The fundamentals behind sandboxing and hardware virtualisation date back from the time of big iron mainframes.
The only novel thing is that this is applied to a personal computer/workstation (to a newer type of hardware that didn't exist back then, but is basically just smaller).
I pay $5,000 for a shoot and 24 hours after I put it on the site some asshole puts it on the tubes
Definite proof that what they need, is some sort of patronage/crowd-sourcing. Get 200 guys pay an average of 25$, the highest bidders get to make request (so they can fullfill whatevet they want they could not easily find elsewhere) - biggest bidders get to take a small part in production (either technical, like holding light. Or a small cameo in a scene. Whatever floats their boat) Then it doesn't matter if the film is on porn hub the next day, he already got his 5000$ back before he even started.
Kind of how the request for weird fetishes on the image board work. (Or some niche-fetish porn site offer paid requests)
If he cannot get 5'000$ in *advance* then there's slim chance that he'll ever be able to them after-ward. If nobody is interested enough to pay to have him film it, he should instead try to think of something original enough to attract attention (and money). Or switch job.
In an over simplified-way that might make physicist angry... ("I am a doctor, Jim ! Not a quantum physicist !")
Design a machine that fires approximately 1 photon per second. Measure and confirm rate of photo firing. Then put double slit in front. Expose picture. As during any given second, there's (an avarage) maximum of only photon, then (in average situation) this single photon should not have any other to interact with. Classic physics should predict only a (predominant) picture of the slit with (nearly no) pattern at all (except for the few outlier situation where 2 photons ended up in flight due to imprecision, but then their pattern should be much fainter). Quantum physics should predict that even if (most of the time) only one photo is in flight, you still see a diffraction pattern predominantly. (and not only faintly for the couple of mis-firing)
In short : do the measurement AND the experiment at 2 different times, otherwise the measurement will destroy the photo.
Now again, this is NOT my area of expertise, so I probably made have of the physicists on/. cringe (and the other half laugh uncontrollably).
There is also a difference in the way you interpret the data.
Redlight and speedcams work very precisely because each vehicle has a license plate, matching an entry into a vehicle database. Either you have a matching number, that should match an exact vehicle (or a counterfeit license plate), or you don't. Identity is binary. Also the devices are usually calibrated to guarantee precise results regarding measurement. in other words, you can at least reliable trust the information (platenumber ; speed) coming out of photo by a speed-cam (then, whether the license plate matches the actual car or was tampered with, and whether the car's registered owner was driving or some 3rd party, is an entirely different question)
Image analysis only give you a likelihood of resemblance. Footage of person may or might not look like the suspect. It has the same kind of imprecision as any visual witness. Some are better than other at identifying suspect, but nobody is perfect. The only certitude you have is that the person doing crime on the video tape looks more or less like a given suspect. The suspect *might be* the perpetrator. But the suspect *might also be* a random look-alike that looked close enough to the person on the tape to fool the algorithm by chance. That's valid both for visual clues (and that even if technologies becomes better at distinguishing people's image than regular witnesses) and for DNA clues (given a big enough pool of DNA samples, any methodology - short of full genome sequencing and that doesn't take into account twins - will eventually start to show random matches) So information from cam can only be used as clues ("a person with a description matching the suspect has been recorded on tape to perpetrate a crime") and compared with other information (alibi: the suspect was at a party at that time and everyone else there saw him. The tape recorded a look-alike).
TL;DR: Technology is never going to replace good old detective work, only give extra information and tools.
That's unless government makes it mandatory to shave everyone's head and tatoo a QR-Code with the SSN.
Some companies can allow that. Many more cannot. Good luck assembling products or working in a restaurant while "working" at home. Hard for a nurse to take care of patients or for a construction worker to build a road from home.
Well, when I was speaking about work from home, i was more thinking about the typical/. crowd, working IT jobs or science jobs. Of course for other people, that might instead be staying at home on a sick leave.
For many people (esp hourly people) missing work is not really a viable option even when sick.
Few employers provide enough paid sick leave to mitigate this problem. At my place of work people have to be considering a hospital visit before they will stay away.
Two very important things:
1- There are countries where sick leave days are the norm. It is mandatory for both employee and companies to be covered by health insurance that will pay for sick days. usually up to 2 days are "no question asked", beyond that a visit at the doctor is mandatory.
But yeah, I now : that is for us "evil communist european countries". Not for your land of the free, make america great again.
2- now, from a purely business logic: which situation does make more sense economically :
A. company doesn't pay any sick days. it's employee try to stay working no matter what. so in case of Influenza virus outbreak they stay at work. they don't treat their sickness, so eventually they are completely useless for a whole week (thanks to flu symptoms), but you dtill have to pay them, just because theu showed up. and by being at work while sick, they contaminate the rest of your workforce l. so everybody ends up being un productive for at least 1 week (except for the few people that could afford to get vaccinated againdt influenza A virus) (and that's hoping that this years strain was correctly predicted).
That is going to cost a lot of productivity to the company.
B. company ponies up and is able to pay at least a pay a few sick day leaves. A sick employee can afford to stay at home. They'll still be away for one week, but at least they won't transmit the disease much and there are only a few employee being absent.
Overall, it costs a bit to the company, but productivity doesn't dip that much down during each year's influenza outbreak.
C. health insurrance mandatory for every one. It costs a bit to the employer and employee to have a health plan. But at least the health plan can cover sick days. And drugs expenses and medical visit. In case of influenza, no only can the employee afford to stay home (and thus avoid to contaminate everyone else), but they can even afford to go to the doctor early enough (not only waiting until it has to be hospital) and thus get proper anti-viral therapy. Meaning that the employee is only away for 1 or 2 days while only having a very mild case if flu, instead of being debilitated by the symptoms and having to stay a whole week in the bed.
So it costs a bit to the componay and the employee, but productivity has barely any noticeable dip, even during outbreaks.
Also, lots of health plan also reimburse preventive medecine, and thus much more employee can afford to get vaccinated. Therefor even less employee are actually concerned by the seasonnal flu outbreak.
Overall, this is also the best scenario for the quality of life of the employee, as they get less sick and are in general better health (also meaning that they don't die too early, and that you don't risk losing trained workforce)
Ultra liberal USA went for A, because "it's not my fault if you're irresponsible and can't afford your health !" type liberals.
The rest of the developped world went for C - but we know, we are freedom-hating socialist scum.
Current Renault Zoé (based on the same platform, but a Renault Clio body bolted on it) : 41kWh : - costs ~9'000$ approx when you decide to buy the battery with the car (as opposed to rent one) - costs ~15'000$ if you decide to buy a new one later on (because you initially opted renting one instead).
According to the summary : that battery would cost ~4100$ approx to built. (Given that the company need to recoup its R&D costs, and the various subcontractors making the batteries (LG Chem, Imecar, Kreisel Electric, etc.) are all building their own equivalent of a megafactory in Europe, the price might seem seems legit).
Does he expect me to just test every morning when I get up from bed?
not every morning, and not everyone. but if the technology can become cheap enough, some at risk situations could be monitored (non vaccined people, who are in contact with lots of others, when there's an active outbreak around).
and there's an immense benefits in detecting early: - have the currently not yet symptomatic sick stay at home to avoid spreading it further (with most viruses, including influenza, you could be transmitting before you start feeling symptoms). - treat it with specific anti-virals (for influenza: the sooner you start, the better the chances of it being effective. starting more than 48h after the onset of symptoms won't make much a difference - you'll still be dick. being able to begin therapy before even the symptoms starts will surely reduce the impact of the flu) thus, if such a tchnology works, you could replace "laying in bed for 1 week" by "working from home for a couple of days" (a lot better quality of time) and from an employer's perspective, it means that 1 sick guy wont contaminate your whole company (a lot less lost days) (of course, for that you need a technology that is precise enough. i somewhat feel that if the analysis is limited to few gazes mentioned, it might tire of too many false alerts).
