In this media wasteland, you could do worse than a show with reasonably good writing that's not violent or cynical. The show is not a 20 minute cereal commercial, is not likely to cause epileptic seizures, and isn't so infantile that an adult can't sit through it. In other words, I approve of it, and I don't approve of... anything, really.
At least one TV network has realized this - and it's getting fairly good ratings (as good as network TV gets these days) because of it - CBS. Sure they have SOME reality shows (who doesn't? And you need a few for the few viewers you have that demand it). But they have a LOT of the more traditional scripted and written shows. And they consistently do a lot better - the scripted shows generally outperform the reality ones these days.
I have yet to come across a voltage regulator that doesn't run hot. Typically, it's one of the hottest components in an electrical circuit. And we're integrated this into a slab of silicon already well-known for getting so hot it can catch fire?
Can someone please tell me why this is a good idea, because all of my experience in electrical engineering says that when things heat up, they become more unstable and prone to failure, and the one thing you do not want going critical is your voltage regulator. If that goes, the whole computer catches fire.
Voltage regulators typically run hot because of the pass transistor. In a linear supply, the pass transistor is what dissipates the difference in voltage from input to output. In a switching supply, the pass transistor is the chopper (the one that converts the DC to AC for the transformer or magnetics). Most VRMs in computers are actually cooled just passively - if you look, the tab is soldered to the motherboard and a huge copper plane on the PCB (usually via-stitched to carry heat to the ground or power plane as appropriate).
HOWEVER, there is one thing the on-die VRM has over the on-motherboard one - a heatsink!
The main CPU core has to be cooled anyhow, so the VRM has a very easy access to the heatsink and can thus dissipate the heat much easier.
I would guess one of the reasons for doing it on board is because of the currents and voltages involved - power supply is becoming a big issue when you have multiple voltage rails (Vcore, Vgpu, Vio (x a dozen or so depending on the IO), etc). Each voltage rail requires at least one pin or multiple pins, and even though we have 1100+ pins, it still means the CPU is pin-limited (in ASICs, you can be pin-limited, where what you can stuff onto an IC is limited by the number of pins you can have, or silicon-limited, where what you can stuff in is limited by silicon area. SoCs and such are typically pin-limited, while memories are silicon limited. Similar concept to IO bound and CPU bound programs, really.)
If instead of dedicating easily 100 pins to power (if your core is 1.8V and you're drawing 90W, you're wanting 50A, and each pin may take.75-1A max meaning you're looking at 50-75 pins, plus power to other onboard stuff), you can dedicate 25 to higher voltage single rail, it's a lot less current per pin, so lower IIR losses, and a lot more pins freed up. If you need 90W, and you feed in say, 15V, you only need 6A, so instead of the 50-75 pins to carry 50A, you only need say, 10 pins. Plus, if you want other voltages, instead of dedicating more pins, the onboard VRM gives you more flexibility - say you want another 60W, that's 150W, at 15V, that's 10A, or 10-15 pins. Still a lot less and giving you 80-odd more pins for further I/O.
80 pins is a lot - you can add a number of additional PCIe channels and lanes with that, for example. Or additional functions.
And you've simplified the motherboard design a lot and gotten rid of a potential source of issues - poor power supply design by motherboard manufacturers (applies mostly to cheap budget boards).
Only if you can accept payments via Google Checkout/Wallet.
You see, the Play market is available in more countries that you can accept payment from - paid apps are simply not shown to those countries. (For a long time after Android first came out, the only supported country was the US, while the phone was available in a number of other countries, including Canada. End result is if you wanted a paid app, you had to pirate it because Google didn't make it available to you).
As a result, many apps went "freemium" because it's either not show up and be pirated, or go free and make up the money selling ads and in-app purchases (the latter being fairly new, which is why many apps use the ad model with fairly... "liberating" permissions required).
So if you're selling in a country that can accept Google payments, you must use it. If not, you're free to engage Paypal and others (which until Google allowed in-app payments most devs used Paypal, and even after IAPs until Google cracked down).
Fact remains, Play is available in more countries than you can accept money from.
Bingo. Why does a system tray notification require admin rights? Every other software installer I've ever downloaded tells you what it's going to install and only asks for admin rights when the installation process itself starts.
Better yet, why isn't it downloading on behalf of the installer and letting the INSTALLER ask for admin rights?
Half the time, it claims there's an update, and then it promptly fails to download it. After giving it admin. Why not attempt to download it ahead of time?
Yes, ask for admin if it would save needless popups (e.g., Windows Update - where installing multiple patches may require admin priviledges - so ask it up front then use that to run the patch installers as admin to avoid bugging the user).
Apple doesn't bother making a better product, they never really have. They have brilliant marketting though such that people who don't know better flock to the thing in droves because it's an iphone. They know what an iphone is, and equate it with smart phone. never mind that the competing phones do more, do it better, and do it at a lower price. The iphone is consistently a year or more behind the major Android players in features, the usability and interface really improved since the first iphone launched (and it badly needs to) and yet they manage to get people to line up every year to buy the new iphone, even when there's no real change from the last one.
Problem is, people aren't sheeple. They can spot marketing BS a mile away.
If a product is all hype and no substance, guess what? It dies. And in the age of instant news, if a product is crap, people know about it the moment it's released. Doesn't matter if it's an iPhone, a movie, a new store, etc. It doesn't matter if the revised edition is better or not.
Hell, see Vista. Despite marketing claims (Project Mojave, anyone?), people still thing it's a POS. Even though in the Windows 7 era, Vista benefits from developers doing crap right and things actually work. (Vista took the crap for Windows 7, you could say).
Hell, Linux still has a "hard to use" reputation, despite it being easier to install and update than Windows (though, some of it is deserved - some support communities can really do better than "RTFM" as a reply. Like "RTFM section X, paragraph y" linked).
Basically, if the iPhone was crap, then people would've known 6 years ago. Or on release of every new model ever since. But since Apple can pull in the profits, they obviously either managed to brainwash everyone (which is a very politically powerful tool wasted on getting people to buy a brand of computers?), or have a product that serves its purpose.
Sure, an iPhone won't meet any measurebator's specs, but do we really need to get back to the megapixel wars, GHz wars, GByte wars, etc? Or has it been definitely proven that more megapixels are better, more GHz are better, more GBs are better (OK, the last one is definitely better).
Hell, look at what Apple started with the "retina" display - more and more DPI until I'm sure few can even see the pixels up close (441 - the 5" 1080p screens). Or phones getting so big that people are making "small phones" to go with your big one. (NOTE: The SGS3, while the best selling Android phone, only has around 10% of the market. The rest of them are the hundreds of other phones Samsung makes with smaller screens, keyboards, crappy processors, etc, and some others by LG, HTC, Motorola, etc).
tl;dr - marketing can sell you a product, but it can't work a miracle. A turd is still a turd and no amount of marketing can spruce it up.
