Forced? Really? I do not think that word means what you think it means.
This reminds me of a friend of mine who was looking for a house to buy and wanted to sign up for a month-to-month lease on their apartment. The landlord came back with a monthly cost 2.5x higher than the standard 12-month price.
Forced. Coerced. Convinced. Pick whatever word you want to describe little or no options left on the table that make financial sense.
Given that a high-end smartphone is a luxury, not a necessity, none of those words apply.
However, they can't actually afford a $6K per month mortgage payment
Then something is jacked up if they qualified for it.
Banks will always loan you more than you can actually afford.
Plus, you can always tell the bank you're planning to rent a room. That won't work if you're clearly in over your head, but if you're on the edge (per their very optimistic model), it will.
I also have doubts that Apple was completely unaware that their phones would, in fact, be sold under two year contracts when they agreed to sell them to the carriers.
I don't see how that's at all relevant to Apple. What financing or other arrangements are made between users and carriers is none of Apple's business. Apple communicated their actual warranty to both carriers and users and made no claims -- to anyone -- that they would warranty phones for the period of some contract between third parties.
I don't think your argument would hold water even if the users were buying the phones from Apple directly with two-year financing from Apple. When you buy a new car from Ford, there's no expectation that the car will continue functioning for the duration of the 72-month purchase contract. You *do* get various levels of warranty on various parts of the vehicle, but those are all completely independent of the financing terms.
Would you argue that someone who bought their car / phone for cash should get less warranty coverage than someone who financed it? This is silly. The one thing has nothing to do with the other.
Either the buyer's an idiot, with a) more money than sense, or b) thinks they'll be rich any minute now; or else they're fucking flippers, who'll redo the kitchen, slap some paint on, and try to resell it.
Most likely it's neither. Instead, it's just someone buying a home that's close to work, for the going rate. Sellers in Silicon Valley don't even bother trying to price their houses accurately, they just ask the minimum they need to get, and know that the price will be bid up substantially. Odds are good that the buyer can't even afford the mortgage and will have to rent out part of the home to make ends meet (this is extremely common in the area).
But, the buyer also knows that unless something changes, buying the house is a good deal for them. They'll get every penny of equity back out when they sell in a few years, and probably get a hefty return on their investment, which is much better than if they'd just kept paying $4K+ per month in rent.
I know a few people who have done exactly this, though their homes have all been around $1.5M, not $2.5M. They buy a house and acquire a mortgage payment of just under $6K per month. $2.5K of that is principal, so their net cost for housing (ignoring the tax deduction, which they don't ignore) is $3.5K per month, which is about what they'd be paying in rent for an apartment large enough for a small family anyway.
However, they can't actually afford a $6K per month mortgage payment. So, they convert the garage into an apartment and rent it out, or rent out the master bedroom/bath. Some single guy pays them $2.5K per month to live in part of their house which lowers their monthly out of pocket cost to $3.5K per month... and lowers their net housing cost to $1K per month.
Bottom line, as long as they get their equity back out when they sell, this scheme is a much better deal than renting. If housing prices rise, they'll do even better and will probably never pay any taxes on the capital gains, which could be substantial. If the bottom drops out of housing prices, they could get severely screwed, of course.
If you are able to verify your vote after you leave the polling station, then someone else is potentially able to watch over your shoulder while you do so, and could therefore make good on any threat they'd provided you to ensure you voted the way they wanted you to.
Nope. It seems obvious that this should be impossible, but it's not. You should read about it.
If something is properly encrypted, disassembling a chip won't help.
I wish. (Really. Designing airtight cryptographic security for phones is my day job.)
The secrets used to encrypt the data must either be embedded in the device or obtained from outside of it, or some mixture of the two. Since the only practical outside source of key material is the user, and users suck at generating, managing and entering high-entropy secrets, the vast majority of the key material must come from the chip.
You can (and devices do) use key stretching, and you can (and devices do) implement various other brute force mitigations in software and hardware, but at the end of the day whatever software does can be replicated and hardware can be broken. So it really all comes down to the (lousy) user password.
Still, disassembling a chip is a pretty high barrier. And the right kind of chip can make it harder yet. At the end of the day there's nothing you can do in a consumer device that will defeat an adversary with serious engineering expertise and willing to spend a few hundred thousand dollars. Or one willing to spend a few hundred dollars and who is able to surreptitiously swap your device for one that they configured to look like yours, but to snarf your password.
I think telecommuting also makes a lot of sense, at least for the employees who find it appealing and can work effectively that way. I've telecommuted full time for about 15 of the last 20 years, and I think it's awesome. But many companies don't like telecommuting for various reasons that I don't fully understand, and I'm not sure they do either (Google doesn't; about one in 2000 Google engineers work remotely. No, that's not an exaggeration, if anything I'm overestimating the number of remote engineers).
Because they quickly realized that if you can do the job from 200 miles away then you can do it from 2000 miles away. All telecommuting did was make offshoring even more attractive.
That's a reason that companies should like telecommuting, not reject it.
Even IBM (before they went nuts and moved everyone to India) and other deep-pocketed companies had them back in the day, and that was when it was harder to stay in touch. The only difference was that the office was in Pittsburgh and not Pune, or Moline and not Mumbai. I remember reading something some time back that mentioned IBM would strategically locate big engineering facilities just far enough away from large business centers to be a short flight or medium length drive. They'd import the workers or hire from local university talent pools, and the execs would be mollified because they still felt like they had control. IBM used to have big facilities in Burlington, VT and Rochester, MN that fit that description perfectly. They probably didn't have to pay anything near what they'd have to pay for people in Westchester or Dutchess County, NY.
I worked for IBM back then, and that approach had its own share of problems. Fairly large ones.
The biggest was that those big engineering offices utterly dominated the local economy. Effectively, they created company towns, which meant that everyone who joined IBM had to move to one of the towns, and everyone who left IBM had to move out of one of the towns. This sucked for employees. The fact that there were several such locations meant that transferring to a different job within the company also frequently required relocation. When I first joined the company, employees said that IBM stood for "I've Been Moved".