We've known what causes people to get sick (germs)
(...which is a gross over simplification. disease are caused by entirely different types of bugd: viruses, bacteria, funghi, protozoa, microscopic animals like worms, etc.)
for over a hundred years now. But it's amazing how many people still put their faith in things like health food diets.
Well, eating properly cooked food and drinking clean water can be proven to reduce some diseases
With proper detection and quarantine the flu could be eliminated from the human population almost entirely.
fat chance.
some diseases which are 100% human-targeting have indeed been eliminated (and you need to add vaccine on the list of things used against them).
but influenza affects tons of other animals (e.g: pigs) of which some are wild (e.g.: birds). And its going to be nearly impossible to control the spread of viral infections among (wild) migrating birds.
you can give specific anti virals to cases of Influenza detected early enough, you can quarantine to diminish transmission when sick (i.e. staying at home when sick.) but you can still have transmitted before starting to feel the symptoms), you can vaccinate to slow down its spread (hoping that the correct sub-type was predicted).
but you can't easily prevent animals carrying the influenza to pass it to human.
(well, in practice it's a tiny bit more complicated than just "passing". the different strains need to hybridize. but you got the general idea).
Personal theory: Most people, consciously or subconsciously, want to be LESS connected to their phones, not more connected.
Also another theory : it boils down to specs. And actually to too many of them. To the point that "trying to get 4G" functional, became detrimental to the core functionality of a watch (tells you the time, without you needing to constantly charge it). Some people might be interested into "a regular watch, that can also give you alerts". But not as many are interested into "we managed to cram a whole smartphone around your wrist !"
Pebble, the kickstarter campaign that got a bat shit insane success raising tremendously more money than initially hoped for, to the point that they got noticed by other manufacturer (like Apple, LG, Samsung, etc.) who realised that smartwatch are "a thing" (again) and thus kickstarted the (current) era of smartwatched : They went for minimalistic specs. Putting as little processing power in the watch itself, counting on the linked smartphone to do any actually heavy computing, counting on e-Ink to even further lower the power consumption. Their point: making a watch, that basically behaves like a watch (you don't need to charge your dumb wrist watch every single night), but has some interesting status display. Basically, they just went for the next logical small-step evolution after the various Citizen's bluetooth-connected watches. (And their result: tremendous success, given the kickstarter result, at time when smartwatch weren't yet considered "a thing" (again))
The other manufacturer saw the success, realised that smartwatch were popular (again), and decided to give another try at it. And attacked it through the same angle they have taken the habit with their smartphone : Pump up the specs to try to beat the competition with a product that looks better on the paper.
More CPU power, better animated OLED colourful and luminous displays, extra health sensors (well those could actually be useful) touch screen instead of button, more connectivity, including Wifi or even Cellular (WTF ?!? You already have such a device in your pocket !), able to answer SMS and Tweets directly from the smartwatch itself (?!), etc. All this to the expense of useful things like : actual battery life.
On the paper, these watches managed to look cool and attractive (so they got relative success) but in practice they are a little bit underwhelming (so after buying them, people didn't go batshit crazy).
So you end up with a small early player that got crushed by everyone else (Pebble).
And a bunch of smartphone makers that try to replicate a whole smartphone, sized down to fit around your wrist.
A few of them managed to get the lead in the market (Apple, probably because someone in their R&D lab was already musing with the idea back at the time of iPod Nano 6gen) but ended up with product that, although they tick every box on the feature check list, don't compare positively with regular watch (show the time, without needing a battery replacement for at least a couple of months).
And, whereas the Apple iPhone 1 managed to convince people that it's suddenly okay to have a cell phone that can go through the day without needing a charge, they didn't manage to pull the same trick with watches that need nightly recharge instead of holding months between battery replacement.
Thus, not everyone wanted to have an Apple Watch.
Because every other manufacturer (beside Pebble) jumped to big specs, but that ended-up being detrimental to the key feature that attracts most people to watch (able to always show time, needing a fresh battery only once every few months).
Another myth that these studies have debunked is that HDDs do better if kept cool. Actually, failure rates are lower for disks kept at the higher end of the rated temperatures. This is one reason that Google runs "hot" datacenters today, with ambient temps over 100F.
Funnily though, I've had very good success with IBM Deskstars (back during the infamous era of "Deathstars" click death) simply by running them cooler. Though again, I did only have a few. So I only have anecdotal evidence. Maybe I was just lucky to have the few Deskstars that didn't went "Dearthstar". But maybe thermal management was indeed exceptionnally a problem on these old drives.
If taken far enough, Romanian carriers will have to raise their prices to account for their share of Danish infrastructure costs, but that means all Romanians would then be paying those costs (while still on lower Romanian salaries) while Danish tax and government budget is being undermined.
The current situation, the prices within country aren't that much high. (a few eurocents difference in the price per minute). Maybe a users calling over a Romanian carriers within Romania will end up paying a few euros less than an users over a Danish carrier within Denmark. I.e.: there is some variation between countries, but not as much as you would think, and specially not as much when compared to :
Roaming costs: They are currently outrageously high. When abroad (e.g.: the Danish users goes to Romania) the costs of roaming can be 10x the normal price (you could be paying *whole euros* per minute - even when, per your reasoning the cost of Romania's infrastructure is certainly not the 9x price difference).
The current situation isn't about asking a couple of cent more to make up for the difference in infrastructure that you mention by donating to the roaming country. The current situation is about milking the travelling users as if they were gold mines. (It probably made sense for the carrier back in a time of less mobility, when probably most of the users roaming abroad would be business chief-whatever-officers type which don't pay much attention to the phone bill that their work is paying) (Nowadays it just painful for all people travelling abroad because everybody has a wireless phone, because smartphone are nearly necessary for lots of everyday tasks, specially when abroad and you don't want to lug your laptop around, and modern society comes with even more geographical mobiliy)
This happens even within the same carrier. E.g.: buy a contract from O2 Germany. And whenever you travel to Czech republic, get charged outrageously, even when connecting to O2 CZ (the same company).
and it's easy enough to just get a local SIM card if you are travelling alot or for extended time,
Cue in mental picture of an avarage european user carrying 10 different SIM cards, and using a phone that has 4 SIM slots, with 2 of them always simultaneously on air.
If there were mass popular demand for it, carriers would already offer at least limited versions of it (potentially most popular in small countries or regions where travel to nearby countries is routine).
And there are such offer : there are "virtual carrier" (that don't own their own network of antenna, but piggy back on other carriers which they have an arrangement with) who have signed agreement with most big carriers in all European countries (it's not as impossible as it sound. There aren't really that many different carriers in Europes. Most of the time it's the same few international companies going each by different name/brands in each country. - see the above O2 example - Sign just a few of those big internationals, and you have deals in nearly every country). By doing this, this virtual carrier are never actually roaming, you've always got at least one local operator with whom your virtual carrier has a deal and you always pay the local price. as a random example : xxSIM
The proposed law is about always forcing this situation on all carrier, no excuse accepted.