This guy's near daily media appearances has certainly inspired many canadians including myself. I have watched many children sing along with his ISS song (not as good as david bowie, but its the thought that counts) and it really inspires. Hopefully helping lots of kids to think about becoming scientists, researchers and yes astronauts. Space can seem so dull sometimes, he really brings it to life.
I may not care for much patriotically these days, but hes really doing canada a service being so media savvy. I am not sure if american astronauts do so much singing, and perhaps its covered extensively by their local media and I just never hear about it. But he really could be one of a kind.
Before he launched in December, Chris mentioned he was going to do the first album recorded in space, I'm hoping this was just a taste of what's coming.
I have to be honest, I've been watching a LOT of Chris' videos that get posted by the CSA (Canadian Space Agency) (an agency facing budget cuts from the Harper Government(tm)). I don't think I've seen anyone from the ISS do so much media relations in their off time.
I know a few other commanders have done media work - Don Pettit did some as well. With the American Physical Society (any physics major should know them) he did a bunch of videos called "Science off the Sphere" (which I apparently finally got my T-Shirt from that).
Chris is definitely very media friendly and has hosted a LOT Of media events while aboard - he even keeps in touch with Discovery Canada's Daily Planet, the longest running science program around. Honestly, Chris Hadfield is awesome!
Safe journey home - your country eagerly awaits your arrival!
If you're in Canada, Daily Planet carried this car earlier this month. They mentioned they were doing test flights.
OF course, you get to see a rather interesting takeoff int he clip. Alas, I think it's Canada only - not sure if the US Discovery channel has it on any of their channels (it's a Canadian production).
The expression "Apple does not have a backdoor per se" basically cannot be proven unless you have a full source code. Moreover, nothing will stop a real hackers from desoldering a flash and attaching it to reader. And also: I've never seen a modern device which does not have some JTAG or similar debug port that can be useful to program the very bootloader that verifies the digital signatures of bootable code. Times when BIOS was pluggable are gone.
Except around the 3Gs era, Apple started hard-encrypting the flash to prevent that very attack.
And JTAG ports can be disabled by software - I've worked on devices where once a fuse was blown, the JTAG lines were disconnected internally and thus inaccessible.
And yes, Apple is the only one that can do it because they hold the keys. If you need to load out special software, only Apple has the private key to sign and run whatever tools they have. And they can probably read out the filesystem, figure out what the keys are and brute force what they need to brute force.
Of course, the article doesn't say Apple has any success at all - perhaps they can crack the 4 digit passcodes that bypass the 10 code self-lockout and erasure. But you can enable a more secure form using a complex passcode. And supposedly you can enable even more sophisticated encrypted and protections.
Or hell, we don't even know how may phones are in the queu or how long it takes Apple, It could take Apple 4 months to decrypt the iPhone. Or they may have a backlog because only one person is decrypting them and he can only do one a day or something.
"Deprived of the ability to use browser plugins, protected content distributors are not, in general, switching to unprotected media. Instead, they're switching away from the Web entirely. " So what is wrong with this, exactly? If you want to distribute DRMed content, you are fully free to use your own means. Let the web stay DRM-free, as it should be.
So you want the Android and iOS way of using the internet then? Where you interact with apps to do your banking, check maps, check what food is nearby, get ratings on movies/shows/food, etc?
You know, Steve Jobs was onto something with the original iPhone - where developers would be forced to do everything in HTML and Javascript. But developers lobbied for a native SDK, leading to app stores and the rest as we know it today. Now we have people complaining that why are people using apps now when before, they could've just used the web? People are making apps that are basically web site containers. Or publishing apps that are basically sets of static HTML content in a web view.
And not on the web.
Hell, people use the Amazon app to buy stuff online. You go into the store, scan the barcode with the amazon app, and click buy. You could do the same by browsing Amazon's web page on your browser, but it's less convenient and easy. And if the trend continues, Amazon may decide that their website links you to their app page to finish the purchase because the vast majority of people are buying stuff through the app. Right now it hasn't happened. But who knows?
In the end, an app based world is one where everything is segregated - you use one app for your sports scores, another for your banking, a third for shopping, etc. And there'll be this dusty app called a "web browser" that once was used for that stuff, but because everyone used apps instead, slowly lost prominence as apps could have better security, could display content better, could do advertising better, etc. etc. etc. I'm sure Netflix's main usage now is apps (for smartphones, set top boxes, etc), and a very tiny portion of it is pure web access. Silverlight is dead. It's only a matter of time before Netflix pulls web access to its content because everyone uses the Netflix app anyway.\
Hell, forums on the web are being taken over by apps - you may have heard of things like Tapatalk and Forum Runner.
Robotic manufacture is going to go crazy in the next decade and it's going to change everything.
Software is needed to make that shit work. If software patents can make (profitable) roadblocks to 3d implementation, then robotic manufacture is going to go someplace besides the USA.
This is so obvious that even the Supreme Court is going to see it. (Congress might need some 'lobbying' to understand it, though.)
Where do you draw the line, though?
If I make a robot that's completely controlled by software so that the only thing mechanical I have are motors and sensors, that means I can only patent whatever's in that heap of metal? If I come up with a clever way of doing something like being able to get precision out of imprecise hardware, it's not patentable? Whereas, if I stuck that software inside a hardware implementation (software runs on hardware - I can make fixed hardware do the same thing), I can now patent it?
That's the tricky part. It means the cheap hardware and doing everything in software solution will lose out to the expensive hardware because the latter can contain patents, but the former will have very little actually patentable, because it's all software.
Or consider something like a software defined radio. Very little hardware, and beyond the basics, everything's in software. But if I invent a clever way of using the radio waves, if I do it in hardware, I can patent the crap out of it, but if I use an SDR, I can't?
The main problem with software patents is software is nebulous. There is nothing like it before. Before the computer age, we had "stuff" you could patent, and "creative works" which you copyrighted. But now software bridges the two - is it a create work because it's written, or is it "stuff" because it's a fundamental part of whatever machinery it controls? Nevermind RTL software written and compiled to ASICs and FPGAs.
Software has basically ended up taking two normally incompatible things (you don't patent books, like you don't copyright machinery) and fused them together. Your clever machinery depends on software for part of its operation - together it's interesting, separately, it's a hunk of metal and writing and both are useless apart.
Nevermind software generating software - is the output of such copyrightable? What if software generated a book? What if the inputs to the software are under other copyright?