Another was that a huge amount of IBM's capital was tied up in real estate. Rumor has it that when Louis Gerstner took over as CEO and surveyed the company's balance sheet, he said "Is this a computer company or a real estate company?". His fix to this problem greatly improved IBM's return on assets, but created its own problems. First, he sold most of the real estate and leased it back. That freed up a lot of capital and increased flexibility. Second, he started a program to push employees out of the office, making many of us telecommuters. Personally, I thought that was awesome, but not everyone likes to work that way. Employees began saying that IBM now meant "I'm By Myself", particularly since another Gerstner initiative was the push into the services business, which meant that employees were often working at customer sites, frequently on small contracts where they might be the only IBMer present for periods of time.
Of course, later Palmisano and then Rometty instituted the new IBM direction, which was to lay off the American employees and move everything overseas, to India, Romania, Brazil, etc.
I think satellite offices make a lot of sense, but I think it's better to locate them in regional tech hubs rather than out in the sticks the way IBM did. That makes it more likely that people can join or leave without having to relocate, giving both company and employees more flexibility. Google (my current employer) does this to some degree, maintaining engineering offices in many of the tech hubs. I think Google should do more of it, but the leadership seems to like at least half of the engineers to be located in Silicon Valley. It does allow employees to transition between teams easily, without having to relocate... but it means they have to live in Silicon Valley.
I think telecommuting also makes a lot of sense, at least for the employees who find it appealing and can work effectively that way. I've telecommuted full time for about 15 of the last 20 years, and I think it's awesome. But many companies don't like telecommuting for various reasons that I don't fully understand, and I'm not sure they do either (Google doesn't; about one in 2000 Google engineers work remotely. No, that's not an exaggeration, if anything I'm overestimating the number of remote engineers).
Just 2-3 weeks ago Jan T Sanner, the minister with responsibility for elections, decided that every single vote had to be counted manually, including all early voting ballots.
In my Virginia Beach precinct, we had electronic voting machines a while ago, but have had paper -- fill in the bubble, then scanned -- ballots for the past several years including the 2016 election. The ballots are scanned on their way into the locked ballot box. This system is easier and faster than the electronic versions were, plus there's a paper trail.
With a bit more it's possible to go a step further, and get election systems that not only have a verifiable paper trail, but which are end-to-end verifiable, allowing any voter to check after the fact whether or not their vote was included in the tally (but without being able to prove to anyone how they voted), and allowing anyone to verify the correctness of the tally. The method relies on applying the concepts and methods of modern cryptographic proofs to the problem of voting. It not only ensures that ballot scanning is not less secure than non-automated methods, it achieves a provable level of election integrity that has never before been possible.
As for intelligence - America has had a strong anti-intellectual bent from its inception "nurtured by the false notion that democracy means that "my ignorance is just as good as your knowledge." " -- Isaac Asimov
Some parts of America have long had the anti-intellectual bent, which they acquired largely in reaction to the pro-intellectual and zealous progressive proselytization of the Yankee northeast. From the 17th century the Yankee northeast revered education above wealth and birth and deeply trusted in collective, democratic community decisionmaking over aristocratic and other tradition, and worked to spread their gospel (originally religious gospel; later a sort of secular humanist gospel).
Meanwhile, the Virginia Tidewater region, settled by second sons of aristocrats interested in establishing and supporting their own hereditary aristocracy, also appreciated education, but of the classical, scholastic, let's-act-like-upper-class-Romans sort, rather than the progressive sort. The Deep South society, modeled after the slave plantations of Barbados, just wanted an ironclad aristocratic structure to keep the slaves from rising. They had little interest in, or opposition to, education for the upper class but actively disliked democracy unless it was tightly restricted. Both southern cultures disagreed with and disliked the Yankees, and resented their pushy progressivism (regarding abolition and much more).
And then there were the Applachian regions, with a culture derived from the first Scottish highlander settlers, violent and almost ridiculously independent. They also were deeply skeptical of the opinions of educated snobs and had no real interest in community much above the level of the family, or any sort of democracy, but just wanted to be left alone.
The other early American cultures, of New York (thoroughly commercial above all), the Midlands (pluralistic and middle class) and New France (egalitarian and tolerant) weren't so anti-intellectual, nor even so anti-Yankee, but all objected to the pushy self-assurance of the Yankee northeast.
Of course, history has mostly vindicated the progressive positions of the 18th-century Yankees, whose ideas were the core of the US Constitution. But their confidence in their own rightness and righteousness, that they knew what was better for others than those others knew themselves, is almost entirely responsible for the anti-intellectual element of American thought.
they know that they could go to a community college for 2 years (and we're in Canada where education doesn't cost as much as in the US), then do an apprenticeship for 2 years, and they'd be making double what they make now, and hardly anyone's doing it
I see what you're saying, but it's hardly proof that they wouldn't be willing to get the education they need if they had to. Right now, they'd like to make more money and have higher status in the organization, but not enough to put in the four years. Given more incentive -- perhaps with some funding to help them make the transition -- I think most of them would do it.
Of course, this presumes that there are any jobs for them to train into. That depends on the progress of AI research.
Current theories of economics are flawed, being based on assumptions of "infinite" that are no longer true.
Current theories of economics are in no way based on assumptions of "infinite" anything, they're very much based on the assumption that resources are scarce, have multiple uses, and that we must decide efficiently between those uses. Actually, it looks like what's going to invalidate current economic theories is that certain resources are going to become effectively "infinite". In particular, labor.
Currently, our economic system is structured around the notion that labor is scarce, and skilled labor is very scarce. That being the case, we need to incentivize people to work, and incentivize them even more to get education/training. So we have a system that more or less says "work or starve", because it is so important to make people work so that we can produce all the stuff that everyone needs to live.
