Better, have the electorate vote to choose which issues matter most to them. Then, let them vote on each issue, and allow the political parties to apportion representatives proportional to the number of people who voted a particular way on those issues. For example, let's say there were three issues (way too small a number, but we'll use it for demonstration purposes): gun control, free speech, and abortion. 75% come out in favor of gun control, 50/50 on abortion, and 90% in favor of free speech. If you had ten legislators, the parties would have to choose at lest 7 who were in favor of gun control, 5 in favor of abortion, and 9 who are pro free speech. The Republicans would be unable to find any in favor of abortion, nor Democrats against, so in practice this would mean five Republicans, at least two of which must support gun control (more if any of the Democrats are against it) and at least four of which are strong free speech advocates (more if any Democrat isn't). In the event that Democrats are able to find anti-abortion candidates or Republicans are able to find pro-choice candidates, then candidates eligible for multiple positions should be chosen at random.
I can't help but marvel at the dichotomy here, too: when men, who are about 50% of the population, represent, say, 75% of the hires, it's evidence of rampant discrimination. When Asians, who represent about 5% of the population, are 90% of the hires, it's just evidence that "the best qualified rise to the top".
Both statements, of course, ridiculous, because they are using the general population as a baseline, rather than the percentage of qualified applicants. Unfortunately, when you look at race, there actually is still a pretty significant difference between the percentage of people graduating with technical degrees and the percentage who end up working in the field. The overrepresentation of Asians is not, of course, anywhere near 18x when looked at in the context of qualified applicants, because a lot more Asians get CS degrees to begin with, so the real number is probably more like 2x. Either way, though, when you compare industry hiring to graduation rate, Asians and whites are statistically overrepresented in the industry and blacks and Hispanics are statistically underrepresented, so there's definitely something odd happening.
That said, you can't necessarily assume that merely because the numbers don't match, discrimination must be at work. After all, there are many non-discrimination-based factors involved, such as cultural differences that motivate people to have different levels of interest in different things. (One could argue that not accommodating those differences is still discrimination, but really, there's a spectrum, and you can't force someone to take an interest in something that they don't care about.)
To determine whether discrimination is actually involved, as a society, we need to look into the motivation behind why people take CS classes and then don't go into CS careers. I think we need a lot more research in this field, because I'm convinced that we can't learn much of anything just by looking at the numbers themselves. The only way we'll really know what's going on is to ask questions that would reveal the degree to which institutional biases, cultural differences, and other factors influenced people's decisions to pursue or not pursue careers in the field. For each person, they should try to determine if he or she:
decided to take CS classes solely because his/her parents pressured him/her.
took CS classes out of enjoyment, but eventually decided that he/she didn't really want to spend his/her life coding.
took CS classes out of a desire to make more money, didn't well, and gave up.
decided that he or she didn't want to live in places with lots of tech jobs.
decided not to move away from a close-knit family (and/or a responsibility to take care of siblings) to work in a far-away city where tech jobs are more readily available.
looked at Bay Area housing prices and worried that he/she couldn't afford to move.
was unable to get sufficient resources from his/her college to help him/her find jobs in the field.
started working in CS and then felt out of place because he/she was the only [woman, black, Hispanic, insert other minority group here] in the workplace.
found his/her passion and decided to do something else. And ask what that passion was.
Ask people why they chose their majors. Ask people whether they've ever considered computer science. If not, encourage them to do so. If so, ask them why they decided not to pursue it. Ask people who went into CS and then dropped out (either before or after getting a degree) why they did. Ask them when they started learning computer science, and whether they had access to CS classes in K-12. Aggregate that data. Until somebody does this on a massive scale, we're all pretty much just grasping at straws. And we need to ask these questions of randomly selected people (both students and adults) every few years, and we need to ask enough people to get statistically relevant data. This, of course,
He actually states that the variations he discusses don't have a major effect, that the effect just causes that attaining the holy grail of a 50/50 split to not be quite possible to attain.
You don't need any magical discussion of human nature to prove that a 50/50 split between genders is impossible to attain, at least in the short to medium term. Fundamentally, it is not possible for an industry to hire more women than graduate with a degree in that field, ignoring the small percentage of self-taught programmers. On average, in the United States, women make up only about 16% of CS grads. So it is physically impossible for the industry average to be more even than 84/16 unless you deliberately leave a lot of men unemployed.