In fact, what might end up happening is a merging of the two - special protections for software only that combine parts of copyrights (the code itself, and any output it generates), and patents (when embedded with hardware to perform a novel task), thus sidestepping the whole issue entirely - software cannot be patented or copyrighted, but it falls under software based IP.
every other tech device - including extremely similar devices, tablets - have come down in price at least four-fold if not ten-fold over the same time period.
phones remain expensive to buy outright because the customers that the phone manufacturers are targetting are their largest customers, the telcos. if new phones were cheap to buy outright, people would be far less inclined to sign up for abusive two year contracts to get a hire-purchase phone (not "free" and not "subsidised" - the price is embedded in your contract)
Incorrect.
Flagship phones are expensive. But you can find unsubsidized Android phones at all price ranges. And no, the most popular phones are NOT the flagships. The SGS3, last year's flagship, sold a bit over 50M units (March 2013). At the time, Google was activating anywhere from 1 (beginning of 2012) to 1.5M (October?) Android phones per day. If we take the numbers, we see roughly 1 in 10 Android phones was an SGS3, the most popular Android phone model ever. That leaves 9 random Android phones out there, and Samsung has around 80% of the Android market, so 7 of the 10 are Samsung Android phones, and the other 2 are from LG, HTC, Motorola, etc. The other flagships don't even register.
Which makes the vast majority of Android phones the cheaper ones. Granted, they're "free with contract" but generally around $200 unsubsidized if you look at the carrier offerings. Of course, they're also useless pieces of crap - either running 2.1 or 2.2 or if you're lucky, 2.3 (though some may have 4.0). Piss-poor WVGA ish screens, sluggish processors (1GHz if you're lucky), low RAM, etc.
Basically, the price has come down, if you want a POS Android phone.
Tablets the same - excluding the Nexus 7, the $200 and under tablets generally are crap. WVGA is common, 1024x600 if you're lucky. Then you have the Nexus 7, the best of the lot, but forced to compete with sub-par. The 10" tablets are already $400+ with the iPad being the most expensive and followed closely by the Samsungs. But they're generally around there.
No, prices have not gone down. All you're seeing is the "build to a cost" nature of it, like the crappy $500 laptops (the ones that are generally good start at $1000 where you can find discrete GPU, >1366x768, and quality). It's just people figured out how to cut corners and make cheap crap.
And no, despite "subsidies going away", what happens is people are opting for the Tab model - where they get the phone "free" and their tab gets paid down off their monthly bill. The only difference is the price is shown externally and you get a small discount on your bill at the end of the contract (for the phone purchase).
And generally, people will choose the cheapest phones still - your flagships will still cost $200 with $400 "tab", while the rest are "free on the tab". Samsung and others will, however, make tons of similar-but-cheaper-crappier phones so you have to specifically ask for an SGS4, not and SGS4 Lite, Micro, Grand, Mega, (Did you know you can still buy an SGS2? Not necessarily the same SGS2 as was the flagship, but they have phones called SGS2).
I came here to post exactly this. Conventional TV channels need to appreciate that technology has created a viable alternative and, like it or not, they have to compete with it. Most of our terrestrial channels know this and have launched catch-up and/or live TV apps on a decent range of platforms, that carry advertising to pay the bills. I'm quite pleased to see that even Sky, that bastion of awfulness, has come to terms with the fact that its business model may be ending and has launched a streaming (live and on demand) subscription package for its channels.
Yes, because streaming one stream per viewer is far more efficient bandwidth wise than broadcasting it to everyone. After all, bandwidth needs of the former scale linearly while it stays flat with the latter. The real problem happens when something big happens and everyone turns on their TV - I don't think handling the 300M-odd TVs is possible right now. (Most streaming networks have far fewer streams than that).
Of course, companies like Netflix do it by not bogging down the Internet - but by creating CDNs and putting them in ISP data centers - because the ISP to subscriber link, despite oversubscription, is still fatter than the ISP to internet peering link.
Anyhow, if you have seen TV, you'll note that each network of channels have been moving their flagship shows to the secondary channels as well - the other channels in the network get older seasons. This is so popular shows people watch together end up being on different channels, so they have to purchase both channels in the a la carte model. (On the bundle model, they're usually "free" because they're bundled together). So while you used to just ignore the dozen other Discovery-owned channels for the main one, soon you have to purchase at least half of them because the main channel has moved new seasons to other channels.
At least that's what I've seen - I've had to check a few channels because they seem to move channels a lot.
You are comparing nuclear power to experimenting and create nuclear weapons... Nuclear Power as it is today is very safe, reliable, and cheap if done correctly. People oppose nuclear power because they are scared because of their ignorance.
Actually, is it as cheap as people say it is? During operation, yes, it's very cheap. But there's a long tail of maintenance that goes on long after the money making days are over. Like cleanup which often takes decades, plus the storage and handling of the waste which takes a long time as well. Those are costs that have to be borne for years after the plant shuts down, and it seems for a lot of them, it's offloaded to taxpayers as the company running them moves onto other things to please shareholders who get the reap the profits and none of the final costs of shutdown.
90% seems awfully high. I was under the impression typical performance was around 40%-70%, depending on the chemistry and materials.
In general, the limiting factor is the quality of the hydrogen - the oxygen can generally come straight from the atmosphere. In this case, it's pure hydrogen stored as a liquid, so fuel cells tend to be fairly efficient.
Take Apple's fuel cells powering their data center, and they're running off natural gas, which is 20% carbon for the most part (by stoichiometry) which is impure and has to be dealt with. And nevermind that natural gas, while being mostly methane, also contains longer gaseous hydrocarbons that provide more "poison" to the cell that has to be dealt with.
Of course, you could refine methane to hydrogen and carbon, but hydrogen itself is rather difficult to transport since it doesn't like to be contained easily.
Not really. Android's the most popular mobile OS, yes, but that's just because it's like PCs - it's being put on devices that Apple won't touch.
Samsung has around 80% of the Android market. But the SGS3, the best selling Android smartphone, only sold 50-odd million phones. Out of the entire Android ecosystem, that's roughly 8-10% of the market - by far the market leader for the model, but a majority of phones sold. That means out of the 8 out of 10 phones Samsung sells, only 1 of them is an SGS3. The other 7? Crap phones you see advertised - SGS2 (yes, they still make them) various flavors, the other crap ones with crappy WVGA screens, or slow processors, or tiny RAM. Basically, super cheap phones the carriers are pushing out to sell everyone a data plan and lock them into whatever crap else there is. (If you look carefully - you'll probably pay MORE for a "dumbphone" than an Android phone). And that's the top seller - the other flagship phones? Not even the radar.