Note that this structure continues right up to the top of the economic ladder, where it's focused not so much on producing goods as on optimizing the flow of resources needed to produce goods... including capital. Bankers and investors are rewarded by making sure that resources are routed away from unprofitable enterprises (those that don't use their resources wisely to generate value for people) and to profitable enterprises (those that do).
(Note that I'm not claiming the system is optimal, merely that it is closer to optimal than anything else we've tried at scale.)
What we're facing now is a radical change of conditions, in which it's possible that basically all unskilled labor, and a great deal of skilled labor -- possibly all of it! -- is going to be replaced by machines which are more efficient than humans in every way. Faster, stronger, tireless, more precise, perhaps even smarter, and, most important, "cheaper", meaning that they consume less resources to create and maintain.
Assuming that the robots actually work for us rather than deciding to eliminate us as wasteful and inefficient, this means that the cost of production is going to fall dramatically. Probably to nearly zero, as the only true constraints on production are raw materials and energy, which typically make up on a tiny, tiny fraction of the cost of most of our goods today.
That's fantastic, since it means that we (humanity, collectively) are going to be dramatically richer than we are today. But our current method of distributing wealth is based on the assumption that labor is scarce and thus needs deep, systematic and even coercive incentives. That assumption is going to become increasingly false, precisely because of post-scarcity "infinite" productivity.
Yeah, slashdot uses HSTS, so your browser should never sent HTTP requests to slashdot.
FWIW http://www.google.com/ does not use HSTS. For this exact reason, I'd guess. So your browser will send HTTP requests to google.com. If the request isn't intercepted you'll get back a 301 redirecting you to https://www.google.com/ so in the normal case it will immediately switch to TLS mode, but in the case that it's intercepted by a login page the login page will work.
I have no idea how long that will continue to be the case.
Yes, I know it is. And if the browser would let me go to the damn page, I could get rid of it.
Go to an HTTP URL. Then the hotel proxy can intercept it without trouble, you can get logged in, then go to the page you actually wanted to visit.
This is a simple, excellent solution for people who understand what's going on. I have no idea how the other 99% of the population is supposed to use hotel Wifi. Hotels may have to find another approach when the vast majority of the web implements HSTS.
Actually, I think what will really happen is that browsers and/or OSes will do Wifi login page detection. Of the OSes that I often use, Android and Mac OS X already do. When a Wifi connection is established, they send a test HTTP request to a particular HTTP-only site. If the data they get is what's expected, great. If they get a Wifi login page, they prompt the user to log in.
Yes wanting to know what my bluray playing is up to when it calls home, makes me a bad person. Check
Your bluray player will accept the self-signed cert provided by your proxy as legitimate, instead of expecting the cert of its actual home server? That's possible, but it just means the bluray player software is broken, and if the manufacturer is really trying to do something sneaky, they'll fix it in the next update as soon as they realize someone is looking.
Google doesn't give a flying fuck about you and your files beyond their ability to mine them for data that they can monetize by selling your privacy away to the highest bidder.
Two errors in the above:
First, Google doesn't data mine Drive files, unless the file in question is marked publicly-accessible by the user.
Actually, I doubt they mine even public docs for information about the doc owner; my guess that the terms of service include that escape hatch for public docs because they get added to the search index, and so searches may turn them up with ads alongside which would constitute a "marketing or promotional campaign", in the words of the ToS.
Second, Google doesn't sell user data, even for products that are advertising-based. Google makes money by using the data to target ads, not by selling the data to advertisers (or anyone else).
Remember: if something is free, you're not the customer... you're the product.
Often true, but many Drive users are paying customers. And the non-paying users are still customers, just customers getting the free loss leader in hopes of convincing them to buy.
The "cloud" is a joke. All it is is you storing YOUR files on someone ELSE'S computers... someone else who doesn't have one one-millionth of the vested interest in your files that you do, even if you DO opt for one of the pay cloud services.
Actually, I'd say that a cloud provider who makes billions on providing cloud services has a great vested interest in keeping your files safe and available. Oh, not so much about your files specifically, but the odds that just yours would get lost are negligible. If they were to lose data, it would hit lots of users... and that could easily cost them billions.
In particular, Google makes lots of money from enterprise users of Drive, Docs, etc. You can bet they're not going to jeopardize that. And, no, a few hours outage once per year or so isn't that big of a deal. That's less downtime than almost any self-hosted large-scale solution will achieve.
Either step up and take responsibility for your own shit, or stop whining.
I take responsibility for my own data, which is why I have a copy of it in Google Drive. Also, another copy on my desktop, one on my laptop and one on my wife's laptop. I used to run my own home file server, with RAID6 plus regular backups, and I dabbled for a while with automatic offset backups using Tahoe LAFS and a backup tool that I wrote. Then I wised up and started doing my offsite backups with a cloud provider and keeping a copy on each machine I use regularly. The result is safer, more reliable and much, much less work.
(Disclaimer/Disclosure: I work for Google. My relationship with Drive, however, is just that of a satisfied user.)
The article and the Reddit thread both talk about a "huge spike" in data usage without including any hard figures. What are we talking about here? 100 MB per day? A gigabyte?
Dunno. I've been running Oreo on and off for about six months now, though, and noticed no data usage while at home. So at least for my usage patterns it's clearly not large. Maybe some apps are worse than others? Or maybe it depends on your home Wifi. I could see that flaky Wifi combined with "mobile data always on" could cause the phone to fall back to using mobile data if the Wifi seems bad. It would do that with mobile data off, too, but it would take longer to make the switch, which might give the Wifi time to start working so you wouldn't end up using mobile data -- just have bad network performance.
With the current record of manufacturers and carriers not giving a damn about porting the Android updates to their products, I'm happy that google is developing a patch, but I'm wondering if anybody will actually receive the patch.
Since only Google devices have received Oreo yet, the patch will be out and in the source before other OEMs upgrade or ship devices.