More to the point, the only way you can achieve a 50/50 split is to leave more than two-thirds of all computer science grads completely idle, and about 81% of all male CS grads unemployed. If you tried to implement this, two things would happen. First, the computer industry would collapse immediately, because it wouldn't be able to hire enough people to meet the immediate demand. Second, the computer industry would collapse even further long-term, because no sane person goes to college for four years known that they have a one in three chance of ever working in the field, and a two in three chance of waiting tables or flipping burgers for the rest of their lives.
The only way to improve on the gender imbalance is to improve on the number of women graduating with CS majors. That, in turn, has to start early in the education process—ideally as early as primary school. Gender imbalance can't be fixed by changing hiring practices and hoping that somehow 12-year-old girls will see how much companies want women programmers, and based on that, will magically take an interest in sitting inside behind a computer screen all day, learning to code. It is something that can only be fixed by getting more women to start learning CS, which mostly happens before kids are even old enough to know what "gender bias" means.
What this means for the world is that we need to shift our focus from trying to get more women into software companies, towards getting women into CS teaching jobs in middle schools and high schools, where studies show that girls are more likely to take an interest in learning CS from women than from men. And we need to focus on getting CS into the curriculum in the first place. (Ironically, Trump is right, but for entirely the wrong reasons.)
Don't get me wrong, I like working at a company that tries hard to recruit women, because the gender balance is healthier, but it isn't doing the industry as a whole any favors, and might even be making things worse, because the pool of applicants is largely a zero-sum game. When one company succeeds, it does so to the detriment of all the other companies. If all the large companies (Google, Apple, Facebook, etc.) managed to reach 50%, you'd have thousands of other companies with zero women programmers. And because most programmers will work for those other companies, most programmers would then perceive computer science to be an even more male-dominated field than they do now.
Just food for thought. I don't have all of the answers for how to fix the diversity problem. I just have the nagging feeling that we aren't even asking the right questions yet.
White skin tends to be slightly brighter than the average background, while black skin tends to be much darker.
... unless you're doing digital photography of people on a painted-black stage. Then, dark skin tends to have similar brightness to the background, and white skin tends to be a blown out pile of poo (unless you under-expose by at least a couple of stops). And, of course, the background ends up at 50% grey, so you can see every scuff mark on the floor. *sigh*
I'm sure the court ruling is a public document. If so, there's nothing to stop you from reading it and finding out for yourself, instead of complaining to a non-lawyer who lives nowhere near TN, much less Nashville.
Yes, the court ruling is a public document. Unfortunately, what you see in the summary is approximately the entire contents of the court ruling. The ruling literally says nothing more than that they overstepped their authority and are permanently enjoined from enforcing the law. It does give a vague hint about the reasons why they can't enforce it on poles that they don't own (saying that it conflicts with federal law, but providing no further details), but it provides no hint whatsoever about the reasoning that resulted in them being permanently enjoined from enforcing it on the poles that they own. The judge could have decided that part by flipping a coin, for all we know; in the absence of a more detailed decision, the decision seems entirely arbitrary, which is unfortunate given how many eyes are on this case nationwide.
But the city, operating in its capacity as the sole owner of a corporation, cannot possibly not have the right to operate that corporation in whatever way it sees fit. Such a conclusion is inherently nonsensical, regardless of what sovereign rights the Tennessee Constitution grants to cities.
What I don't understand is how it is ultra vires. The city of Nashville government OWNS Nashville Electric Service. The city owns the poles, and thus should have the legal right to make up any rules it wants regarding what people can do with poles that it owns. It's absolutely insane for the courts to hold otherwise.
Feinstein won't ever get kicked to the curb, because the Republicans can't seem to find anybody to run who isn't either racist, xenophobic, homophobic, or some combination of the above. All they would have to do is find one single fiscally conservative, socially moderate Republican to run, and she'd be gone, because I don't know any Democrats (at least in the Bay Area) who wouldn't vote against her in a heartbeat if the alternative weren't just to the right of Breitbart.
That's quite bizarre. The federal maximum duration for contempt is, by law, only 18 months, and the Pennsylvania statute allows for only a maximum of 90 days, so if he is being held longer than 90 days + 18 months, then he needs to hire better lawyers.
You obviously never bought Apple products while Jobs was alive. Apple made plenty of mistakes while he was there.