Today, the vast majority of Linux kernels running are running on ARM processors.
And on MIPS. a.k.a. "That other CPU that you'll also find in router and other embed devices, beside ARMs".
But yeah, the typical household of some "only having a couple of laptops, all running Windows" family, actually has way much more instance of Linux running unnoticed on the ARM & MIPS embedded in various small gadget that nobody pays attention to, in addition to maybe a few distant cousins of BSD (any Apple iDevice, and Sony Playstations) :
- routers, smartphones, tablets, smart-appliance (smart TVs), etc. Forget about the "Year of Linux on the Desktop". The "Year of Linux on anything else beside the Desktop" has been achieved a long time ago.
A much simpler way to do it, that won't require you to hack standard system command-line tools, would be to use some copy-on-write or log-structured files system (e.g.: BTRFS, ZFS, etc. depending of your taste), and use snapshots to keep older versions of your file tree. If anything goes wrong you can still recover from a previous snapshot.
Some Linux distributions (like: opensuse) have tools (like snapper) that can automate this task for you (and opensuse uses snapper to similarily snapshot system upgrades).
In case of parallel construction, law enforcement need to construct. They need to have an actually credible explanation of how they got the evidence.
Get evidence through illegal mean A. Then invent a new mean B, that will give the same information (that you know in advance thanks to A), but that would appear to have worked without you knowing the information and could realistically happen even if A didn't happen. Do B, then get the same information, but this time legally. When asked by a judge, say you used method B and that's how you got the information (completely ignore the fact that you did A first, and that's how you got the idea to actually do B).
This isn't exactly true... Any artificial locomotion (any camera movement disconnected from your physical movement) in VR is still a big problem for many people (myself included) because it makes them sick.
Yes, there are a lot of *small details* that are getting refined. (and more so : most of them are subjective. I happen to be lucky enough not to suffer from kinetosis (sea-sickness). So I don't give such a big damn fuck about some of those details)
But still you don't need to invent a whole new genre to be able to exploit them. - back at the time when Forte's VFX1 headset was available for testing at a local computer shop, I could simply load up Descent and play test it. (Though some of my fried had spacial orientations problems, it was still *just quirks* that needed working out). - back when I managed to get an indecently good price on eBay for eMagine's Z800 Visor, I could still pop in Quake3 and enjoy it. (apparently someone did buy it during the launch promotions, and was disappointed and looking to get rid of it for an even lower price. Totally worth it, even if I needed to swap an upgraded ROM. Got even some hacks to get Bino to play 3D movies on it). - nowadays, there exist games that in addition of a normal computer screen + mouse setup can be played with VR headsets like Occulus or Vive. I've mentioned Doom above. Nobody needed to invent Doom or RE7 as a new genre just to be able to actually use these headsets. Yes, you would need to iron a few quirks out, just to be able to increase the quality of the users' experience. But a genre suitable for first person perspective had existed for quite a lot of time before that could be directly tapped into.
First person perspective already makes sense for tons of games which already exists (FPS, but also 3D puzzle games, etc.)
Well of course, if you take some poor schmuck who suffers from seasickness, they'll be rather happy about the new development in VR-headset. And if the poor schmuck is REALLY affected A LOT by it, they'll still need more quirks to get ironed out. But nobody needs to wait for new type of games to get invented, the currently available genre are already good enough. (In fact, back in the days of Z800 3d visor and such, there was a small cottage industry of special custom drivers designed to enable stereo 3D in games that weren't designed for it. Just a proof that you don't need to invent news games for VR).
Whereas, as I've stated above, motion games are an entirely new type of user input that needs new type of games to be adapted to it. That goes beyond the simple quirks to iron out (needing a way to avoid nintendo's equivalent of a tennis elbow). You need to invent new games where the whole concept of "dancing in front of a webcam" or "flailing your arms around" makes sense. There are some attempts : party games, sport games, DDR-like games (which seem to be the only tiny niche genre that already used motions before it was "a thing").
So I stand by my above comment: VR headsets are in a much easier places (quirks that you mention) than motions games (actually find a interesting way to integrate the technology into some AAA games, in a way that doesn't feel tackled on).
So VR headsets are definitely not the same kind of "passing fad" that motions controllers are currently, as the top poster implied.
It's more a question of market size and profitability: - Can you bring it's owning costs... -...and work around the quirks (motion sickness) -...so that the technology gets accessible to a big enough mass of people -...to be actually worth the resource spent on developing AAA games for it ?
And if you look into it: 1. costs - modern headsets try to offset the cost of device by using of-the-shelf parts (the price of high resolution small OLED screens has dramatically dropped, thanks to smartphone and tablets. Same for high rest cameras and accelerome
The point of high sampling frequencies is that you can then filter out any unwanted distortions in an easy way.
Yes *sampling* at 192kHz makes sense for the record studio : it's much easier to make a very good filter that filters everything above the hearing range while leaving everything nicely intact underneath when done as a software plugin, instead of an actual physical filter.
What I'm saying is that *playing back* at 192kHz as some sound cards and audio formats are able doesn't make much sense for a human to enjoy music. (Unless you're a *dog* enjoying music. Or a human doing some *weird scientific experiment* where you actually need a 192kHz signal - e.g.: feeding the output to some RF modulator to create some radio signal).
Or in other words : - a 24bits 192kHz *ADC* - like you mention - is *useful* - for the exact technical reason you give (filtering out the ultra-sonics at a later stage in software) - a 24bits 192kHz *DAC* - like the above poster mentioned - has no purpose for the standard "human listens to music" situation (and can even be detrimental to the quality of the output) - a garden variety 16bits 48kHz DAC is already well enough for most everyday situations (playing an already processed song)
...which entirely doable for an analog medium, and would be probably the way they'll do it if they decide to add DRM to analog mediums.
e.g.: - Audio tapes and vinyls come with a printed code on a label. - (let's say a QR-Code on the side of the tape, and a barcode around the edge of the round label in the middle of the vinyl) - that code is read by the player - (a small cheap webcam aligned inside the tape-deck exactly where the QR-Code would line once the tape is loaded ; a laser picking up the barcode as the vinyl spins around) - step one : analog player will refuse to play an analog medium whose code it can't check the signature. - (similar to how lock-out chips where used in game cartridges. Medium can still play as-is in old school players, but your lobbyist can vote a law to tax them) - step two : in addition to that, the code contain an encrypted parameter that controls some form of analog encryption (like the Nagravision that was used in the past with analog TV channels. It basically just flips/shifts the frequencies of the spectrum around)
(And it's the perfect parallel to the way DRM disturbs user experience : - analog audio encryption can be annoying to users (just like DRM region lock-out, or HDCP buggy drops can be annoying) because the wrapped frequencies might be less optimally conveyed by the analog medium (specially on Vinyls, a little bit less on tapes where the audio is modulated anyway) though you can still market your scheme as shifting the frequencies around to better positions. - analog audio encryption is just as useless as DVD's CSS because it can easily be broken and decrypted. And in fact *was* back in the nagravision days).
Which we aren't physiologically equipped(*) to be able to enjoy any way.
(*): Except to hear the hearing-range ( < ~10-15kHz ) distortions that are caused by throwing ultra-sonics (192kHz sampling means up to 96kHz frequencies) at a setup that was never meant for it.
the ONLY valid reason is that the mix is intentionally different, which makes zero sense.