All of your points are trumped by the simple fact that when people believe their network to be secure, they don't adequately secure the endpoints behind the firewall -- and that the network is never secure. And I don't mean that in a "perfectly hermetic" sense, I mean that in a practical "attackers can always get in" sense. With the exception of databases on laptops, this is the single largest root cause of leaked corporate data. The problem here is that you're talking about theory, and I'm talking about practice -- what actually happens. I've never seen a case of data theft or leakage where the post-mortem says "We needed better firewalls and a VPN". It's always "We needed better security on the database". 90% of the time, the root problem is that the database wasn't correctly configured, or wasn't patched.
The BeyondCorp approach wasn't developed in an ivory tower as a theoretical exercise, it was developed in response to a series of internal "Orange Team" attacks, where SECOPS guys team up with other engineers to try to penetrate the systems. What they found, time and time again, year after year, is that it's always possible to get into the network, that the protection provided by the firewalls is an illusion.
For example, in your "secure" network, do you physically disable all USB ports? Slipping a little USB dongle of some sort to an employee and getting them to plug it into their computer is a classic -- and trivial -- way of penetrating a network. One particularly humorous real-world example I saw was at a major bank, where the attacker managed to social engineer some information about employment anniversaries. He then took some off-the-shelf USB-powered plasma globes, put the bank's logo on them, stuck a microcontroller inside and shipped them to the employees on their anniversaries. Every single one of them plugged it in, and within minutes the attacker was roaming the network at will.
And of course your "secure" network uses 802.11X authentication on every network port, right? Another classic is to get into the building and plug a small device into an open network port.
The old standbys are the best, of course: malware delivered via email, in a PDF, etc.
Your firewall defeats script kiddies doing portscans, and it protects you from worms. That's it. It doesn't protect you from real attackers... and the defenses that will protect you from real attackers will also protect you from kiddies and worms.
And I'm sad to say, even Macs and Linux systems have had remote access exploits
Indeed they do, and that's why you really don't want to depend on a firewall. Your real defense against remote exploits is always on the machines themselves. Keep the running services to the minimal required set, keep the systems patched. Simple, but hard, tedious work. You're still vulnerable to 0days... but do you seriously think that an attacker in possession of good 0days is going to be stopped by a firewall? Not a chance. You're vulnerable to those now.
As I said previously, there's nothing wrong with firewalls and VPNs, if you also adequately secure everything behind the firewall. Firewalls provide a very minor, though real, defense in depth. My workstation is behind a stateful firewall, as required by corporate security policy for remote workstations. But the working assumption is that my workstation is sitting on the open Internet, and managed accordingly. For that matter, Google's corporate network is firewalled. But again, the assumption is that any attacker who wants to can get on the network.
Also: have you, personally, ever tried to activate and enforce a single-sign-on technology across a whole company? I have, many times in my career. It's also not cheap or a trivial task.
Yes, I've done several such projects, most with the added complication of smart cards. But while SSO has security benefits, its primary value is not security, so you're going to end up doing it anyway.
Sort of. That's part of why people have a VPN. The other part is to associate a specific user to a specific endpoint, not an IP, and to provide an encrypted sheath for interactions with company resources to make DNS leakages and TLS MITM attacks more difficult.
(Per my sig, I don't normally read or respond to ACs. I happened to see this one, though, and it's good so I'll answer.)
The specific endpoint in Google's model is at least as strong as that provided by a VPN. It's a per-device client-side digital certificate. On devices with a TPM, the private key is in the TPM, which attests the specific identity of the device. VPN solutions may or may not provide that level of endpoint validation.
Regarding the encrypted sheath, TLS provides it. Regarding TLS MITM attacks, yes, you do have to make sure that the proxies are kept ahead of the latest TLS weaknesses, and to keep them configured to simply reject any clients that try to downgrade. That's not too difficult, though, since you only have a small number of them to manage. Also, I think TLS has finally stabilized.
Regarding DNS leakage, that's a legitimate advantage of VPNs, but not a strong one for corporate use. Google doesn't actually bother to address it in most cases, but where it matters the solution is simple: put all services under one hostname, so the only information that leaks is that you're talking to the company. The fact that every request goes through a proxy makes this pretty easy, since the proxy can take responsibility for looking at the URL path prefix and routing the request to the appropriate backend. DNS leakage is a bigger issue for personal use, but there are lots of reasons to use VPNs for personal traffic (assuming you trust the VPN provider, because what you're not giving to the four different coffee shops, etc., you're concentrating for the VPN provider).
Does everything need to be spelled out for you? What is the point of Trustzone if it can be tampered with. Maybe you should go and do some reading on Trustzone technology and its purpose.
Note that this sequence of operations won't work on most phones launched with Marshmallow or later.
Step 2, factory reset, will clear a critical section of the replay-protected memory block (RPMB). That block stores the rollback protection status of Android Keymaster keys (Keymaster is a TrustZone -- or similar -- app that manages important cryptographic keys). Wiping it will make all such keys permanently unusable, cryptographically, and those keys are used to protect the device encryption keys.
So, when you get to step 7 and restore, you'll be restoring data that is encrypted with keys that you cannot recover.
If, however, you can tamper TrustZone in step 4 so that it, say, always generates the same, known, key for disk encryption, then give it to your target and wait for them to put sensitive data on it, then take it back, dump the flash and decrypt, then you can get the user's data. Oh, you'd also need to brute force the user's password, but that's not hard because phone passwords suck, and you could do it off-device.
Alternatively, if you could rewrite the RPMB data between step 6 and 7, you could "reactivate" the keys, but that would require finding a way to read it before step 2.
Slashdot Mirror
people are forced into a 2-year cellular contract
Forced? Really? I do not think that word means what you think it means.
This reminds me of a friend of mine who was looking for a house to buy and wanted to sign up for a month-to-month lease on their apartment. The landlord came back with a monthly cost 2.5x higher than the standard 12-month price.
Forced. Coerced. Convinced. Pick whatever word you want to describe little or no options left on the table that make financial sense.