They made plenty of mistakes. I can list a lot of them. But with the possible exception of the lack of copy and paste on the original iPhone and its indented headphone jack, I can't really think of any that were as glaringly obvious to the casual observer as their recent mistakes.
also, Steve would never have removed the headphone jack, because he would have tried Bluetooth for maybe a week
AirPods would have been one of his favorite things, no question.
Maybe, but that's not all people use Bluetooth with. There's also cars, where Bluetooth barely works at all.
Which you can still do because Apple ships an adaptor for a wired connection with every phone. Not that I ever use it because in my car I attach through the USB connection that pretty much every car made in the last 6-7 years has.
My phone is plugged into my new (three weeks old) car through the USB connection. It still needs Bluetooth for audio. It only works without Bluetooth if your car pays Apple $$$ for CarPlay integration. I suspect SJ would be driving a Tesla these days, and... you guessed it. No CarPlay.
ESC key is still there by default, an app has to specifically override the touch bar to remove.
But when it is gone, it is gone. Also, you can't feel when it is gone, because it is a touchscreen, which makes it an accessibility nightmare for the visually disabled. At the very least, it should have been an option (on both models, not just the 13").
You are SERIOUSLY underestimating how long Apple has been working on FaceID and how much Jobs would have loved it. I cannot believe someone on Slashdot is claiming that Steve "no buttons" Jobs would have been against removing the last button and installing a 1000x improvement in biometrics.
How long they have worked on it? What does that have to do with whether it's a good feature or whether SJ would like it? Do you know how long they had been working on the Newton when SJ killed it? I mean someone could polish a turd for thirty years and it would still be a shiny turd. That's entirely the wrong metric. The fact that something is hard or takes a long time does not make it a good idea. If it were, then everyone would have beards down to their waists.
The problems with FaceID are fairly fundamental. First, by default, FaceID requires you to look directly at the phone to unlock it, and it inherently requires you to hold the phone in a spot where it can see your face. That's kind of clumsy in a lot of situations. Second, the lack of a home button is a big turnoff. It is quite challenging to swipe up from the bottom edge of an iPhone in a case (which most iPhones are), making the lack of a hardware home button a serious pain in the backside. It's hard enough trying to get into the control center with my device in a case (which I don't use very often for precisely that reason). I can't imagine having to do that every time I want to hit the home button or unlock the device.
Don't get me wrong, I think FaceID is a great idea in principle, and arguably even in implementation (even though the screen notch is ugly as h***, IMO), but that doesn't make up for the lack of a home button or TouchID. The device would be much, much better if it had both, and given a choice between the two, I would choose TouchID over FaceID every day and twice on Sunday, because it maps onto the way I use my device better. Your mileage may vary, of course.
Wrong, this is EXACTLY how you lead. You go places other people are afraid to tread, you make a few people angry, but you move everyone else forward.
Leading doesn't just mean doing things that other people are afraid to do. It also means making d**n sure you're doing the right thing. All of the things I listed above are situations where they did something that, IMO, is the polar opposite of the right thing. The difference between courage and arrogance is common sense.
The impossibility defense becomes practical at some point, as in, "I can't produce the decryption keys, because I have been in jail too long and don't remember them."
Besides, at least at the federal level, there's an 18-month maximum for contempt of court. (Some state laws allow for longer durations.)
The director is not stupid. He is, however, responsible for doing his job, and part of that job is to articulate the argument that encryption makes his job (and the job of his staff) more difficult and to indirectly provide cover when the next attack (and there will always be a next attack) succeeds due to lack of access to some data.
Of course, the thing is, they can have all the data in the world, and there's still no plausible way to sift through it. The flow of information exceeds what can be feasibly checked for terrorist intent by tens of orders of magnitude. It isn't just a little bit impossible. In a hundred years, we won't have computers that could sift through all the data we produce today. Thus, in the real world, breaking crypto can never prevent the next attack. All it can do is tell you more about the people who committed the last one.
That matters because of the difference between theory and practice:
In theory, if decrypting someone's data somehow could lead you to people who were going to commit the next attack, then arresting them could break the next attack.
In practice, you already know who someone was communicating with even without breaking the crypto on the actual messages, so the act of breaking the crypto can never lead you to the people who were going to commit the next attack.
At best, the only thing breakimg crypto can do is save you from having to investigate all the other people that the person was communicating with who weren't going to commit the next attack. And while that's useful from a cost-cutting point of view, a national security issue it ain't.