So by marketshare, Android phones routinely outsell iOS by 3 to 1 or so, or more.
However, an interesting stat keeps popping up - by data usage, iOS users tend to use their data plans - especially websurfing traffic where Safari tends to out-do Android by 2 to 1. So it seems almost exclusively all the Android traffic is driven by high end flagship phone users (makes sense - they're the ones who tend to be willing to pay and actively choose Android over iOS - the other Android users tend to pick it because they don't want to pay the $200 an iPhone costs (don't argue about the "cheap" models - if you go into any store, only the iPhone 5 is prominently displayed), and carriers are literally shoving phones out the.). Other than mobile OS choice, I'd say flagship Android users would be fairly similar demographically to iPhone users.
This has resulted in an interesting situation - advertisers who can't target just flagship Android phone users, are basically paying more for iOS ads. And Google, despite having a majority of phones running Android, still makes more money from iOS.
I think the program GP has in mind is Torque (that can read OBD 2 data and do many other things). iphone lacks the serial port profile for bluetooth so it's quite a sure bet it can't do that.
Odd.
I mean, a quick search through Appshopper brings up many OBD-II reader programs.
Chromebooks are like tablets. They are generally complementary products for desktops and laptops, not replacement products. They are a device you can leave on your coffee table and do a little web browsing and email from your couch. Similarly your guests can pick it up and use it from the couch. It can be a box to help keep others off of your main pc or laptop, less malware risk.
Problem is, Chromebooks are crappy laptops - they have a laptop formfactor (not a tablet one) which makes them inconvenient to use in say a living room when you want to surf on the couch (the keyboard gets in the way).
That's where tablets rule - because you can hold them in one hand and poke at them with the other and not have fiddly bits in the way.
As a laptop, they're awful locked down things that pretend to act like a regular laptop, but aren't.
So basically they're not as friendly as a tablet formfactor, and they're not as useful as a laptop, thus making them rather... useless overall.
Why are manufacturers paying this extortion rather than banding together and trying to fight it like any other patent troll?
What is Google's position on this and why aren't they indemnifying manufacturers that use Android or fighting this themselves?
Because the extortion is cheaper than fighting it out in court. $5/device isn't a lot of money - even if your device sold in SGS3 quantities (over 50M) that's $250M. A good patent lawsuit on the patents Microsoft asserts would run way bigger than that (I think Samsung spent at least that much on the Apple lawsuit).
Thus, it's cheaper overall than to fight it. As for other manufacturers - well, considering Samsung would benefit the most as they own practically 80% of the Android market (while the SGS3 may be a bestseller in Samsung's lineup by selling 50M units, Samsung shipped tons more Android phones - so much that the SGS3 is barely 10-12% of the entire lineup - the rest are all the other Android phones Samsung sells - SGS2, the freebies, the crappies, etc). So the primary beneficiary would be Samsung, something which I think HTC, LG, etc., might be opposed to. Google as well, since I'm not so sure they like how Samsung is starting to dictate how Android should work (or the whole separate ecosystem Samsung is building with their app store).
As for why Google isn't doing it? Well, what's in it for Google? They make no money off Android other than ads sold by showing them in apps. Heck, recent surveys have shown that iOS data traffic still beats Android traffic by 2 to 1, despite Android outselling iOS by 3 to 1 or more. Thus iOS uses are more likely to see ads and Google STILL makes more money from iOS that way. (And it seems advertisers value an iOS user more - they will pay more for an ad impression to an iOS user than Android, apparently 2 to 3 times as much money). Google only does Android to prevent themselves being shut out of a revenue stream (ads on mobiles).
Anyway all this is a older cellphone gaming on your TV with a controller. I am not sure I would qualify this as a console.
Well, I'm sure a very popular thing to do would be to put whatever MAME is best on Android on it, plug in a USB hard drive and play your emulators on the big screen that way. For $99 and a hard drive, you really cannot beat it as a retrogaming console. Of course, those who staked their lives on selling apps are screwed, but no biggie.
Well...it's not to be used to find where people lives, but if it keeps wifi history, then it means it can find where people is *right now* While it can have legit purposes, this could be bad in hands of stalkers, thieves and other criminals, specially when the potential victims have no clue. Also other less criminal but very potentially annoying/conflictive uses if this becomes common knowledge.
Well, given it needs an access point with the same name, I'd say "right now" would be within the range of the access point. Which means they're within about 100 feet or so.
Also, it requires having a network with the same name, so you'll need to make a Linksys network, a Netgear network, etc in order to find one that someone is using.
Though, modern wifi routers people use for internet (i.e., ISP provided) usually use the last 3 octets of the MAC as part of the SSID unless changed. So things aren't as easy. Hell, I haven't seen a Linksys SSID in a long while.
You can do auto- and crosscorrelation in linear time in frequency domain.
... and given a time domain signal, how are you going to switch it to the frequency domain without using an FFT or DCT? The goal is to avoid using a transform...
I think to be fair if most of us were developing a military network in the 60s/70s we might think that 16.7 million addresses is enough. The real blunder was assigning millions and millions to companies and institutions that didn't need them, but again when there is no management structure and no money to set one up and it's a research project anyway...
The biggest failure has been our inability to do anything about it. My ISP hasn't even heard of IPv6 and they are one of the largest.
Well, back then, we used class-based addressing. 0-127 were/8's, 128-172 were/16's and 173-224 were/24s. 225-255 were speciality.
If you were a small business, a class C would be adequate - you were unlikely to ever have more than 256 computers. A large number of companies had between 256 and 65536 computers and thus needed a class B. However, large companies like Apple, Microsoft, IBM, Ford, etc., would potentially easily exceed 64ki (see what I did there?) computers and thus end up with a class A block. Of course, you also needed a new block when you got your 256th computer, or 64ki'th computer.
It was only later (mid 90s or so) that this allocation system was a bit too coarse-grained, so a move to CIDR was done because why waste a whole chunk of IPs just because you exceeded a small number? If you only had 257 things, you still needed a class B and wasted over 65k addresses.
Of course, NAT helped reduce the need even more when most people discovered they don't need complete end-to-end connectivity (which was broken anyways, and will remain so pretty much forever even with IPv6) thanks to firewalls and other things.
Of course, most people never had the niceties of having to renumber their networks, either. Something IPv6 doesn't address properly either - it assumes autoconfig works, or DHCPv6 works, but doesn't allow for the possibility of a bug making things screw up horribly. (The addition of link local and private addresses only serves to confuse - you're going to bet people will wonder why some can see the new server but others can't, etc).
It's one of the things NAT did that made life a bit more bearable - isolating internal network addressing from external addressing. And something that the initial Arpanet guys realized was a huge problem because companies were hooking up their networks only to have a storm of address conflicts.
