apps'; it was a package, and their claim is that half the package was rotten and the other half is of little use to them without the underdelivered component.
Given that Apple is reputed to be a brutal and efficient taskmaster of its suppliers, I'd imagine that either the school district will fail, or Apple will gouge it out of Pearson; but to the best of my understanding there is logic behind complaining to Apple, given the terms under which the devices were purchased.
If you are working on the (probably bullshit) theory that these devices improve educational outcomes, there is a civil rights interest in ensuring that students who made poor prenatal choices still have the opportunity to get a decent education.
However, that assumption is under-supported(even if they were free, it's not news that electronic gizmos are good for slacking off with, so they might have a negative effect unless the school actually has a good plan in mind; and since they aren't free, they are being chosen to the exclusion of other possible educational aids), and if it doesn't happen to be true; then there isn't much of a civil rights case for access to toys. If anything, devices for slacking off probably amplify the effects of differing qualities of home life, since parental attention will have a major effect on how much slacking you can get away with.
The superintendent at the time 'resigned' over the controversy; but depending on the outcome of the FBI's ongoing investigation into the circumstances of the bidding process, he may or may not be looking at further consequences.
Pearson is a company that brings a sort of defense contractor vibe to the educational sector. They are huge, superb at landing contracts, excellent at writing contracts that promise somewhat less than they appear to; but not so hot on delivering, much less on time or on budget.
Anyone buying a zillion ipads for school children without realizing that they'll be using them mostly to screw around on the internet within about five minutes is certainly an idiot; and Pearson certainly can't take the blame for that; but their failure to deliver some curriculum slurry and a terrible textbook app or two within the agreed upon time? That's the sort of thing they do.
It would not much fail to surprise me if it wasn't done this way; but something like those seatback location/direction displays require relatively little data transfer(you wouldn't need more than the 4800 baud NMEA spew you'd get from a standard GPS device, and you could likely get away with less) and no responses from the seatback unit; so you could do everything you'd need over an isolated, intrinsically unidirectional, link.
Put the avionics on the emitter side of an optoisolator, blindly blinking out location and heading data, the controller for the seatback entertainment system on the receiver side, listening, and you get an arrangement where there simply isn't anything to attack at the software level(you could probably hose in flight entertainment for the entire aircraft one way or another; but boredom isn't very lethal); and where physical attacks might be possible; but (by choosing an optoisolator and the location of the interface between the critical and noncritical side) can be made quite difficult.
That, plus iterative improvement by using the mk2 rock you produced by banging mk1 rocks together to shape a mk3 rock, and so on; is pretty much the truth.
Ants are awesome; but really a different flavor: they(along with termites) manage to get extremely impressive results from emergent behavior among swarms of really, really, dumb individuals. Very cool, particularly fascinating if you are trying to get good results from multiple agents without tacking on an unweildy command and control system.
The sort of tool use in TFA is interesting because it suggests fairly advanced cognition(and sometimes communication and transmission of learned techniques). Ants aren't so hot at that.
Oh no, his behavior was a flagrant violation of the rules; and they can selectively destroy his ability to interact with 8x8 black and white tiled surfaces using a cold war mycotoxin for all I care.
If anything, my intended thesis(that even a relatively weak or computationally limited computer could be a substantial aid to a reasonably skilled human) suggests that even modest machine assistance is quite dramatic cheating in terms of allowing you to beat people above your skill level. Being reasonably good just serves to make the computer's job markedly easier in this case, which then allows it to make your job markedly easier, which defeats the point of involving humans at all.
His Nexus 6 wrapped in toilet paper was hidden in the adjacent stall...
I don't know what the state of the art is(and it would presumably vary a bit depending on whether the phone was running the analysis or just acting as a nice UI for a remote server); but it's possible that he was using the computer to augment competent-or-better human play; but not replace it entirely.
Chess has a large enough search space that full scale brute force and ignorance is quite a challenge; but more constrained states(like the actual state of the board partway through a game, or after the use of one of the standard openings) have correspondingly smaller search spaces. A computer program could also assist in checking proposed moves for 'something dumb you would usually recognize; but sometimes miss under stress'. Again, by being one move 'in the future'(since you are just testing, without commitment) you reduce the difficulty of the analyzing the game; but get potentially useful error checking.
I don't know how thoroughly, in terms of human level of play, the game has been beaten with various amounts of hardware at your disposal; but it is definitely beaten enough that a good player with access to a machine is likely to play a better game, possibly a markedly better one.
Those won't really be 'advances' until we have micro-fusion cells and sintered unobtanium nanotube heatsinks.