Even if you assume that they'll do their jobs perfectly, there would still the problem that any back door is essentially guaranteed to eventually be discovered by bad actors and used against the public at large. If the NSA gets their way, we won't be able to do banking online, because it won't be possible to secure the transactions. We won't be able to use credit cards at stores, because it won't be possible to secure the transactions. Basically, imagine a global information apocalypse, and then multiply by 1,000, and you're still not scared enough. Their proposal would be a ticking time bomb that at some arbitrary point in the future would quite literally bring about the end of modern civilization as we know it.
And it would only affect the good guys—the people who have nothing to hide. The bad guys—the people who are actually trying to hide things from law enforcement—would still use unbreakable encryption. After all, the punishment for breaking a crypto law can't practically exceed the punishment they would get if they handed over proof of two decades of drug smuggling, contract murders, etc. Better to go to jail on that minor charge for a year or two than for the rest of your life. So there's absolutely no incentive for the bad guys to follow the law, which means they won't.
This isn't even one of those situations where you can justify it by secondary effects. Folks scream about gun control even though reducing the number of weapons in the hands of the good guys does reduce the number of weapons in the hands of bad guys by reducing the number of weapons out there in the world that can easily be stolen, de-serialed, and sold on the black market. This doesn't even have that advantage, because you don't have to steal crypto software. It costs nothing to make a copy of a piece of software (assuming it isn't commercial software), so the bad guys won't have any trouble getting real crypto even if they take away everyone else's access.
And even if somehow they could magically fix all of those problems with a crypto system based on rainbows and unicorn farts, breaking everyone's crypto still wouldn't buy them much. At best, in the hypothetical situation where someone committed a terrorist attack, they might be able to determine whether the people that person contacted were terrorists or not, instead of having to investigate all of them. So it would save a relatively small amount of investigative effort. And in exchange for that tiny savings by our government, they want us all to give up every shred of privacy—every shred of information security—and send us hurtling headlong towards the end of the world as we know it.
No, what they are proposing is approximately the single most stupid thing ever to come out of any branch of government. This tops the ban on carrying soft drinks through airport security. This tops the ban on pocketknives. This tops the California cities that limit the number of electric vehicle parking places at businesses in the hopes that somehow it will magically reduce road congestion by making people drive their gas guzzlers. It is completely unjustifiable through any logic, no matter how far you try to stretch it—completely and utterly bonkers. Sad.
The "disinhibition effect" is driven by anonymity. People are "mean" to strangers online. Teenagers spend most of their time socializing with close friends.
No, read that article again. Anonymity is just one of several factors that contributes to the effect, and it occurs even without anonymity. It turns out that actually seeing the look on someone's face when you hurt his/her feelings results in a lot more empathy than a text message sent ten minutes later, and not seeing that person in the flesh until the next day.
Asserting that something is "self-evident" is very different from providing actual evidence. I have seen no evidence of causative harm from teenagers socializing online, rather than say, watching TV.
You aren't paying attention, then. Studies have shown that teenagers today are significantly more narcissistic, on average than twenty or thirty years ago. This may or may not be visible by looking at specific individuals in isolation, but in aggregate, the effect is very real and well documented. And that's precisely the effect that one would expect from a loss of empathy, which is precisely the effect that one would predict from people socializing online too much and in person too little. I mean, this isn't absolute proof, but it is about as close as you can get without a randomly selected control group.
There's actually a lot of objective evidence that people are a lot more mean online than they are in person. There's even a name for this difference—the online disinhibition effect. It should be self-evident, then, that doing most of your socializing online will lead to people not being as nice, and in aggregate, will cause significant societal harm.
This is not to say that parents need to micromanage their kids, but there definitely comes a point at which parents do need to actually parent, by telling their kids to put down the phone and actually talk to other people. That said, I'm not sure how Apple could address that—Facebook, Twitter, Snapchat, etc., sure, but not Apple. The problem isn't the hardware, and thus can't realistically be solved by the hardware, I don't think.
Funny. There was an NVIDIA driver security fix on Linux just a few days ago. So nice FUD, but you really need to be a lot more paranoid. GPU drivers are updated for security vulnerabilities often enough for it to have been the first thing I thought of when I saw this story, which is to say that those updates occur way more often than they reasonably should.