Except these disks are more standard. They're basically an SSD and a HDD hooked to a SATA multiplexer (that lets you connect more than one SATA device to a SATA port. NOTE: Note all controllers support MUXes. Also, both drives share the bandwidth of the upstream port).
So plug this into a Windows PC and install the drivers, and two drives become one. Plug it into a Linux PC and you see two drives. Plug it into a Windows PC without drivers and again, you get two drives.
Schools have historically shoved their schedules extra-early so that extracurricular events like sports can occur before the sun goes down, but after school.
Solution, abolish sports. You go to school to learn. You can play with your balls on your own time.
You also get a LOT of pushback from the students themselves - while they like the late start, the don't like the evening endings because it means less dollars for their pockets to buy stuff with (i.e., part time jobs, babysitting, chores, etc). Turns out more than a few teens earn a few bucks doing a couple of hours of babysitting on the side before the parents come home. (It's generally a good time for homework and study as well, so... multitask and make money).
Parents generally hate it as well because it means two trips to school - one to drop off the preteens, then to drop off the teen, then pick up the preteen, and then the teen. Which generally screws up the schedule of the teen who has to wake up early anyways. (Schools do like it though - it means you can stagger lunch hours a bit and have a lot less conflict and congestion).
While I agree that the discovery and reporting of these vulns is important, they kinda crossed the line with the break in. They didn't need to compromise the system to know it was vulnerable (in order to report it). It's obvious that Google's reward program is intended to find vulns in Google products. It does not however, give a free license for hackers to break into anything Google owns, especially third party building control systems.
Then again, by compromising the devices, they could launch an attack behind the firewall. After all, there's a difference between read-only access (there was that company saying ADS-B was vulnerable then posting about internet-accessible AIS (marine Automatic Identification System) data saying they could find the location of any ship on the internet - including Navy and Coast Guard. Duh, that's what AIS is for! And it's not like it can't be turned off if operationally necessary), and full read-write access.
Read only access is a lot less scary (big whoop, it's 21 C in the office today, versus 20 yesterday, and the fan on duct #132 is acting up), than read-write (oh, it's a hot day in Sydney, I'm sure Google would love if it I could set this office to 15C and this one to 35C, turn the fan above the meeting room to max).
Sometimes you have to break in to figure out if you have full access or just limited access - because the limited access may be neat, but not useful at all (like AIS data - it's not terribly useful when it's hooked to an AIS receiver).
Also, some of these vulnerabilities may not be terribly important to Google - because Google properly firewalled it off. Or maybe it is because it's behind the firewall. You can bet a lot of other building automation systems may not have the internet savvy that Google has. Or maybe a misconfiguration in Google's network or someone's PC could serve as a launch point.
Slashdot Mirror
At least one TV network has realized this - and it's getting fairly good ratings (as good as network TV gets these days) because of it - CBS. Sure they have SOME reality shows (who doesn't? And you need a few for the few viewers you have that demand it). But they have a LOT of the more traditional scripted and written shows. And they consistently do a lot better - the scripted shows generally outperform the reality ones these days.
Voltage regulators typically run hot because of the pass transistor. In a linear supply, the pass transistor is what dissipates the difference in voltage from input to output. In a switching supply, the pass transistor is the chopper (the one that converts the DC to AC for the transformer or magnetics). Most VRMs in computers are actually cooled just passively - if you look, the tab is soldered to the motherboard and a huge copper plane on the PCB (usually via-stitched to carry heat to the ground or power plane as appropriate).
HOWEVER, there is one thing the on-die VRM has over the on-motherboard one - a heatsink!
The main CPU core has to be cooled anyhow, so the VRM has a very easy access to the heatsink and can thus dissipate the heat much easier.
I would guess one of the reasons for doing it on board is because of the currents and voltages involved - power supply is becoming a big issue when you have multiple voltage rails (Vcore, Vgpu, Vio (x a dozen or so depending on the IO), etc). Each voltage rail requires at least one pin or multiple pins, and even though we have 1100+ pins, it still means the CPU is pin-limited (in ASICs, you can be pin-limited, where what you can stuff onto an IC is limited by the number of pins you can have, or silicon-limited, where what you can stuff in is limited by silicon area. SoCs and such are typically pin-limited, while memories are silicon limited. Similar concept to IO bound and CPU bound programs, really.)
If instead of dedicating easily 100 pins to power (if your core is 1.8V and you're drawing 90W, you're wanting 50A, and each pin may take .75-1A max meaning you're looking at 50-75 pins, plus power to other onboard stuff), you can dedicate 25 to higher voltage single rail, it's a lot less current per pin, so lower IIR losses, and a lot more pins freed up. If you need 90W, and you feed in say, 15V, you only need 6A, so instead of the 50-75 pins to carry 50A, you only need say, 10 pins. Plus, if you want other voltages, instead of dedicating more pins, the onboard VRM gives you more flexibility - say you want another 60W, that's 150W, at 15V, that's 10A, or 10-15 pins. Still a lot less and giving you 80-odd more pins for further I/O.
80 pins is a lot - you can add a number of additional PCIe channels and lanes with that, for example. Or additional functions.
And you've simplified the motherboard design a lot and gotten rid of a potential source of issues - poor power supply design by motherboard manufacturers (applies mostly to cheap budget boards).
Only if you can accept payments via Google Checkout/Wallet.
You see, the Play market is available in more countries that you can accept payment from - paid apps are simply not shown to those countries. (For a long time after Android first came out, the only supported country was the US, while the phone was available in a number of other countries, including Canada. End result is if you wanted a paid app, you had to pirate it because Google didn't make it available to you).
As a result, many apps went "freemium" because it's either not show up and be pirated, or go free and make up the money selling ads and in-app purchases (the latter being fairly new, which is why many apps use the ad model with fairly ... "liberating" permissions required).
So if you're selling in a country that can accept Google payments, you must use it. If not, you're free to engage Paypal and others (which until Google allowed in-app payments most devs used Paypal, and even after IAPs until Google cracked down).
Fact remains, Play is available in more countries than you can accept money from.
Better yet, why isn't it downloading on behalf of the installer and letting the INSTALLER ask for admin rights?
Half the time, it claims there's an update, and then it promptly fails to download it. After giving it admin. Why not attempt to download it ahead of time?
Yes, ask for admin if it would save needless popups (e.g., Windows Update - where installing multiple patches may require admin priviledges - so ask it up front then use that to run the patch installers as admin to avoid bugging the user).
Problem is, people aren't sheeple. They can spot marketing BS a mile away.