Given that a high-end smartphone is a luxury, not a necessity, none of those words apply.
However, they can't actually afford a $6K per month mortgage payment
Then something is jacked up if they qualified for it.
Banks will always loan you more than you can actually afford.
Plus, you can always tell the bank you're planning to rent a room. That won't work if you're clearly in over your head, but if you're on the edge (per their very optimistic model), it will.
I also have doubts that Apple was completely unaware that their phones would, in fact, be sold under two year contracts when they agreed to sell them to the carriers.
I don't see how that's at all relevant to Apple. What financing or other arrangements are made between users and carriers is none of Apple's business. Apple communicated their actual warranty to both carriers and users and made no claims -- to anyone -- that they would warranty phones for the period of some contract between third parties.
I don't think your argument would hold water even if the users were buying the phones from Apple directly with two-year financing from Apple. When you buy a new car from Ford, there's no expectation that the car will continue functioning for the duration of the 72-month purchase contract. You *do* get various levels of warranty on various parts of the vehicle, but those are all completely independent of the financing terms.
Would you argue that someone who bought their car / phone for cash should get less warranty coverage than someone who financed it? This is silly. The one thing has nothing to do with the other.
people are forced into a 2-year cellular contract
Forced? Really? I do not think that word means what you think it means.
Either the buyer's an idiot, with a) more money than sense, or b) thinks they'll be rich any minute now; or else they're fucking flippers, who'll redo the kitchen, slap some paint on, and try to resell it.
Most likely it's neither. Instead, it's just someone buying a home that's close to work, for the going rate. Sellers in Silicon Valley don't even bother trying to price their houses accurately, they just ask the minimum they need to get, and know that the price will be bid up substantially. Odds are good that the buyer can't even afford the mortgage and will have to rent out part of the home to make ends meet (this is extremely common in the area).
But, the buyer also knows that unless something changes, buying the house is a good deal for them. They'll get every penny of equity back out when they sell in a few years, and probably get a hefty return on their investment, which is much better than if they'd just kept paying $4K+ per month in rent.
I know a few people who have done exactly this, though their homes have all been around $1.5M, not $2.5M. They buy a house and acquire a mortgage payment of just under $6K per month. $2.5K of that is principal, so their net cost for housing (ignoring the tax deduction, which they don't ignore) is $3.5K per month, which is about what they'd be paying in rent for an apartment large enough for a small family anyway.
However, they can't actually afford a $6K per month mortgage payment. So, they convert the garage into an apartment and rent it out, or rent out the master bedroom/bath. Some single guy pays them $2.5K per month to live in part of their house which lowers their monthly out of pocket cost to $3.5K per month... and lowers their net housing cost to $1K per month.
Bottom line, as long as they get their equity back out when they sell, this scheme is a much better deal than renting. If housing prices rise, they'll do even better and will probably never pay any taxes on the capital gains, which could be substantial. If the bottom drops out of housing prices, they could get severely screwed, of course.
You really need to read the underlying paper. The system actually has the properties you claim it cannot have.
If you are able to verify your vote after you leave the polling station, then someone else is potentially able to watch over your shoulder while you do so, and could therefore make good on any threat they'd provided you to ensure you voted the way they wanted you to.
Nope. It seems obvious that this should be impossible, but it's not. You should read about it.
If something is properly encrypted, disassembling a chip won't help.
I wish. (Really. Designing airtight cryptographic security for phones is my day job.)
The secrets used to encrypt the data must either be embedded in the device or obtained from outside of it, or some mixture of the two. Since the only practical outside source of key material is the user, and users suck at generating, managing and entering high-entropy secrets, the vast majority of the key material must come from the chip.
You can (and devices do) use key stretching, and you can (and devices do) implement various other brute force mitigations in software and hardware, but at the end of the day whatever software does can be replicated and hardware can be broken. So it really all comes down to the (lousy) user password.
Still, disassembling a chip is a pretty high barrier. And the right kind of chip can make it harder yet. At the end of the day there's nothing you can do in a consumer device that will defeat an adversary with serious engineering expertise and willing to spend a few hundred thousand dollars. Or one willing to spend a few hundred dollars and who is able to surreptitiously swap your device for one that they configured to look like yours, but to snarf your password.
I think telecommuting also makes a lot of sense, at least for the employees who find it appealing and can work effectively that way. I've telecommuted full time for about 15 of the last 20 years, and I think it's awesome. But many companies don't like telecommuting for various reasons that I don't fully understand, and I'm not sure they do either (Google doesn't; about one in 2000 Google engineers work remotely. No, that's not an exaggeration, if anything I'm overestimating the number of remote engineers).
Because they quickly realized that if you can do the job from 200 miles away then you can do it from 2000 miles away. All telecommuting did was make offshoring even more attractive.
That's a reason that companies should like telecommuting, not reject it.
Even IBM (before they went nuts and moved everyone to India) and other deep-pocketed companies had them back in the day, and that was when it was harder to stay in touch. The only difference was that the office was in Pittsburgh and not Pune, or Moline and not Mumbai. I remember reading something some time back that mentioned IBM would strategically locate big engineering facilities just far enough away from large business centers to be a short flight or medium length drive. They'd import the workers or hire from local university talent pools, and the execs would be mollified because they still felt like they had control. IBM used to have big facilities in Burlington, VT and Rochester, MN that fit that description perfectly. They probably didn't have to pay anything near what they'd have to pay for people in Westchester or Dutchess County, NY.
I worked for IBM back then, and that approach had its own share of problems. Fairly large ones.
The biggest was that those big engineering offices utterly dominated the local economy. Effectively, they created company towns, which meant that everyone who joined IBM had to move to one of the towns, and everyone who left IBM had to move out of one of the towns. This sucked for employees. The fact that there were several such locations meant that transferring to a different job within the company also frequently required relocation. When I first joined the company, employees said that IBM stood for "I've Been Moved".