Defense in depth requires that you address holes at any level, regardless of whether they can be exploited by someone without breaking some other security. When applied to meatspace, you're basically arguing that there's no need to fix a stuck-open bank vault door, because the exterior doors of the bank have locks and alarms.
Now it may not always take to a great depth but the basic fact is that everyone technical working at Apple is looking at product design using very similar criteria to what Jobs would be using. Apple has become a kind of like a large somewhat lossy AI simulating Jobs mind, using employees as the processors.
It obviously doesn't take. If it did, they wouldn't make so many utterly stupid and basic mistakes in their design. For example, the current Apple TV has a USB-C port and comes with a USB (non-C) to lightning cable to charge its remote control. Worse, even if you buy an adapter, the USB-C port on the device still won't provide power for charging the remote control. They literally provide a device that looks like it should be capable of being used standalone, but actually can't be used unless you either have a computer or a cell phone charger to charge up the remote control. (Well, it will work for a few weeks until the remote control's battery runs down. And I guess some TVs have a USB debug port, though not all.)
I mean, that's the sort of obvious, trivial oversight that would have caused Steve to throw the device across the room and say, "Come back when you have a product that people can use."
Also, Steve would never have removed the headphone jack, because he would have tried Bluetooth for maybe a week before throwing the phone at an engineer and saying, "This doesn't f**king work." My iPhone 6s won't reliably connect to either of my cars (different brands, different years). It just randomly fails, often requiring power-cycling both devices to get it going again. It takes one second to plug in a wire. When Bluetooth doesn't work, it takes three to five minutes to get it working again, and because it fails about once per week, on average, it is at least an order of magnitude slower than just using a wired connection. Given the poor reliability, I doubt Bluetooth will be Steve-ready for at least another five to ten years. It just sucks horribly.
And I'm pretty sure Steve would never have shipped the touchbar with an escape key (which is critical for proper keyboard navigation of dialog boxes in OS X, not to mention vi), nor would he have shipped the iPhone X without a fingerprint scanner. As somebody who has used a fingerprint scanner on the back, it works just fine. Their irrational fear of doing what the rest of the industry is doing resulted in a product that is considerably less usable than the previous generation. This is not how you lead.
I could continue ranting for hours like this. It isn't that Apple doesn't produce good products. It's just that the polish is no longer there. They don't spend the extra time to fix all the things that would have caused Steve to tell them that the product isn't ready. Their engineering teams are simultaneously too detail-oriented and not detail-oriented enough, missing the forest for the trees. And the only way to fix that is to bring in somebody with the same sort of eye for what details matter to real users that Steve had—to bring in a gatekeeper and champion of user-friendliness.
Every employee always has a smartphone on them. Always.
People do sometimes forget their phones at home, and not everybody has a phone that can access internal resources, so that's not necessarily true. However, every employee does have a badge (or else they couldn't get into their own building). It would not be that hard to have a battery-powered badge reader that, when you tap your badge on it, unlocks a locking hub, and re-locks it when you tap the same badge a second time. That approach has several benefits:
It's a lot faster to tap a badge than to get out your smartphone and screw around with it for three minutes while you're trying to grab a bike.
If a bike gets stolen and recovered, you know exactly which employee forgot to re-lock the bike, and you can slap that person around a bit (unlike with a fixed code).
The cellular hardware required for validating the badge would also give you the ability to track the bike.
Of course, that device would have to get charged every day, but I think they pick up the bikes every night and redistribute them every morning anyway, so charging them up wouldn't be such a big deal. Alternatively, they could have badge readers on the bike racks, and require you to return the bikes to those racks. That would have all the same advantages without the need to charge the bikes, and would also have the advantage of guaranteeing that you can find the bikes in a consistent spot, rather than them being in any of a dozen different places around any given building. Of course, it would be a big disadvantage if any of those racks ever became full.:-)
Which probably contains security holes that make it unfit for use in a data center. Sticking with an old version of GPU drivers is simply not a viable option, and anybody even suggesting otherwise should be stripped naked and dragged through the streets behind a Brinks truck on national television for all to see. It is the computer security equivalent of saying, "It doesn't matter if the fuel tank in that Ford Pinto is so thin and right in front of the rear bumper."