It does make a sense commercially. If your mix looks like a set of solid bars constantly locked to the top on the spectrum of your radio/CD-player, it "sounds" subjectively louder and attracts more the attention. And when you're in the market of selling CDs, you *DO* want to attract the attention of customer to these CDs that you are selling.
LP are marketed to audio snobs who'll be more attracted to the "less loud"/more natural mix. You don't try to attract them with loudness, you attract them with the fact that it IS an LP and with the less shitty mix.
So yeah, the mix found in commercial CDs and files is much shittier to what you can find on LP (even if the technological limitations are actually the opposite), but that's because that's the mix with which you're the most likely to attract suckers to whom to sell those shitty mix.
You're grey hat hacker. You burst into the mail of someone who happens to be a criminal and who plans to rob a bank. You would like to have the police know it (so they can prevent the robbing), but you cannot admit that you got the information by hacking into mail servers (that would be illegal, give you trouble, and not necessarily be receivable as proof). So instead you manage to invent a different reason for the police to "happen to be there by chance" (ask a friend working at the police if they couldn't make a pause with their coworker to come into the bank to retreive money. Or make a false alarm go off somewhere that will cause the police to arrive, etc.)
The basic fact is exactly the same (the criminal XyZ want to rob a bank). You keep the truth, you just try to hide how you got it and try to invent another way by which this truth can be discovered. It looks as if method B uncovered the truth (by chance, etc.) whereas in fact it was method A, and method B was setup purposefully knowing that it will also be able to uncover the truth that you knew already in advance.
But the basic information is truth.
In short : Parallel construction is "hide the actual source of the info, come up with a believable way for the info to be discovered elsewhere 'by chance', but the info is legit".
Alternative fact :
You're a politician, A. You don't like politician B. So you start spreading rumours that politician B is actually a former criminal who got caught while trying to rob a bank.
Newspaper might be doubting your information. So they start digging into record and discover... that politician B never caught for a bank robbery or anything else, and in fact has never been discovered to be a criminal at all. When they come back to you trying to get you admit that you lied, you answer that your informations are simply "alternative facts". That you were "having the impression that politician B could have been a criminal" and that impressions aren't objective but personal, you're entitled to have your own impression and the journalist can't come criticising you for your impressions because these are hard to judge objectively and it's hard to give a definitive answer that politician B has never been and will never be a criminal.
The basic information (politician B == bank robber) is complete bullshit, but you come up with a weird reasoning of why the things you said can be considered as facts, and why those facts differ from that fact that other people pretend can be objectively judged.
In short : It's insane troll logic trying to justify why you want your lies to be believed as facts.
We've seen supposedly revolutionary new technologies come along before and then falter when it turns out that all they are good for is playing casual or party-games. The Wii's motion control sold a hell of a lot of consoles on the basis of Wii Sports. However, before too long, it dawned on people that Wii Sports was pretty much the limit of the device's capabilities. Similarly the Kinect had a lot of early success on the basis of some party games, but every attempt to integrate it into a proper game was either irrelevant or disastrous (Steel Battalion says hi). It's becoming increasingly clear that if any of these new technologies are going to "stick", then they need to be something you could realistically use to play a major AAA title; a Dragon Age or a Call of Duty (not that I'm a big fan of either of those). VR had looked like it was headed in the same direction as the Wii/Kinect; an initial burst of hype, then growing disillusionment.
I don't necessarily agree with your last statement.
Wii and Kinect (and countless other "revolutions in game play(tm)") share something that they don't have in common with VR. They completely change the way game are played, and that's a problem for an industry that has perfected other different way to play. That's not the case with VR.
When you boil down to the core concepts, Wii and Kinect are about completely throwing the input interface out of the window and trying something completely different : you do NOT use a classical controller, i.e.: you do not push button on a pad/joystick or use a keyboard/mouse pair. Instead you use body motions (wave you hands around with wiimotes in them, jump around in front of Kinect's camera, etc.) The industry has spend decades finding ways to get them most out of a joystick/gamepad or out of a keybaord/mouse pair. With have vast libraries of games optimized for them. We have complete *genre* of game which seem to be married to some type of input interface (2D platformers are strongest on gamepads - optionally keyboard as a fall back. Whereas mouse/keyboard is the king for strategy games and FPS - with maybe analog sticks as a pale substiture for the second) There's just not enough know-how and enough experience in how to successfully take advantage of new types of "motion"-controllers. And you though that bringing FPS from the keyboard/mouse to the joypad was a difficult task ? That's why, except for a few very special games that were thaugh from the gound up, every other game feels like having the motion element show-horned into an unrelated and incompatible game.
When you boil down to the core concept, VR is just... a game seen from the perspective of the hero. That's it. Like every single other FPS game since MidiMaze / Wolfenstein / Doom / Quake, etc. There's no fundamentally different idea that hasn't existed for the past few decade. That's why you've already seen tons of attempts of adding VR to games with first-person perspective. The difference is only in details. There are VAST amounts of details making VR much better as an output interface for first-person perspective, and also making the current crop of VR devices better than past attempts - but these are only details. But again you don't need to revolutionize the gaming industry to at least make attempts at using the technology.
To make things short: - in 2017 you can simply decide to play Doom with a VR helmet instead of a computer screen. Beside the quality of the experience (immersion, etc.) no big deal, game design isn't profoundly affected by the choice. - in 2017 you cannot simply decide to play Doom by dancing in front of a Kinect instead of using a controller. You'd need a totally different type of game.
The limiting factor of Wii/Kinect/etc. are: - can we suddenly now invent completely new way to play game to leverage these weird interfaces ? (hard problem)
The limiting factor of VR: - will the quality be good enough so that people won't automa
If that's true then why is this even a Pi thing? If the heavy lifting is in the cloud then my phone or my watch can do it.
The RaspPi are cheap.
If you want to do tons of things (like use AI for voice commands in your home, image recognition on your security cams, etc...) you either : - buy an expensive phone or smartwatch for each.
or: - buy a cheap RasbPi for each.
As the number of project increase the order of magnitude in price between a Pi and a smartphone start to really get noticeable.
Or is the problem that all developers are working on Pi hardware now? Where did all the devs go??
So yeah, most of the "hack at home" style of developers currently experiment with RaspPis.
I guess it boils down for A.I. students to publish which type and their respective sizes of problems can be handled by the Raspberry Pi?
Most of the time : Nope.
It boils down to the fact that most complex A.I. problems are nowadays tackled using neural nets. Which require usually orders of magnitude more power that what's available on a mere embed CPUs. You usually want a couple of servers in your basement each equipped with graphics cards running the neural nets over some GPGPU solution (OpenCL, CUDA, etc.)
And that's without taking into account the massive datasets needed to train said neural nets. e.g.: for Speach-to-Text For Google, with thousands of users all streaming their voice commands to the mothership, they have a huge corpus to refine their models. For us random geeks, we basically are limited to a few small open corpuses like voxforge.
So there are very few AI thinks that you can directly accomplish on the RaspPi itself.
Slashdot Mirror
The fundamentals behind sandboxing and hardware virtualisation date back from the time of big iron mainframes.
The only novel thing is that this is applied to a personal computer/workstation (to a newer type of hardware that didn't exist back then, but is basically just smaller).
(and keep the cryptographic keys, just in case backups fail)
I pay $5,000 for a shoot and 24 hours after I put it on the site some asshole puts it on the tubes
Definite proof that what they need, is some sort of patronage/crowd-sourcing.