Another was that a huge amount of IBM's capital was tied up in real estate. Rumor has it that when Louis Gerstner took over as CEO and surveyed the company's balance sheet, he said "Is this a computer company or a real estate company?". His fix to this problem greatly improved IBM's return on assets, but created its own problems. First, he sold most of the real estate and leased it back. That freed up a lot of capital and increased flexibility. Second, he started a program to push employees out of the office, making many of us telecommuters. Personally, I thought that was awesome, but not everyone likes to work that way. Employees began saying that IBM now meant "I'm By Myself", particularly since another Gerstner initiative was the push into the services business, which meant that employees were often working at customer sites, frequently on small contracts where they might be the only IBMer present for periods of time.
Of course, later Palmisano and then Rometty instituted the new IBM direction, which was to lay off the American employees and move everything overseas, to India, Romania, Brazil, etc.
I think satellite offices make a lot of sense, but I think it's better to locate them in regional tech hubs rather than out in the sticks the way IBM did. That makes it more likely that people can join or leave without having to relocate, giving both company and employees more flexibility. Google (my current employer) does this to some degree, maintaining engineering offices in many of the tech hubs. I think Google should do more of it, but the leadership seems to like at least half of the engineers to be located in Silicon Valley. It does allow employees to transition between teams easily, without having to relocate... but it means they have to live in Silicon Valley.
I think telecommuting also makes a lot of sense, at least for the employees who find it appealing and can work effectively that way. I've telecommuted full time for about 15 of the last 20 years, and I think it's awesome. But many companies don't like telecommuting for various reasons that I don't fully understand, and I'm not sure they do either (Google doesn't; about one in 2000 Google engineers work remotely. No, that's not an exaggeration, if anything I'm overestimating the number of remote engineers).
Just 2-3 weeks ago Jan T Sanner, the minister with responsibility for elections, decided that every single vote had to be counted manually, including all early voting ballots.
He should go one step further and implement end-to-end verifiable voting.
In my Virginia Beach precinct, we had electronic voting machines a while ago, but have had paper -- fill in the bubble, then scanned -- ballots for the past several years including the 2016 election. The ballots are scanned on their way into the locked ballot box. This system is easier and faster than the electronic versions were, plus there's a paper trail.
With a bit more it's possible to go a step further, and get election systems that not only have a verifiable paper trail, but which are end-to-end verifiable, allowing any voter to check after the fact whether or not their vote was included in the tally (but without being able to prove to anyone how they voted), and allowing anyone to verify the correctness of the tally. The method relies on applying the concepts and methods of modern cryptographic proofs to the problem of voting. It not only ensures that ballot scanning is not less secure than non-automated methods, it achieves a provable level of election integrity that has never before been possible.
Check it out.
As for intelligence - America has had a strong anti-intellectual bent from its inception "nurtured by the false notion that democracy means that "my ignorance is just as good as your knowledge." " -- Isaac Asimov
Some parts of America have long had the anti-intellectual bent, which they acquired largely in reaction to the pro-intellectual and zealous progressive proselytization of the Yankee northeast. From the 17th century the Yankee northeast revered education above wealth and birth and deeply trusted in collective, democratic community decisionmaking over aristocratic and other tradition, and worked to spread their gospel (originally religious gospel; later a sort of secular humanist gospel).
Meanwhile, the Virginia Tidewater region, settled by second sons of aristocrats interested in establishing and supporting their own hereditary aristocracy, also appreciated education, but of the classical, scholastic, let's-act-like-upper-class-Romans sort, rather than the progressive sort. The Deep South society, modeled after the slave plantations of Barbados, just wanted an ironclad aristocratic structure to keep the slaves from rising. They had little interest in, or opposition to, education for the upper class but actively disliked democracy unless it was tightly restricted. Both southern cultures disagreed with and disliked the Yankees, and resented their pushy progressivism (regarding abolition and much more).
And then there were the Applachian regions, with a culture derived from the first Scottish highlander settlers, violent and almost ridiculously independent. They also were deeply skeptical of the opinions of educated snobs and had no real interest in community much above the level of the family, or any sort of democracy, but just wanted to be left alone.
The other early American cultures, of New York (thoroughly commercial above all), the Midlands (pluralistic and middle class) and New France (egalitarian and tolerant) weren't so anti-intellectual, nor even so anti-Yankee, but all objected to the pushy self-assurance of the Yankee northeast.
Of course, history has mostly vindicated the progressive positions of the 18th-century Yankees, whose ideas were the core of the US Constitution. But their confidence in their own rightness and righteousness, that they knew what was better for others than those others knew themselves, is almost entirely responsible for the anti-intellectual element of American thought.
they know that they could go to a community college for 2 years (and we're in Canada where education doesn't cost as much as in the US), then do an apprenticeship for 2 years, and they'd be making double what they make now, and hardly anyone's doing it
I see what you're saying, but it's hardly proof that they wouldn't be willing to get the education they need if they had to. Right now, they'd like to make more money and have higher status in the organization, but not enough to put in the four years. Given more incentive -- perhaps with some funding to help them make the transition -- I think most of them would do it.
Of course, this presumes that there are any jobs for them to train into. That depends on the progress of AI research.
Current theories of economics are flawed, being based on assumptions of "infinite" that are no longer true.
Current theories of economics are in no way based on assumptions of "infinite" anything, they're very much based on the assumption that resources are scarce, have multiple uses, and that we must decide efficiently between those uses. Actually, it looks like what's going to invalidate current economic theories is that certain resources are going to become effectively "infinite". In particular, labor.
Currently, our economic system is structured around the notion that labor is scarce, and skilled labor is very scarce. That being the case, we need to incentivize people to work, and incentivize them even more to get education/training. So we have a system that more or less says "work or starve", because it is so important to make people work so that we can produce all the stuff that everyone needs to live.