If this is legal (I'm pretty sure it isn't), then it's way past time for some serious changes to copyright law and contract law. No sane society can afford to allow a company to make arbitrary changes the license agreement on critical device drivers that are required for hardware to function properly and that must be kept up-to-date to keep a system secure. After all, if they can change these terms of sale retroactively, what's to stop them from deciding three years from now that the Tesla V100 drivers are no longer licensed for data center use, and you're required to upgrade to the Tesla V600 if you want to keep using it in a data center? One year from now? Six months?
Even if NVIDIA manages to find a way to avoid losing every lawsuit that arises from this suicidally stupid decision, I have to wonder why in h*** any data center purchaser in his/her right should mind even CONSIDER NVIDIA hardware in the future, knowing that NVIDIA might arbitrarily change their licensing terms in a way that forces them to sell all their hardware at a loss and replace it at any time?
This really should bankrupt NVIDIA in a just world. It's that heinous. And IMO, someone should be fired for even suggesting such an appalling change to their hardware licensing retroactively.
This guy repeatedly, intentionally, with malice aforethought, put people's lives in danger. He may not have pulled the trigger, but IMO, he should still be found guilty of felony murder (in the first degree) and tried accordingly. Watching him get a lethal injection sentence *might* be enough to deter others who still think it's fun....
Yeah, this is what I don't get. Anyone who would actually have something to hide would not carry it unencrypted across the border, because they would know that the border security people might decide to search it. So apart from catching the most incredibly stupid criminals (who would probably get caught for other reasons even without this search), the only thing this rather bizarre policy will do is cause Americans to become lackadaisical about our fourth amendment rights. Then again, maybe that's the point.
Slashdot Mirror
Better, have the electorate vote to choose which issues matter most to them. Then, let them vote on each issue, and allow the political parties to apportion representatives proportional to the number of people who voted a particular way on those issues. For example, let's say there were three issues (way too small a number, but we'll use it for demonstration purposes): gun control, free speech, and abortion. 75% come out in favor of gun control, 50/50 on abortion, and 90% in favor of free speech. If you had ten legislators, the parties would have to choose at lest 7 who were in favor of gun control, 5 in favor of abortion, and 9 who are pro free speech. The Republicans would be unable to find any in favor of abortion, nor Democrats against, so in practice this would mean five Republicans, at least two of which must support gun control (more if any of the Democrats are against it) and at least four of which are strong free speech advocates (more if any Democrat isn't). In the event that Democrats are able to find anti-abortion candidates or Republicans are able to find pro-choice candidates, then candidates eligible for multiple positions should be chosen at random.
I can't help but marvel at the dichotomy here, too: when men, who are about 50% of the population, represent, say, 75% of the hires, it's evidence of rampant discrimination. When Asians, who represent about 5% of the population, are 90% of the hires, it's just evidence that "the best qualified rise to the top".
Both statements, of course, ridiculous, because they are using the general population as a baseline, rather than the percentage of qualified applicants. Unfortunately, when you look at race, there actually is still a pretty significant difference between the percentage of people graduating with technical degrees and the percentage who end up working in the field. The overrepresentation of Asians is not, of course, anywhere near 18x when looked at in the context of qualified applicants, because a lot more Asians get CS degrees to begin with, so the real number is probably more like 2x. Either way, though, when you compare industry hiring to graduation rate, Asians and whites are statistically overrepresented in the industry and blacks and Hispanics are statistically underrepresented, so there's definitely something odd happening.
That said, you can't necessarily assume that merely because the numbers don't match, discrimination must be at work. After all, there are many non-discrimination-based factors involved, such as cultural differences that motivate people to have different levels of interest in different things. (One could argue that not accommodating those differences is still discrimination, but really, there's a spectrum, and you can't force someone to take an interest in something that they don't care about.)
To determine whether discrimination is actually involved, as a society, we need to look into the motivation behind why people take CS classes and then don't go into CS careers. I think we need a lot more research in this field, because I'm convinced that we can't learn much of anything just by looking at the numbers themselves. The only way we'll really know what's going on is to ask questions that would reveal the degree to which institutional biases, cultural differences, and other factors influenced people's decisions to pursue or not pursue careers in the field. For each person, they should try to determine if he or she:
Ask people why they chose their majors. Ask people whether they've ever considered computer science. If not, encourage them to do so. If so, ask them why they decided not to pursue it. Ask people who went into CS and then dropped out (either before or after getting a degree) why they did. Ask them when they started learning computer science, and whether they had access to CS classes in K-12. Aggregate that data. Until somebody does this on a massive scale, we're all pretty much just grasping at straws. And we need to ask these questions of randomly selected people (both students and adults) every few years, and we need to ask enough people to get statistically relevant data. This, of course,
You'd be surprised how often event videographers shoot footage of concerts, dance programs, etc. on a black stage. It seems pretty average to me.