Well, if you have a taste for brutal irony, newer revisions of the eDP spec include the option to use lossy compression(but it's, um, 'visually lossless', we swear!) to reduce the amount of data you need to send to the screen, and the power costs of the link... Where better than the suckers overpaying for resolution to introduce such a feature?
This is true; but if you've bothered to produce a small, very high resolution, display anyway; you'd have to really not care in order to be unwilling to sell some of the product you already make to somebody willing to buy it.
That likely excludes 'quantity: 1' orders from random hobbyists; but if you can hit the minimum order quantity, your money is a lot more important than your intended application.
The curious quirk, in this case, is that you probably won't even be able to get a 'theoretically better; but not perceptably so and definitely not worth the price' product; you'll almost inevitably get a worse one.
Any phone/tablet SoC with claims to being remotely high end is already some mixture of thermally constrained and deliberately crippled to save the device's battery life. If you demand their full performance, they'll throttle within minutes; and if they somehow had the thermal headroom to avoid that, they'd flatten the battery in a some egregiously short time.
Assuming reasonably equal tech(ie. not a 1920x1080 phone from two years ago against a phone from next year with this screen) the higher resolution device will have worse battery life(or a visibly larger battery) and be at a greater risk of annoying frame rate/responsiveness issues in any applications that try to do complex GPU work at native resolution. Some amount of this is accepted, since visible giant eyeball-slashing pixels suck; but the returns on graphical prettiness diminish, while the power and thermal costs just keep on scaling...
At least audiophile nonsense is generally good at what it does, if you ignore the price tag and the nonsense; this will be actively worse than a similar device based on a slightly less ambitious screen.
Speaking of updating outdated regulations... Is it time to give some thought to the amount of precious, precious, spectrum we dedicate to broadcasting low quality audio using extraordinarily archaic techniques? Sure, I appreciate being able to tune in to talk radio with nothing but a chunk of germanium and the patience to poke it until it agrees to start rectifying; but I need a better reason than that to operate a dinosaur preserve.
What principled justification would there be for excluding 'religious' and 'talk' stations from payment? One would think that any 'religious' station would either be a nonprofit or deserve to pay like any other business; and 'talk' is huge business, and presumably not a terribly heavy consumer of music.
I can take a few guesses about the pragmatic political considerations for those exemptions; but they aren't exactly complementary.
If you plan on having the medium survive your house burning down, it'll either have to be something really exotic(CNCed cuneiform tablets?) or something boring inside a sufficiently fireproof safe (which can get costly; but are a well recognized product category).
If it gets to the point where the fire and/or water are in contact with your storage medium, luck might save you; but the odds are lousy enough that it doesn't really qualify as a plan.
You really should consider off-site storage. This doesn't have to mean 'in the cloud', anything that gets updated very infrequently can be dumped to some backup medium and shoved in a safe deposit box.
ark is helpful, if not always detailed enough; but it make Intel model numbers non-cryptic in much the same way that DNS makes IP addresses human readable.
So, when ICANN floated the gTLD idea, everyone told them that it was pointless bullshit that would only end in trademark wrangling, shakedowns, and vast swaths of slum domains used for little more than scamming.
They decided to go ahead anyway.
Now they are shocked, hurt, and betrayed that someone would be using one of the new TLDs for less than upstanding purposes. What utter fools.
Intel model numbering has often been a bit cryptic, and worse more recently as they've spawned new product lines and taken advantage of their lead over AMD by market-segmenting with incredible precision, producing parts that differ by a single feature enabled or disabled, or have the same clock speed but different 'turbo' speeds, or any number of similar permutations.
As though that isn't enough fun, now even expert level knowledge of the model numbers won't tell you how fast it is because the OEM can gimp it to suit their chassis design. It's a good thing that basically all modern CPUs are really fast, or this would be downright depressing.
I believe you mean 'enable all Apple devices for socially harmonious lawful remote management'.
Hey, they weren't spying on our SSLed services today, so we still totally trust them! Also, have you seen how lucrative the Chinese market could be?
There's no market for video of 'banal interaction between officer, person, goes boringly', except possibly as stock footage; but there would definitely be an appetite for footage of people getting shot by cops under justified circumstances. The broadcast media are a little squeamish about showing kills, so they'd be a little more circumspect; but online? hell yeah. It's not as though Russian dash-cam footage is newsworthy in the slightest, and that's a youtube phenomenon. Lovingly curated collections of perp snuff film would be like candy.
If you were running a comparison based on media hits, you'd have to compensate for the fact that justified shootings would tend to be boring local news, while especially lurid unjustified ones would be national; but online video availability and local news crime blotter coverage would reflect the availability of video in either case.