If a product is all hype and no substance, guess what? It dies. And in the age of instant news, if a product is crap, people know about it the moment it's released. Doesn't matter if it's an iPhone, a movie, a new store, etc. It doesn't matter if the revised edition is better or not.
Hell, see Vista. Despite marketing claims (Project Mojave, anyone?), people still thing it's a POS. Even though in the Windows 7 era, Vista benefits from developers doing crap right and things actually work. (Vista took the crap for Windows 7, you could say).
Hell, Linux still has a "hard to use" reputation, despite it being easier to install and update than Windows (though, some of it is deserved - some support communities can really do better than "RTFM" as a reply. Like "RTFM section X, paragraph y" linked).
Basically, if the iPhone was crap, then people would've known 6 years ago. Or on release of every new model ever since. But since Apple can pull in the profits, they obviously either managed to brainwash everyone (which is a very politically powerful tool wasted on getting people to buy a brand of computers?), or have a product that serves its purpose.
Sure, an iPhone won't meet any measurebator's specs, but do we really need to get back to the megapixel wars, GHz wars, GByte wars, etc? Or has it been definitely proven that more megapixels are better, more GHz are better, more GBs are better (OK, the last one is definitely better).
Hell, look at what Apple started with the "retina" display - more and more DPI until I'm sure few can even see the pixels up close (441 - the 5" 1080p screens). Or phones getting so big that people are making "small phones" to go with your big one. (NOTE: The SGS3, while the best selling Android phone, only has around 10% of the market. The rest of them are the hundreds of other phones Samsung makes with smaller screens, keyboards, crappy processors, etc, and some others by LG, HTC, Motorola, etc).
tl;dr - marketing can sell you a product, but it can't work a miracle. A turd is still a turd and no amount of marketing can spruce it up.
Before he launched in December, Chris mentioned he was going to do the first album recorded in space, I'm hoping this was just a taste of what's coming.
I have to be honest, I've been watching a LOT of Chris' videos that get posted by the CSA (Canadian Space Agency) (an agency facing budget cuts from the Harper Government(tm)). I don't think I've seen anyone from the ISS do so much media relations in their off time.
I know a few other commanders have done media work - Don Pettit did some as well. With the American Physical Society (any physics major should know them) he did a bunch of videos called "Science off the Sphere" (which I apparently finally got my T-Shirt from that).
Chris is definitely very media friendly and has hosted a LOT Of media events while aboard - he even keeps in touch with Discovery Canada's Daily Planet, the longest running science program around. Honestly, Chris Hadfield is awesome!
Safe journey home - your country eagerly awaits your arrival!
(Alas, Canada's first astronaut was snubbed recently...)
If you're in Canada, Daily Planet carried this car earlier this month. They mentioned they were doing test flights.
OF course, you get to see a rather interesting takeoff int he clip. Alas, I think it's Canada only - not sure if the US Discovery channel has it on any of their channels (it's a Canadian production).
Guess we might see an update shortly.
Except around the 3Gs era, Apple started hard-encrypting the flash to prevent that very attack.
And JTAG ports can be disabled by software - I've worked on devices where once a fuse was blown, the JTAG lines were disconnected internally and thus inaccessible.
And yes, Apple is the only one that can do it because they hold the keys. If you need to load out special software, only Apple has the private key to sign and run whatever tools they have. And they can probably read out the filesystem, figure out what the keys are and brute force what they need to brute force.
Of course, the article doesn't say Apple has any success at all - perhaps they can crack the 4 digit passcodes that bypass the 10 code self-lockout and erasure. But you can enable a more secure form using a complex passcode. And supposedly you can enable even more sophisticated encrypted and protections.
Or hell, we don't even know how may phones are in the queu or how long it takes Apple, It could take Apple 4 months to decrypt the iPhone. Or they may have a backlog because only one person is decrypting them and he can only do one a day or something.
So you want the Android and iOS way of using the internet then? Where you interact with apps to do your banking, check maps, check what food is nearby, get ratings on movies/shows/food, etc?
You know, Steve Jobs was onto something with the original iPhone - where developers would be forced to do everything in HTML and Javascript. But developers lobbied for a native SDK, leading to app stores and the rest as we know it today. Now we have people complaining that why are people using apps now when before, they could've just used the web? People are making apps that are basically web site containers. Or publishing apps that are basically sets of static HTML content in a web view.
And not on the web.
Hell, people use the Amazon app to buy stuff online. You go into the store, scan the barcode with the amazon app, and click buy. You could do the same by browsing Amazon's web page on your browser, but it's less convenient and easy. And if the trend continues, Amazon may decide that their website links you to their app page to finish the purchase because the vast majority of people are buying stuff through the app. Right now it hasn't happened. But who knows?
In the end, an app based world is one where everything is segregated - you use one app for your sports scores, another for your banking, a third for shopping, etc. And there'll be this dusty app called a "web browser" that once was used for that stuff, but because everyone used apps instead, slowly lost prominence as apps could have better security, could display content better, could do advertising better, etc. etc. etc. I'm sure Netflix's main usage now is apps (for smartphones, set top boxes, etc), and a very tiny portion of it is pure web access. Silverlight is dead. It's only a matter of time before Netflix pulls web access to its content because everyone uses the Netflix app anyway.\
Hell, forums on the web are being taken over by apps - you may have heard of things like Tapatalk and Forum Runner.
Where do you draw the line, though?
If I make a robot that's completely controlled by software so that the only thing mechanical I have are motors and sensors, that means I can only patent whatever's in that heap of metal? If I come up with a clever way of doing something like being able to get precision out of imprecise hardware, it's not patentable? Whereas, if I stuck that software inside a hardware implementation (software runs on hardware - I can make fixed hardware do the same thing), I can now patent it?
That's the tricky part. It means the cheap hardware and doing everything in software solution will lose out to the expensive hardware because the latter can contain patents, but the former will have very little actually patentable, because it's all software.
Or consider something like a software defined radio. Very little hardware, and beyond the basics, everything's in software. But if I invent a clever way of using the radio waves, if I do it in hardware, I can patent the crap out of it, but if I use an SDR, I can't?
The main problem with software patents is software is nebulous. There is nothing like it before. Before the computer age, we had "stuff" you could patent, and "creative works" which you copyrighted. But now software bridges the two - is it a create work because it's written, or is it "stuff" because it's a fundamental part of whatever machinery it controls? Nevermind RTL software written and compiled to ASICs and FPGAs.
Software has basically ended up taking two normally incompatible things (you don't patent books, like you don't copyright machinery) and fused them together. Your clever machinery depends on software for part of its operation - together it's interesting, separately, it's a hunk of metal and writing and both are useless apart.