You don't need any magical discussion of human nature to prove that a 50/50 split between genders is impossible to attain, at least in the short to medium term. Fundamentally, it is not possible for an industry to hire more women than graduate with a degree in that field, ignoring the small percentage of self-taught programmers. On average, in the United States, women make up only about 16% of CS grads. So it is physically impossible for the industry average to be more even than 84/16 unless you deliberately leave a lot of men unemployed.
More to the point, the only way you can achieve a 50/50 split is to leave more than two-thirds of all computer science grads completely idle, and about 81% of all male CS grads unemployed. If you tried to implement this, two things would happen. First, the computer industry would collapse immediately, because it wouldn't be able to hire enough people to meet the immediate demand. Second, the computer industry would collapse even further long-term, because no sane person goes to college for four years known that they have a one in three chance of ever working in the field, and a two in three chance of waiting tables or flipping burgers for the rest of their lives.
The only way to improve on the gender imbalance is to improve on the number of women graduating with CS majors. That, in turn, has to start early in the education process—ideally as early as primary school. Gender imbalance can't be fixed by changing hiring practices and hoping that somehow 12-year-old girls will see how much companies want women programmers, and based on that, will magically take an interest in sitting inside behind a computer screen all day, learning to code. It is something that can only be fixed by getting more women to start learning CS, which mostly happens before kids are even old enough to know what "gender bias" means.
What this means for the world is that we need to shift our focus from trying to get more women into software companies, towards getting women into CS teaching jobs in middle schools and high schools, where studies show that girls are more likely to take an interest in learning CS from women than from men. And we need to focus on getting CS into the curriculum in the first place. (Ironically, Trump is right, but for entirely the wrong reasons.)
Don't get me wrong, I like working at a company that tries hard to recruit women, because the gender balance is healthier, but it isn't doing the industry as a whole any favors, and might even be making things worse, because the pool of applicants is largely a zero-sum game. When one company succeeds, it does so to the detriment of all the other companies. If all the large companies (Google, Apple, Facebook, etc.) managed to reach 50%, you'd have thousands of other companies with zero women programmers. And because most programmers will work for those other companies, most programmers would then perceive computer science to be an even more male-dominated field than they do now.
Just food for thought. I don't have all of the answers for how to fix the diversity problem. I just have the nagging feeling that we aren't even asking the right questions yet.
... unless you're doing digital photography of people on a painted-black stage. Then, dark skin tends to have similar brightness to the background, and white skin tends to be a blown out pile of poo (unless you under-expose by at least a couple of stops). And, of course, the background ends up at 50% grey, so you can see every scuff mark on the floor. *sigh*
So... you're saying this story is a bunch of hot air?
I'm sure the court ruling is a public document. If so, there's nothing to stop you from reading it and finding out for yourself, instead of complaining to a non-lawyer who lives nowhere near TN, much less Nashville.
Yes, the court ruling is a public document. Unfortunately, what you see in the summary is approximately the entire contents of the court ruling. The ruling literally says nothing more than that they overstepped their authority and are permanently enjoined from enforcing the law. It does give a vague hint about the reasons why they can't enforce it on poles that they don't own (saying that it conflicts with federal law, but providing no further details), but it provides no hint whatsoever about the reasoning that resulted in them being permanently enjoined from enforcing it on the poles that they own. The judge could have decided that part by flipping a coin, for all we know; in the absence of a more detailed decision, the decision seems entirely arbitrary, which is unfortunate given how many eyes are on this case nationwide.
But the city, operating in its capacity as the sole owner of a corporation, cannot possibly not have the right to operate that corporation in whatever way it sees fit. Such a conclusion is inherently nonsensical, regardless of what sovereign rights the Tennessee Constitution grants to cities.
What I don't understand is how it is ultra vires. The city of Nashville government OWNS Nashville Electric Service. The city owns the poles, and thus should have the legal right to make up any rules it wants regarding what people can do with poles that it owns. It's absolutely insane for the courts to hold otherwise.