Note that this structure continues right up to the top of the economic ladder, where it's focused not so much on producing goods as on optimizing the flow of resources needed to produce goods... including capital. Bankers and investors are rewarded by making sure that resources are routed away from unprofitable enterprises (those that don't use their resources wisely to generate value for people) and to profitable enterprises (those that do).
(Note that I'm not claiming the system is optimal, merely that it is closer to optimal than anything else we've tried at scale.)
What we're facing now is a radical change of conditions, in which it's possible that basically all unskilled labor, and a great deal of skilled labor -- possibly all of it! -- is going to be replaced by machines which are more efficient than humans in every way. Faster, stronger, tireless, more precise, perhaps even smarter, and, most important, "cheaper", meaning that they consume less resources to create and maintain.
Assuming that the robots actually work for us rather than deciding to eliminate us as wasteful and inefficient, this means that the cost of production is going to fall dramatically. Probably to nearly zero, as the only true constraints on production are raw materials and energy, which typically make up on a tiny, tiny fraction of the cost of most of our goods today.
That's fantastic, since it means that we (humanity, collectively) are going to be dramatically richer than we are today. But our current method of distributing wealth is based on the assumption that labor is scarce and thus needs deep, systematic and even coercive incentives. That assumption is going to become increasingly false, precisely because of post-scarcity "infinite" productivity.
Yeah, slashdot uses HSTS, so your browser should never sent HTTP requests to slashdot.
FWIW http://www.google.com/ does not use HSTS. For this exact reason, I'd guess. So your browser will send HTTP requests to google.com. If the request isn't intercepted you'll get back a 301 redirecting you to https://www.google.com/ so in the normal case it will immediately switch to TLS mode, but in the case that it's intercepted by a login page the login page will work.
I have no idea how long that will continue to be the case.
>Because it is a MitM attack?
Yes, I know it is. And if the browser would let me go to the damn page, I could get rid of it.
Go to an HTTP URL. Then the hotel proxy can intercept it without trouble, you can get logged in, then go to the page you actually wanted to visit.
This is a simple, excellent solution for people who understand what's going on. I have no idea how the other 99% of the population is supposed to use hotel Wifi. Hotels may have to find another approach when the vast majority of the web implements HSTS.
Actually, I think what will really happen is that browsers and/or OSes will do Wifi login page detection. Of the OSes that I often use, Android and Mac OS X already do. When a Wifi connection is established, they send a test HTTP request to a particular HTTP-only site. If the data they get is what's expected, great. If they get a Wifi login page, they prompt the user to log in.
Yes wanting to know what my bluray playing is up to when it calls home, makes me a bad person. Check
Your bluray player will accept the self-signed cert provided by your proxy as legitimate, instead of expecting the cert of its actual home server? That's possible, but it just means the bluray player software is broken, and if the manufacturer is really trying to do something sneaky, they'll fix it in the next update as soon as they realize someone is looking.
The real question here is: if Philo T. Farnsworth were alive today to see what's become of television, would he be happy or sad?
Amazed and impressed. 4K, 5K, 8K, vast color palettes, high-contrast, incredibly-thin screens... the technology is pretty amazing.
Google doesn't give a flying fuck about you and your files beyond their ability to mine them for data that they can monetize by selling your privacy away to the highest bidder.
Two errors in the above:
First, Google doesn't data mine Drive files, unless the file in question is marked publicly-accessible by the user.
Actually, I doubt they mine even public docs for information about the doc owner; my guess that the terms of service include that escape hatch for public docs because they get added to the search index, and so searches may turn them up with ads alongside which would constitute a "marketing or promotional campaign", in the words of the ToS.
Second, Google doesn't sell user data, even for products that are advertising-based. Google makes money by using the data to target ads, not by selling the data to advertisers (or anyone else).
Remember: if something is free, you're not the customer... you're the product.
Often true, but many Drive users are paying customers. And the non-paying users are still customers, just customers getting the free loss leader in hopes of convincing them to buy.
The "cloud" is a joke. All it is is you storing YOUR files on someone ELSE'S computers... someone else who doesn't have one one-millionth of the vested interest in your files that you do, even if you DO opt for one of the pay cloud services.
Actually, I'd say that a cloud provider who makes billions on providing cloud services has a great vested interest in keeping your files safe and available. Oh, not so much about your files specifically, but the odds that just yours would get lost are negligible. If they were to lose data, it would hit lots of users... and that could easily cost them billions.
In particular, Google makes lots of money from enterprise users of Drive, Docs, etc. You can bet they're not going to jeopardize that. And, no, a few hours outage once per year or so isn't that big of a deal. That's less downtime than almost any self-hosted large-scale solution will achieve.
Either step up and take responsibility for your own shit, or stop whining.
I take responsibility for my own data, which is why I have a copy of it in Google Drive. Also, another copy on my desktop, one on my laptop and one on my wife's laptop. I used to run my own home file server, with RAID6 plus regular backups, and I dabbled for a while with automatic offset backups using Tahoe LAFS and a backup tool that I wrote. Then I wised up and started doing my offsite backups with a cloud provider and keeping a copy on each machine I use regularly. The result is safer, more reliable and much, much less work.
(Disclaimer/Disclosure: I work for Google. My relationship with Drive, however, is just that of a satisfied user.)
The article and the Reddit thread both talk about a "huge spike" in data usage without including any hard figures. What are we talking about here? 100 MB per day? A gigabyte?
Dunno. I've been running Oreo on and off for about six months now, though, and noticed no data usage while at home. So at least for my usage patterns it's clearly not large. Maybe some apps are worse than others? Or maybe it depends on your home Wifi. I could see that flaky Wifi combined with "mobile data always on" could cause the phone to fall back to using mobile data if the Wifi seems bad. It would do that with mobile data off, too, but it would take longer to make the switch, which might give the Wifi time to start working so you wouldn't end up using mobile data -- just have bad network performance.
With the current record of manufacturers and carriers not giving a damn about porting the Android updates to their products, I'm happy that google is developing a patch, but I'm wondering if anybody will actually receive the patch.