Get 200 guys pay an average of 25$, the highest bidders get to make request (so they can fullfill whatevet they want they could not easily find elsewhere) - biggest bidders get to take a small part in production (either technical, like holding light. Or a small cameo in a scene. Whatever floats their boat)
Then it doesn't matter if the film is on porn hub the next day, he already got his 5000$ back before he even started.
Kind of how the request for weird fetishes on the image board work.
(Or some niche-fetish porn site offer paid requests)
If he cannot get 5'000$ in *advance* then there's slim chance that he'll ever be able to them after-ward.
If nobody is interested enough to pay to have him film it, he should instead try to think of something original enough to attract attention (and money).
Or switch job.
ng * g ^-1 :
nanograms [of the polluant, found] per gram [of the medium].
10 ng * g ^-1 :
if you take 1 gram of the soil of the bottom of the pacific ocean, you'll find that it contains 10 ng of the plastic polluant.
Why vote for the "lesser of 2 evils", when you could vote for Cthulhu ?
In an over simplified-way that might make physicist angry... ("I am a doctor, Jim ! Not a quantum physicist !")
Design a machine that fires approximately 1 photon per second.
Measure and confirm rate of photo firing.
Then put double slit in front.
Expose picture.
As during any given second, there's (an avarage) maximum of only photon,
then (in average situation) this single photon should not have any other to interact with.
Classic physics should predict only a (predominant) picture of the slit with (nearly no) pattern at all (except for the few outlier situation where 2 photons ended up in flight due to imprecision, but then their pattern should be much fainter).
Quantum physics should predict that even if (most of the time) only one photo is in flight, you still see a diffraction pattern predominantly. (and not only faintly for the couple of mis-firing)
In short : do the measurement AND the experiment at 2 different times, otherwise the measurement will destroy the photo.
Now again, this is NOT my area of expertise, so I probably made have of the physicists on /. cringe (and the other half laugh uncontrollably).
There is also a difference in the way you interpret the data.
Redlight and speedcams work very precisely because each vehicle has a license plate, matching an entry into a vehicle database.
Either you have a matching number, that should match an exact vehicle (or a counterfeit license plate), or you don't.
Identity is binary.
Also the devices are usually calibrated to guarantee precise results regarding measurement.
in other words, you can at least reliable trust the information (platenumber ; speed) coming out of photo by a speed-cam
(then, whether the license plate matches the actual car or was tampered with, and whether the car's registered owner was driving or some 3rd party, is an entirely different question)
Image analysis only give you a likelihood of resemblance.
Footage of person may or might not look like the suspect.
It has the same kind of imprecision as any visual witness. Some are better than other at identifying suspect, but nobody is perfect.
The only certitude you have is that the person doing crime on the video tape looks more or less like a given suspect.
The suspect *might be* the perpetrator.
But the suspect *might also be* a random look-alike that looked close enough to the person on the tape to fool the algorithm by chance.
That's valid both for visual clues (and that even if technologies becomes better at distinguishing people's image than regular witnesses) and for DNA clues (given a big enough pool of DNA samples, any methodology - short of full genome sequencing and that doesn't take into account twins - will eventually start to show random matches)
So information from cam can only be used as clues ("a person with a description matching the suspect has been recorded on tape to perpetrate a crime") and compared with other information (alibi: the suspect was at a party at that time and everyone else there saw him. The tape recorded a look-alike).
TL;DR: Technology is never going to replace good old detective work, only give extra information and tools.
That's unless government makes it mandatory to shave everyone's head and tatoo a QR-Code with the SSN.
Some companies can allow that. Many more cannot. Good luck assembling products or working in a restaurant while "working" at home. Hard for a nurse to take care of patients or for a construction worker to build a road from home.
Well, when I was speaking about work from home, i was more thinking about the typical /. crowd, working IT jobs or science jobs.
Of course for other people, that might instead be staying at home on a sick leave.
For many people (esp hourly people) missing work is not really a viable option even when sick.
Few employers provide enough paid sick leave to mitigate this problem. At my place of work people have to be considering a hospital visit before they will stay away.
Two very important things:
1- There are countries where sick leave days are the norm.
It is mandatory for both employee and companies to be covered by health insurance that will pay for sick days.
usually up to 2 days are "no question asked", beyond that a visit at the doctor is mandatory.
But yeah, I now : that is for us "evil communist european countries". Not for your land of the free, make america great again.
2- now, from a purely business logic: which situation does make more sense economically :
A. company doesn't pay any sick days. it's employee try to stay working no matter what. so in case of Influenza virus outbreak they stay at work. they don't treat their sickness, so eventually they are completely useless for a whole week (thanks to flu symptoms), but you dtill have to pay them, just because theu showed up. and by being at work while sick, they contaminate the rest of your workforce l. so everybody ends up being un productive for at least 1 week (except for the few people that could afford to get vaccinated againdt influenza A virus) (and that's hoping that this years strain was correctly predicted).
That is going to cost a lot of productivity to the company.
B. company ponies up and is able to pay at least a pay a few sick day leaves. A sick employee can afford to stay at home. They'll still be away for one week, but at least they won't transmit the disease much and there are only a few employee being absent.
Overall, it costs a bit to the company, but productivity doesn't dip that much down during each year's influenza outbreak.
C. health insurrance mandatory for every one. It costs a bit to the employer and employee to have a health plan.
But at least the health plan can cover sick days. And drugs expenses and medical visit.
In case of influenza, no only can the employee afford to stay home (and thus avoid to contaminate everyone else), but they can even afford to go to the doctor early enough (not only waiting until it has to be hospital) and thus get proper anti-viral therapy. Meaning that the employee is only away for 1 or 2 days while only having a very mild case if flu, instead of being debilitated by the symptoms and having to stay a whole week in the bed.
So it costs a bit to the componay and the employee, but productivity has barely any noticeable dip, even during outbreaks.
Also, lots of health plan also reimburse preventive medecine, and thus much more employee can afford to get vaccinated. Therefor even less employee are actually concerned by the seasonnal flu outbreak.
Overall, this is also the best scenario for the quality of life of the employee, as they get less sick and are in general better health (also meaning that they don't die too early, and that you don't risk losing trained workforce)
Ultra liberal USA went for A, because "it's not my fault if you're irresponsible and can't afford your health !" type liberals.
The rest of the developped world went for C - but we know, we are freedom-hating socialist scum.
No number for Leaf.
Current Renault Zoé (based on the same platform, but a Renault Clio body bolted on it) :
41kWh :
- costs ~9'000$ approx when you decide to buy the battery with the car (as opposed to rent one)
- costs ~15'000$ if you decide to buy a new one later on (because you initially opted renting one instead).
According to the summary : that battery would cost ~4100$ approx to built.
(Given that the company need to recoup its R&D costs, and the various subcontractors making the batteries (LG Chem, Imecar, Kreisel Electric, etc.) are all building their own equivalent of a megafactory in Europe, the price might seem seems legit).
Does he expect me to just test every morning when I get up from bed?
not every morning, and not everyone.
but if the technology can become cheap enough, some at risk situations could be monitored (non vaccined people, who are in contact with lots of others, when there's an active outbreak around).
and there's an immense benefits in detecting early:
- have the currently not yet symptomatic sick stay at home to avoid spreading it further (with most viruses, including influenza, you could be transmitting before you start feeling symptoms).