Nevermind software generating software - is the output of such copyrightable? What if software generated a book? What if the inputs to the software are under other copyright?
In fact, what might end up happening is a merging of the two - special protections for software only that combine parts of copyrights (the code itself, and any output it generates), and patents (when embedded with hardware to perform a novel task), thus sidestepping the whole issue entirely - software cannot be patented or copyrighted, but it falls under software based IP.
Incorrect.
Flagship phones are expensive. But you can find unsubsidized Android phones at all price ranges. And no, the most popular phones are NOT the flagships. The SGS3, last year's flagship, sold a bit over 50M units (March 2013). At the time, Google was activating anywhere from 1 (beginning of 2012) to 1.5M (October?) Android phones per day. If we take the numbers, we see roughly 1 in 10 Android phones was an SGS3, the most popular Android phone model ever. That leaves 9 random Android phones out there, and Samsung has around 80% of the Android market, so 7 of the 10 are Samsung Android phones, and the other 2 are from LG, HTC, Motorola, etc. The other flagships don't even register.
Which makes the vast majority of Android phones the cheaper ones. Granted, they're "free with contract" but generally around $200 unsubsidized if you look at the carrier offerings. Of course, they're also useless pieces of crap - either running 2.1 or 2.2 or if you're lucky, 2.3 (though some may have 4.0). Piss-poor WVGA ish screens, sluggish processors (1GHz if you're lucky), low RAM, etc.
Basically, the price has come down, if you want a POS Android phone.
Tablets the same - excluding the Nexus 7, the $200 and under tablets generally are crap. WVGA is common, 1024x600 if you're lucky. Then you have the Nexus 7, the best of the lot, but forced to compete with sub-par. The 10" tablets are already $400+ with the iPad being the most expensive and followed closely by the Samsungs. But they're generally around there.
No, prices have not gone down. All you're seeing is the "build to a cost" nature of it, like the crappy $500 laptops (the ones that are generally good start at $1000 where you can find discrete GPU, >1366x768, and quality). It's just people figured out how to cut corners and make cheap crap.
And no, despite "subsidies going away", what happens is people are opting for the Tab model - where they get the phone "free" and their tab gets paid down off their monthly bill. The only difference is the price is shown externally and you get a small discount on your bill at the end of the contract (for the phone purchase).
And generally, people will choose the cheapest phones still - your flagships will still cost $200 with $400 "tab", while the rest are "free on the tab". Samsung and others will, however, make tons of similar-but-cheaper-crappier phones so you have to specifically ask for an SGS4, not and SGS4 Lite, Micro, Grand, Mega, (Did you know you can still buy an SGS2? Not necessarily the same SGS2 as was the flagship, but they have phones called SGS2).
Yes, because streaming one stream per viewer is far more efficient bandwidth wise than broadcasting it to everyone. After all, bandwidth needs of the former scale linearly while it stays flat with the latter. The real problem happens when something big happens and everyone turns on their TV - I don't think handling the 300M-odd TVs is possible right now. (Most streaming networks have far fewer streams than that).
Of course, companies like Netflix do it by not bogging down the Internet - but by creating CDNs and putting them in ISP data centers - because the ISP to subscriber link, despite oversubscription, is still fatter than the ISP to internet peering link.
Anyhow, if you have seen TV, you'll note that each network of channels have been moving their flagship shows to the secondary channels as well - the other channels in the network get older seasons. This is so popular shows people watch together end up being on different channels, so they have to purchase both channels in the a la carte model. (On the bundle model, they're usually "free" because they're bundled together). So while you used to just ignore the dozen other Discovery-owned channels for the main one, soon you have to purchase at least half of them because the main channel has moved new seasons to other channels.
At least that's what I've seen - I've had to check a few channels because they seem to move channels a lot.
Actually, is it as cheap as people say it is? During operation, yes, it's very cheap. But there's a long tail of maintenance that goes on long after the money making days are over. Like cleanup which often takes decades, plus the storage and handling of the waste which takes a long time as well. Those are costs that have to be borne for years after the plant shuts down, and it seems for a lot of them, it's offloaded to taxpayers as the company running them moves onto other things to please shareholders who get the reap the profits and none of the final costs of shutdown.
In general, the limiting factor is the quality of the hydrogen - the oxygen can generally come straight from the atmosphere. In this case, it's pure hydrogen stored as a liquid, so fuel cells tend to be fairly efficient.
Take Apple's fuel cells powering their data center, and they're running off natural gas, which is 20% carbon for the most part (by stoichiometry) which is impure and has to be dealt with. And nevermind that natural gas, while being mostly methane, also contains longer gaseous hydrocarbons that provide more "poison" to the cell that has to be dealt with.
Of course, you could refine methane to hydrogen and carbon, but hydrogen itself is rather difficult to transport since it doesn't like to be contained easily.
Not really. Android's the most popular mobile OS, yes, but that's just because it's like PCs - it's being put on devices that Apple won't touch.
Samsung has around 80% of the Android market. But the SGS3, the best selling Android smartphone, only sold 50-odd million phones. Out of the entire Android ecosystem, that's roughly 8-10% of the market - by far the market leader for the model, but a majority of phones sold. That means out of the 8 out of 10 phones Samsung sells, only 1 of them is an SGS3. The other 7? Crap phones you see advertised - SGS2 (yes, they still make them) various flavors, the other crap ones with crappy WVGA screens, or slow processors, or tiny RAM. Basically, super cheap phones the carriers are pushing out to sell everyone a data plan and lock them into whatever crap else there is. (If you look carefully - you'll probably pay MORE for a "dumbphone" than an Android phone). And that's the top seller - the other flagship phones? Not even the radar.
So by marketshare, Android phones routinely outsell iOS by 3 to 1 or so, or more.
However, an interesting stat keeps popping up - by data usage, iOS users tend to use their data plans - especially websurfing traffic where Safari tends to out-do Android by 2 to 1. So it seems almost exclusively all the Android traffic is driven by high end flagship phone users (makes sense - they're the ones who tend to be willing to pay and actively choose Android over iOS - the other Android users tend to pick it because they don't want to pay the $200 an iPhone costs (don't argue about the "cheap" models - if you go into any store, only the iPhone 5 is prominently displayed), and carriers are literally shoving phones out the.). Other than mobile OS choice, I'd say flagship Android users would be fairly similar demographically to iPhone users.
This has resulted in an interesting situation - advertisers who can't target just flagship Android phone users, are basically paying more for iOS ads. And Google, despite having a majority of phones running Android, still makes more money from iOS.
Odd.
I mean, a quick search through Appshopper brings up many OBD-II reader programs.