Since only Google devices have received Oreo yet, the patch will be out and in the source before other OEMs upgrade or ship devices.
All of your points are trumped by the simple fact that when people believe their network to be secure, they don't adequately secure the endpoints behind the firewall -- and that the network is never secure. And I don't mean that in a "perfectly hermetic" sense, I mean that in a practical "attackers can always get in" sense. With the exception of databases on laptops, this is the single largest root cause of leaked corporate data. The problem here is that you're talking about theory, and I'm talking about practice -- what actually happens. I've never seen a case of data theft or leakage where the post-mortem says "We needed better firewalls and a VPN". It's always "We needed better security on the database". 90% of the time, the root problem is that the database wasn't correctly configured, or wasn't patched.
The BeyondCorp approach wasn't developed in an ivory tower as a theoretical exercise, it was developed in response to a series of internal "Orange Team" attacks, where SECOPS guys team up with other engineers to try to penetrate the systems. What they found, time and time again, year after year, is that it's always possible to get into the network, that the protection provided by the firewalls is an illusion.
For example, in your "secure" network, do you physically disable all USB ports? Slipping a little USB dongle of some sort to an employee and getting them to plug it into their computer is a classic -- and trivial -- way of penetrating a network. One particularly humorous real-world example I saw was at a major bank, where the attacker managed to social engineer some information about employment anniversaries. He then took some off-the-shelf USB-powered plasma globes, put the bank's logo on them, stuck a microcontroller inside and shipped them to the employees on their anniversaries. Every single one of them plugged it in, and within minutes the attacker was roaming the network at will.
And of course your "secure" network uses 802.11X authentication on every network port, right? Another classic is to get into the building and plug a small device into an open network port.
The old standbys are the best, of course: malware delivered via email, in a PDF, etc.
Your firewall defeats script kiddies doing portscans, and it protects you from worms. That's it. It doesn't protect you from real attackers... and the defenses that will protect you from real attackers will also protect you from kiddies and worms.
And I'm sad to say, even Macs and Linux systems have had remote access exploits
Indeed they do, and that's why you really don't want to depend on a firewall. Your real defense against remote exploits is always on the machines themselves. Keep the running services to the minimal required set, keep the systems patched. Simple, but hard, tedious work. You're still vulnerable to 0days... but do you seriously think that an attacker in possession of good 0days is going to be stopped by a firewall? Not a chance. You're vulnerable to those now.
As I said previously, there's nothing wrong with firewalls and VPNs, if you also adequately secure everything behind the firewall. Firewalls provide a very minor, though real, defense in depth. My workstation is behind a stateful firewall, as required by corporate security policy for remote workstations. But the working assumption is that my workstation is sitting on the open Internet, and managed accordingly. For that matter, Google's corporate network is firewalled. But again, the assumption is that any attacker who wants to can get on the network.
Also: have you, personally, ever tried to activate and enforce a single-sign-on technology across a whole company? I have, many times in my career. It's also not cheap or a trivial task.
Yes, I've done several such projects, most with the added complication of smart cards. But while SSO has security benefits, its primary value is not security, so you're going to end up doing it anyway.
Sort of. That's part of why people have a VPN. The other part is to associate a specific user to a specific endpoint, not an IP, and to provide an encrypted sheath for interactions with company resources to make DNS leakages and TLS MITM attacks more difficult.
(Per my sig, I don't normally read or respond to ACs. I happened to see this one, though, and it's good so I'll answer.)
The specific endpoint in Google's model is at least as strong as that provided by a VPN. It's a per-device client-side digital certificate. On devices with a TPM, the private key is in the TPM, which attests the specific identity of the device. VPN solutions may or may not provide that level of endpoint validation.
Regarding the encrypted sheath, TLS provides it. Regarding TLS MITM attacks, yes, you do have to make sure that the proxies are kept ahead of the latest TLS weaknesses, and to keep them configured to simply reject any clients that try to downgrade. That's not too difficult, though, since you only have a small number of them to manage. Also, I think TLS has finally stabilized.
Regarding DNS leakage, that's a legitimate advantage of VPNs, but not a strong one for corporate use. Google doesn't actually bother to address it in most cases, but where it matters the solution is simple: put all services under one hostname, so the only information that leaks is that you're talking to the company. The fact that every request goes through a proxy makes this pretty easy, since the proxy can take responsibility for looking at the URL path prefix and routing the request to the appropriate backend. DNS leakage is a bigger issue for personal use, but there are lots of reasons to use VPNs for personal traffic (assuming you trust the VPN provider, because what you're not giving to the four different coffee shops, etc., you're concentrating for the VPN provider).
Do you really not understand?
1. Backup phone
2. Factory reset
3. Unlock bootloader
4. Tamper Trustzone
5. Factory reset
6. Lock bootloader
7. Restore
Does everything need to be spelled out for you? What is the point of Trustzone if it can be tampered with. Maybe you should go and do some reading on Trustzone technology and its purpose.
Note that this sequence of operations won't work on most phones launched with Marshmallow or later.
Step 2, factory reset, will clear a critical section of the replay-protected memory block (RPMB). That block stores the rollback protection status of Android Keymaster keys (Keymaster is a TrustZone -- or similar -- app that manages important cryptographic keys). Wiping it will make all such keys permanently unusable, cryptographically, and those keys are used to protect the device encryption keys.
So, when you get to step 7 and restore, you'll be restoring data that is encrypted with keys that you cannot recover.
If, however, you can tamper TrustZone in step 4 so that it, say, always generates the same, known, key for disk encryption, then give it to your target and wait for them to put sensitive data on it, then take it back, dump the flash and decrypt, then you can get the user's data. Oh, you'd also need to brute force the user's password, but that's not hard because phone passwords suck, and you could do it off-device.
Alternatively, if you could rewrite the RPMB data between step 6 and 7, you could "reactivate" the keys, but that would require finding a way to read it before step 2.