- treat it with specific anti-virals (for influenza: the sooner you start, the better the chances of it being effective. starting more than 48h after the onset of symptoms won't make much a difference - you'll still be dick. being able to begin therapy before even the symptoms starts will surely reduce the impact of the flu)
thus, if such a tchnology works, you could replace "laying in bed for 1 week" by "working from home for a couple of days" (a lot better quality of time)
and from an employer's perspective, it means that 1 sick guy wont contaminate your whole company (a lot less lost days)
(of course, for that you need a technology that is precise enough. i somewhat feel that if the analysis is limited to few gazes mentioned, it might tire of too many false alerts).
We've known what causes people to get sick (germs)
(...which is a gross over simplification. disease are caused by entirely different types of bugd: viruses, bacteria, funghi, protozoa, microscopic animals like worms, etc.)
for over a hundred years now. But it's amazing how many people still put their faith in things like health food diets.
Well, eating properly cooked food and drinking clean water can be proven to reduce some diseases
With proper detection and quarantine the flu could be eliminated from the human population almost entirely.
fat chance.
some diseases which are 100% human-targeting have indeed been eliminated (and you need to add vaccine on the list of things used against them).
but influenza affects tons of other animals (e.g: pigs) of which some are wild (e.g.: birds). And its going to be nearly impossible to control the spread of viral infections among (wild) migrating birds.
you can give specific anti virals to cases of Influenza detected early enough, you can quarantine to diminish transmission when sick (i.e. staying at home when sick.) but you can still have transmitted before starting to feel the symptoms), you can vaccinate to slow down its spread (hoping that the correct sub-type was predicted).
but you can't easily prevent animals carrying the influenza to pass it to human.
(well, in practice it's a tiny bit more complicated than just "passing". the different strains need to hybridize. but you got the general idea).
Personal theory: Most people, consciously or subconsciously, want to be LESS connected to their phones, not more connected.
Also another theory :
it boils down to specs.
And actually to too many of them.
To the point that "trying to get 4G" functional, became detrimental to the core functionality of a watch (tells you the time, without you needing to constantly charge it).
Some people might be interested into "a regular watch, that can also give you alerts". But not as many are interested into "we managed to cram a whole smartphone around your wrist !"
Pebble, the kickstarter campaign that got a bat shit insane success raising tremendously more money than initially hoped for, to the point that they got noticed by other manufacturer (like Apple, LG, Samsung, etc.) who realised that smartwatch are "a thing" (again) and thus kickstarted the (current) era of smartwatched :
They went for minimalistic specs. Putting as little processing power in the watch itself, counting on the linked smartphone to do any actually heavy computing, counting on e-Ink to even further lower the power consumption.
Their point: making a watch, that basically behaves like a watch (you don't need to charge your dumb wrist watch every single night), but has some interesting status display.
Basically, they just went for the next logical small-step evolution after the various Citizen's bluetooth-connected watches.
(And their result: tremendous success, given the kickstarter result, at time when smartwatch weren't yet considered "a thing" (again))
The other manufacturer saw the success, realised that smartwatch were popular (again),
and decided to give another try at it.
And attacked it through the same angle they have taken the habit with their smartphone :
Pump up the specs to try to beat the competition with a product that looks better on the paper.
More CPU power, better animated OLED colourful and luminous displays, extra health sensors (well those could actually be useful) touch screen instead of button, more connectivity, including Wifi or even Cellular (WTF ?!? You already have such a device in your pocket !), able to answer SMS and Tweets directly from the smartwatch itself (?!), etc.
All this to the expense of useful things like : actual battery life.
On the paper, these watches managed to look cool and attractive (so they got relative success) but in practice they are a little bit underwhelming (so after buying them, people didn't go batshit crazy).
So you end up with a small early player that got crushed by everyone else (Pebble).
And a bunch of smartphone makers that try to replicate a whole smartphone, sized down to fit around your wrist.
A few of them managed to get the lead in the market (Apple, probably because someone in their R&D lab was already musing with the idea back at the time of iPod Nano 6gen) but ended up with product that, although they tick every box on the feature check list, don't compare positively with regular watch (show the time, without needing a battery replacement for at least a couple of months).
And, whereas the Apple iPhone 1 managed to convince people that it's suddenly okay to have a cell phone that can go through the day without needing a charge, they didn't manage to pull the same trick with watches that need nightly recharge instead of holding months between battery replacement.
Thus, not everyone wanted to have an Apple Watch.
Because every other manufacturer (beside Pebble) jumped to big specs, but that ended-up being detrimental to the key feature that attracts most people to watch (able to always show time, needing a fresh battery only once every few months).
Another myth that these studies have debunked is that HDDs do better if kept cool. Actually, failure rates are lower for disks kept at the higher end of the rated temperatures. This is one reason that Google runs "hot" datacenters today, with ambient temps over 100F.
Funnily though, I've had very good success with IBM Deskstars (back during the infamous era of "Deathstars" click death) simply by running them cooler.
Though again, I did only have a few. So I only have anecdotal evidence.
Maybe I was just lucky to have the few Deskstars that didn't went "Dearthstar".
But maybe thermal management was indeed exceptionnally a problem on these old drives.
If taken far enough, Romanian carriers will have to raise their prices to account for their share of Danish infrastructure costs, but that means all Romanians would then be paying those costs (while still on lower Romanian salaries) while Danish tax and government budget is being undermined.
The current situation, the prices within country aren't that much high. (a few eurocents difference in the price per minute).
Maybe a users calling over a Romanian carriers within Romania will end up paying a few euros less than an users over a Danish carrier within Denmark.
I.e.: there is some variation between countries, but not as much as you would think, and specially not as much when compared to :
Roaming costs: They are currently outrageously high.
When abroad (e.g.: the Danish users goes to Romania) the costs of roaming can be 10x the normal price (you could be paying *whole euros* per minute - even when, per your reasoning the cost of Romania's infrastructure is certainly not the 9x price difference).
The current situation isn't about asking a couple of cent more to make up for the difference in infrastructure that you mention by donating to the roaming country.
The current situation is about milking the travelling users as if they were gold mines.
(It probably made sense for the carrier back in a time of less mobility, when probably most of the users roaming abroad would be business chief-whatever-officers type which don't pay much attention to the phone bill that their work is paying)
(Nowadays it just painful for all people travelling abroad because everybody has a wireless phone, because smartphone are nearly necessary for lots of everyday tasks, specially when abroad and you don't want to lug your laptop around, and modern society comes with even more geographical mobiliy)
This happens even within the same carrier. E.g.: buy a contract from O2 Germany. And whenever you travel to Czech republic, get charged outrageously, even when connecting to O2 CZ (the same company).
and it's easy enough to just get a local SIM card if you are travelling alot or for extended time,
Cue in mental picture of an avarage european user carrying 10 different SIM cards, and using a phone that has 4 SIM slots, with 2 of them always simultaneously on air.
If there were mass popular demand for it, carriers would already offer at least limited versions of it (potentially most popular in small countries or regions where travel to nearby countries is routine).
And there are such offer : there are "virtual carrier" (that don't own their own network of antenna, but piggy back on other carriers which they have an arrangement with) who have signed agreement with most big carriers in all European countries (it's not as impossible as it sound. There aren't really that many different carriers in Europes. Most of the time it's the same few international companies going each by different name/brands in each country. - see the above O2 example - Sign just a few of those big internationals, and you have deals in nearly every country).