Dash Command
Dyno Chart
Rev
and dozens of others (ignoring the ones that don't use an OBD-II reader and are merely databases).
and here are some iOS compatible readers:
Kiwi 2 Wifi
OBD Key.
Yes, most of them use wifi instead of Bluetooth. While iOS does support serial port profile, they don't expose it for whatever reason.
Problem is, Chromebooks are crappy laptops - they have a laptop formfactor (not a tablet one) which makes them inconvenient to use in say a living room when you want to surf on the couch (the keyboard gets in the way).
That's where tablets rule - because you can hold them in one hand and poke at them with the other and not have fiddly bits in the way.
As a laptop, they're awful locked down things that pretend to act like a regular laptop, but aren't.
So basically they're not as friendly as a tablet formfactor, and they're not as useful as a laptop, thus making them rather... useless overall.
Because the extortion is cheaper than fighting it out in court. $5/device isn't a lot of money - even if your device sold in SGS3 quantities (over 50M) that's $250M. A good patent lawsuit on the patents Microsoft asserts would run way bigger than that (I think Samsung spent at least that much on the Apple lawsuit).
Thus, it's cheaper overall than to fight it. As for other manufacturers - well, considering Samsung would benefit the most as they own practically 80% of the Android market (while the SGS3 may be a bestseller in Samsung's lineup by selling 50M units, Samsung shipped tons more Android phones - so much that the SGS3 is barely 10-12% of the entire lineup - the rest are all the other Android phones Samsung sells - SGS2, the freebies, the crappies, etc). So the primary beneficiary would be Samsung, something which I think HTC, LG, etc., might be opposed to. Google as well, since I'm not so sure they like how Samsung is starting to dictate how Android should work (or the whole separate ecosystem Samsung is building with their app store).
As for why Google isn't doing it? Well, what's in it for Google? They make no money off Android other than ads sold by showing them in apps. Heck, recent surveys have shown that iOS data traffic still beats Android traffic by 2 to 1, despite Android outselling iOS by 3 to 1 or more. Thus iOS uses are more likely to see ads and Google STILL makes more money from iOS that way. (And it seems advertisers value an iOS user more - they will pay more for an ad impression to an iOS user than Android, apparently 2 to 3 times as much money). Google only does Android to prevent themselves being shut out of a revenue stream (ads on mobiles).
Well, I'm sure a very popular thing to do would be to put whatever MAME is best on Android on it, plug in a USB hard drive and play your emulators on the big screen that way. For $99 and a hard drive, you really cannot beat it as a retrogaming console. Of course, those who staked their lives on selling apps are screwed, but no biggie.
Well, given it needs an access point with the same name, I'd say "right now" would be within the range of the access point. Which means they're within about 100 feet or so.
Also, it requires having a network with the same name, so you'll need to make a Linksys network, a Netgear network, etc in order to find one that someone is using.
Though, modern wifi routers people use for internet (i.e., ISP provided) usually use the last 3 octets of the MAC as part of the SSID unless changed. So things aren't as easy. Hell, I haven't seen a Linksys SSID in a long while.
Well, back then, we used class-based addressing. 0-127 were /8's, 128-172 were /16's and 173-224 were /24s. 225-255 were speciality.
If you were a small business, a class C would be adequate - you were unlikely to ever have more than 256 computers. A large number of companies had between 256 and 65536 computers and thus needed a class B. However, large companies like Apple, Microsoft, IBM, Ford, etc., would potentially easily exceed 64ki (see what I did there?) computers and thus end up with a class A block. Of course, you also needed a new block when you got your 256th computer, or 64ki'th computer.
It was only later (mid 90s or so) that this allocation system was a bit too coarse-grained, so a move to CIDR was done because why waste a whole chunk of IPs just because you exceeded a small number? If you only had 257 things, you still needed a class B and wasted over 65k addresses.
Of course, NAT helped reduce the need even more when most people discovered they don't need complete end-to-end connectivity (which was broken anyways, and will remain so pretty much forever even with IPv6) thanks to firewalls and other things.
Of course, most people never had the niceties of having to renumber their networks, either. Something IPv6 doesn't address properly either - it assumes autoconfig works, or DHCPv6 works, but doesn't allow for the possibility of a bug making things screw up horribly. (The addition of link local and private addresses only serves to confuse - you're going to bet people will wonder why some can see the new server but others can't, etc).
It's one of the things NAT did that made life a bit more bearable - isolating internal network addressing from external addressing. And something that the initial Arpanet guys realized was a huge problem because companies were hooking up their networks only to have a storm of address conflicts.
Except these disks are more standard. They're basically an SSD and a HDD hooked to a SATA multiplexer (that lets you connect more than one SATA device to a SATA port. NOTE: Note all controllers support MUXes. Also, both drives share the bandwidth of the upstream port).
So plug this into a Windows PC and install the drivers, and two drives become one. Plug it into a Linux PC and you see two drives. Plug it into a Windows PC without drivers and again, you get two drives.
You also get a LOT of pushback from the students themselves - while they like the late start, the don't like the evening endings because it means less dollars for their pockets to buy stuff with (i.e., part time jobs, babysitting, chores, etc). Turns out more than a few teens earn a few bucks doing a couple of hours of babysitting on the side before the parents come home. (It's generally a good time for homework and study as well, so ... multitask and make money).
Parents generally hate it as well because it means two trips to school - one to drop off the preteens, then to drop off the teen, then pick up the preteen, and then the teen. Which generally screws up the schedule of the teen who has to wake up early anyways. (Schools do like it though - it means you can stagger lunch hours a bit and have a lot less conflict and congestion).
Then again, by compromising the devices, they could launch an attack behind the firewall. After all, there's a difference between read-only access (there was that company saying ADS-B was vulnerable then posting about internet-accessible AIS (marine Automatic Identification System) data saying they could find the location of any ship on the internet - including Navy and Coast Guard. Duh, that's what AIS is for! And it's not like it can't be turned off if operationally necessary), and full read-write access.
Read only access is a lot less scary (big whoop, it's 21 C in the office today, versus 20 yesterday, and the fan on duct #132 is acting up), than read-write (oh, it's a hot day in Sydney, I'm sure Google would love if it I could set this office to 15C and this one to 35C, turn the fan above the meeting room to max).
Sometimes you have to break in to figure out if you have full access or just limited access - because the limited access may be neat, but not useful at all (like AIS data - it's not terribly useful when it's hooked to an AIS receiver).
Also, some of these vulnerabilities may not be terribly important to Google - because Google properly firewalled it off. Or maybe it is because it's behind the firewall. You can bet a lot of other building automation systems may not have the internet savvy that Google has. Or maybe a misconfiguration in Google's network or someone's PC could serve as a launch point.