By doing this, this virtual carrier are never actually roaming, you've always got at least one local operator with whom your virtual carrier has a deal and you always pay the local price.
as a random example : xxSIM
The proposed law is about always forcing this situation on all carrier, no excuse accepted.
Today, the vast majority of Linux kernels running are running on ARM processors.
And on MIPS.
a.k.a. "That other CPU that you'll also find in router and other embed devices, beside ARMs".
But yeah, the typical household of some "only having a couple of laptops, all running Windows" family,
actually has way much more instance of Linux running unnoticed on the ARM & MIPS embedded in various small gadget that nobody pays attention to,
in addition to maybe a few distant cousins of BSD (any Apple iDevice, and Sony Playstations) :
- routers, smartphones, tablets, smart-appliance (smart TVs), etc.
Forget about the "Year of Linux on the Desktop".
The "Year of Linux on anything else beside the Desktop" has been achieved a long time ago.
A much simpler way to do it, that won't require you to hack standard system command-line tools,
would be to use some copy-on-write or log-structured files system (e.g.: BTRFS, ZFS, etc. depending of your taste),
and use snapshots to keep older versions of your file tree.
If anything goes wrong you can still recover from a previous snapshot.
Some Linux distributions (like: opensuse) have tools (like snapper) that can automate this task for you (and opensuse uses snapper to similarily snapshot system upgrades).
In case of parallel construction, law enforcement need to construct. They need to have an actually credible explanation of how they got the evidence.
Get evidence through illegal mean A.
Then invent a new mean B, that will give the same information (that you know in advance thanks to A), but that would appear to have worked without you knowing the information and could realistically happen even if A didn't happen.
Do B, then get the same information, but this time legally.
When asked by a judge, say you used method B and that's how you got the information (completely ignore the fact that you did A first, and that's how you got the idea to actually do B).
This isn't exactly true... Any artificial locomotion (any camera movement disconnected from your physical movement) in VR is still a big problem for many people (myself included) because it makes them sick.
Yes, there are a lot of *small details* that are getting refined. (and more so : most of them are subjective. I happen to be lucky enough not to suffer from kinetosis (sea-sickness). So I don't give such a big damn fuck about some of those details)
But still you don't need to invent a whole new genre to be able to exploit them.
- back at the time when Forte's VFX1 headset was available for testing at a local computer shop, I could simply load up Descent and play test it.
(Though some of my fried had spacial orientations problems, it was still *just quirks* that needed working out).
- back when I managed to get an indecently good price on eBay for eMagine's Z800 Visor, I could still pop in Quake3 and enjoy it.
(apparently someone did buy it during the launch promotions, and was disappointed and looking to get rid of it for an even lower price. Totally worth it, even if I needed to swap an upgraded ROM. Got even some hacks to get Bino to play 3D movies on it).
- nowadays, there exist games that in addition of a normal computer screen + mouse setup can be played with VR headsets like Occulus or Vive. I've mentioned Doom above.
Nobody needed to invent Doom or RE7 as a new genre just to be able to actually use these headsets. Yes, you would need to iron a few quirks out, just to be able to increase the quality of the users' experience.
But a genre suitable for first person perspective had existed for quite a lot of time before that could be directly tapped into.
First person perspective already makes sense for tons of games which already exists (FPS, but also 3D puzzle games, etc.)
Well of course, if you take some poor schmuck who suffers from seasickness, they'll be rather happy about the new development in VR-headset.
And if the poor schmuck is REALLY affected A LOT by it, they'll still need more quirks to get ironed out.
But nobody needs to wait for new type of games to get invented, the currently available genre are already good enough.
(In fact, back in the days of Z800 3d visor and such, there was a small cottage industry of special custom drivers designed to enable stereo 3D in games that weren't designed for it. Just a proof that you don't need to invent news games for VR).
Whereas, as I've stated above, motion games are an entirely new type of user input that needs new type of games to be adapted to it.
That goes beyond the simple quirks to iron out (needing a way to avoid nintendo's equivalent of a tennis elbow).
You need to invent new games where the whole concept of "dancing in front of a webcam" or "flailing your arms around" makes sense.
There are some attempts : party games, sport games, DDR-like games (which seem to be the only tiny niche genre that already used motions before it was "a thing").
So I stand by my above comment:
VR headsets are in a much easier places (quirks that you mention)
than motions games (actually find a interesting way to integrate the technology into some AAA games, in a way that doesn't feel tackled on).
So VR headsets are definitely not the same kind of "passing fad" that motions controllers are currently, as the top poster implied.
It's more a question of market size and profitability : ...and work around the quirks (motion sickness) ...so that the technology gets accessible to a big enough mass of people ...to be actually worth the resource spent on developing AAA games for it ?
- Can you bring it's owning costs...
-
-
-
And if you look into it :
1. costs
- modern headsets try to offset the cost of device by using of-the-shelf parts (the price of high resolution small OLED screens has dramatically dropped, thanks to smartphone and tablets. Same for high rest cameras and accelerome
The point of high sampling frequencies is that you can then filter out any unwanted distortions in an easy way.
Yes *sampling* at 192kHz makes sense for the record studio :
it's much easier to make a very good filter that filters everything above the hearing range while leaving everything nicely intact underneath when done as a software plugin, instead of an actual physical filter.
What I'm saying is that *playing back* at 192kHz as some sound cards and audio formats are able doesn't make much sense for a human to enjoy music.
(Unless you're a *dog* enjoying music. Or a human doing some *weird scientific experiment* where you actually need a 192kHz signal - e.g.: feeding the output to some RF modulator to create some radio signal).
Or in other words :
- a 24bits 192kHz *ADC* - like you mention - is *useful* - for the exact technical reason you give (filtering out the ultra-sonics at a later stage in software)
- a 24bits 192kHz *DAC* - like the above poster mentioned - has no purpose for the standard "human listens to music" situation (and can even be detrimental to the quality of the output)
- a garden variety 16bits 48kHz DAC is already well enough for most everyday situations (playing an already processed song)
"Digital" rights management, idiot.
...which entirely doable for an analog medium, and would be probably the way they'll do it if they decide to add DRM to analog mediums.
e.g.:
- Audio tapes and vinyls come with a printed code on a label.
- (let's say a QR-Code on the side of the tape, and a barcode around the edge of the round label in the middle of the vinyl)
- that code is read by the player
- (a small cheap webcam aligned inside the tape-deck exactly where the QR-Code would line once the tape is loaded ; a laser picking up the barcode as the vinyl spins around)
- step one : analog player will refuse to play an analog medium whose code it can't check the signature.
- (similar to how lock-out chips where used in game cartridges. Medium can still play as-is in old school players, but your lobbyist can vote a law to tax them)
- step two : in addition to that, the code contain an encrypted parameter that controls some form of analog encryption (like the Nagravision that was used in the past with analog TV channels. It basically just flips/shifts the frequencies of the spectrum around)
(And it's the perfect parallel to the way DRM disturbs user experience :
- analog audio encryption can be annoying to users (just like DRM region lock-out, or HDCP buggy drops can be annoying) because the wrapped frequencies might be less optimally conveyed by the analog medium (specially on Vinyls, a little bit less on tapes where the audio is modulated anyway) though you can still market your scheme as shifting the frequencies around to better positions.
- analog audio encryption is just as useless as DVD's CSS because it can easily be broken and decrypted. And in fact *was* back in the nagravision days).
dacs can do 24/192k and even DSD.
Which we aren't physiologically equipped(*) to be able to enjoy any way.
(*): Except to hear the hearing-range ( < ~10-15kHz ) distortions that are caused by throwing ultra-sonics (192kHz sampling means up to 96kHz frequencies) at a setup that was never meant for it.
the ONLY valid reason is that the mix is intentionally different, which makes zero sense.
It does make a sense commercially.
If your mix looks like a set of solid bars constantly locked to the top on the spectrum of your radio/CD-player, it "sounds" subjectively louder and attracts more the attention.
And when you're in the market of selling CDs, you *DO* want to attract the attention of customer to these CDs that you are selling.
LP are marketed to audio snobs who'll be more attracted to the "less loud"/more natural mix. You don't try to attract them with loudness, you attract them with the fact that it IS an LP and with the less shitty mix.
So yeah, the mix found in commercial CDs and files is much shittier to what you can find on LP (even if the technological limitations are actually the opposite), but that's because that's the mix with which you're the most likely to attract suckers to whom to sell those shitty mix.
Parallel construction :
You're grey hat hacker.
You burst into the mail of someone who happens to be a criminal and who plans to rob a bank.
You would like to have the police know it (so they can prevent the robbing),
but you cannot admit that you got the information by hacking into mail servers (that would be illegal, give you trouble, and not necessarily be receivable as proof).
So instead you manage to invent a different reason for the police to "happen to be there by chance" (ask a friend working at the police if they couldn't make a pause with their coworker to come into the bank to retreive money. Or make a false alarm go off somewhere that will cause the police to arrive, etc.)
The basic fact is exactly the same (the criminal XyZ want to rob a bank).
You keep the truth, you just try to hide how you got it and try to invent another way by which this truth can be discovered.
It looks as if method B uncovered the truth (by chance, etc.) whereas in fact it was method A, and method B was setup purposefully knowing that it will also be able to uncover the truth that you knew already in advance.
But the basic information is truth.
In short : Parallel construction is "hide the actual source of the info, come up with a believable way for the info to be discovered elsewhere 'by chance', but the info is legit".
Alternative fact :
You're a politician, A. You don't like politician B.
So you start spreading rumours that politician B is actually a former criminal who got caught while trying to rob a bank.
Newspaper might be doubting your information. So they start digging into record and discover... that politician B never caught for a bank robbery or anything else, and in fact has never been discovered to be a criminal at all.
When they come back to you trying to get you admit that you lied,
you answer that your informations are simply "alternative facts". That you were "having the impression that politician B could have been a criminal" and that impressions aren't objective but personal, you're entitled to have your own impression and the journalist can't come criticising you for your impressions because these are hard to judge objectively and it's hard to give a definitive answer that politician B has never been and will never be a criminal.
The basic information (politician B == bank robber) is complete bullshit, but you come up with a weird reasoning of why the things you said can be considered as facts, and why those facts differ from that fact that other people pretend can be objectively judged.
In short : It's insane troll logic trying to justify why you want your lies to be believed as facts.
We've seen supposedly revolutionary new technologies come along before and then falter when it turns out that all they are good for is playing casual or party-games.
The Wii's motion control sold a hell of a lot of consoles on the basis of Wii Sports. However, before too long, it dawned on people that Wii Sports was pretty much the limit of the device's capabilities. Similarly the Kinect had a lot of early success on the basis of some party games, but every attempt to integrate it into a proper game was either irrelevant or disastrous (Steel Battalion says hi). It's becoming increasingly clear that if any of these new technologies are going to "stick", then they need to be something you could realistically use to play a major AAA title; a Dragon Age or a Call of Duty (not that I'm a big fan of either of those).
VR had looked like it was headed in the same direction as the Wii/Kinect; an initial burst of hype, then growing disillusionment.
I don't necessarily agree with your last statement.
Wii and Kinect (and countless other "revolutions in game play(tm)") share something that they don't have in common with VR.
They completely change the way game are played, and that's a problem for an industry that has perfected other different way to play.
That's not the case with VR.
When you boil down to the core concepts, Wii and Kinect are about completely throwing the input interface out of the window and trying something completely different : you do NOT use a classical controller, i.e.: you do not push button on a pad/joystick or use a keyboard/mouse pair.
Instead you use body motions (wave you hands around with wiimotes in them, jump around in front of Kinect's camera, etc.)
The industry has spend decades finding ways to get them most out of a joystick/gamepad or out of a keybaord/mouse pair. With have vast libraries of games optimized for them. We have complete *genre* of game which seem to be married to some type of input interface
(2D platformers are strongest on gamepads - optionally keyboard as a fall back. Whereas mouse/keyboard is the king for strategy games and FPS - with maybe analog sticks as a pale substiture for the second)
There's just not enough know-how and enough experience in how to successfully take advantage of new types of "motion"-controllers.
And you though that bringing FPS from the keyboard/mouse to the joypad was a difficult task ? That's why, except for a few very special games that were thaugh from the gound up, every other game feels like having the motion element show-horned into an unrelated and incompatible game.
When you boil down to the core concept, VR is just... a game seen from the perspective of the hero. That's it.
Like every single other FPS game since MidiMaze / Wolfenstein / Doom / Quake, etc.
There's no fundamentally different idea that hasn't existed for the past few decade.
That's why you've already seen tons of attempts of adding VR to games with first-person perspective.
The difference is only in details.
There are VAST amounts of details making VR much better as an output interface for first-person perspective, and also making the current crop of VR devices better than past attempts - but these are only details.
But again you don't need to revolutionize the gaming industry to at least make attempts at using the technology.
To make things short :
- in 2017 you can simply decide to play Doom with a VR helmet instead of a computer screen. Beside the quality of the experience (immersion, etc.) no big deal, game design isn't profoundly affected by the choice.
- in 2017 you cannot simply decide to play Doom by dancing in front of a Kinect instead of using a controller. You'd need a totally different type of game.
The limiting factor of Wii/Kinect/etc. are :
- can we suddenly now invent completely new way to play game to leverage these weird interfaces ?
(hard problem)
The limiting factor of VR :
- will the quality be good enough so that people won't automa
If that's true then why is this even a Pi thing? If the heavy lifting is in the cloud then my phone or my watch can do it.
The RaspPi are cheap.
If you want to do tons of things (like use AI for voice commands in your home, image recognition on your security cams, etc...)
you either :
- buy an expensive phone or smartwatch for each.
or:
- buy a cheap RasbPi for each.
As the number of project increase the order of magnitude in price between a Pi and a smartphone start to really get noticeable.
Or is the problem that all developers are working on Pi hardware now? Where did all the devs go??
So yeah, most of the "hack at home" style of developers currently experiment with RaspPis.
I guess it boils down for A.I. students to publish which type and their respective sizes of problems can be handled by the Raspberry Pi?
Most of the time : Nope.
It boils down to the fact that most complex A.I. problems are nowadays tackled using neural nets.
Which require usually orders of magnitude more power that what's available on a mere embed CPUs.
You usually want a couple of servers in your basement each equipped with graphics cards running the neural nets over some GPGPU solution (OpenCL, CUDA, etc.)
And that's without taking into account the massive datasets needed to train said neural nets.
e.g.: for Speach-to-Text
For Google, with thousands of users all streaming their voice commands to the mothership, they have a huge corpus to refine their models.
For us random geeks, we basically are limited to a few small open corpuses like voxforge.
So there are very few AI thinks that you can directly accomplish on the RaspPi itself.