Feinstein won't ever get kicked to the curb, because the Republicans can't seem to find anybody to run who isn't either racist, xenophobic, homophobic, or some combination of the above. All they would have to do is find one single fiscally conservative, socially moderate Republican to run, and she'd be gone, because I don't know any Democrats (at least in the Bay Area) who wouldn't vote against her in a heartbeat if the alternative weren't just to the right of Breitbart.
That's quite bizarre. The federal maximum duration for contempt is, by law, only 18 months, and the Pennsylvania statute allows for only a maximum of 90 days, so if he is being held longer than 90 days + 18 months, then he needs to hire better lawyers.
They made plenty of mistakes. I can list a lot of them. But with the possible exception of the lack of copy and paste on the original iPhone and its indented headphone jack, I can't really think of any that were as glaringly obvious to the casual observer as their recent mistakes.
Maybe, but that's not all people use Bluetooth with. There's also cars, where Bluetooth barely works at all.
My phone is plugged into my new (three weeks old) car through the USB connection. It still needs Bluetooth for audio. It only works without Bluetooth if your car pays Apple $$$ for CarPlay integration. I suspect SJ would be driving a Tesla these days, and... you guessed it. No CarPlay.
But when it is gone, it is gone. Also, you can't feel when it is gone, because it is a touchscreen, which makes it an accessibility nightmare for the visually disabled. At the very least, it should have been an option (on both models, not just the 13").
How long they have worked on it? What does that have to do with whether it's a good feature or whether SJ would like it? Do you know how long they had been working on the Newton when SJ killed it? I mean someone could polish a turd for thirty years and it would still be a shiny turd. That's entirely the wrong metric. The fact that something is hard or takes a long time does not make it a good idea. If it were, then everyone would have beards down to their waists.
The problems with FaceID are fairly fundamental. First, by default, FaceID requires you to look directly at the phone to unlock it, and it inherently requires you to hold the phone in a spot where it can see your face. That's kind of clumsy in a lot of situations. Second, the lack of a home button is a big turnoff. It is quite challenging to swipe up from the bottom edge of an iPhone in a case (which most iPhones are), making the lack of a hardware home button a serious pain in the backside. It's hard enough trying to get into the control center with my device in a case (which I don't use very often for precisely that reason). I can't imagine having to do that every time I want to hit the home button or unlock the device.
Don't get me wrong, I think FaceID is a great idea in principle, and arguably even in implementation (even though the screen notch is ugly as h***, IMO), but that doesn't make up for the lack of a home button or TouchID. The device would be much, much better if it had both, and given a choice between the two, I would choose TouchID over FaceID every day and twice on Sunday, because it maps onto the way I use my device better. Your mileage may vary, of course.
Leading doesn't just mean doing things that other people are afraid to do. It also means making d**n sure you're doing the right thing. All of the things I listed above are situations where they did something that, IMO, is the polar opposite of the right thing. The difference between courage and arrogance is common sense.
The impossibility defense becomes practical at some point, as in, "I can't produce the decryption keys, because I have been in jail too long and don't remember them."
Besides, at least at the federal level, there's an 18-month maximum for contempt of court. (Some state laws allow for longer durations.)
Of course, the thing is, they can have all the data in the world, and there's still no plausible way to sift through it. The flow of information exceeds what can be feasibly checked for terrorist intent by tens of orders of magnitude. It isn't just a little bit impossible. In a hundred years, we won't have computers that could sift through all the data we produce today. Thus, in the real world, breaking crypto can never prevent the next attack. All it can do is tell you more about the people who committed the last one.
That matters because of the difference between theory and practice:
At best, the only thing breakimg crypto can do is save you from having to investigate all the other people that the person was communicating with who weren't going to commit the next attack. And while that's useful from a cost-cutting point of view, a national security issue it ain't.
Even if you assume that they'll do their jobs perfectly, there would still the problem that any back door is essentially guaranteed to eventually be discovered by bad actors and used against the public at large. If the NSA gets their way, we won't be able to do banking online, because it won't be possible to secure the transactions. We won't be able to use credit cards at stores, because it won't be possible to secure the transactions. Basically, imagine a global information apocalypse, and then multiply by 1,000, and you're still not scared enough. Their proposal would be a ticking time bomb that at some arbitrary point in the future would quite literally bring about the end of modern civilization as we know it.
And it would only affect the good guys—the people who have nothing to hide. The bad guys—the people who are actually trying to hide things from law enforcement—would still use unbreakable encryption. After all, the punishment for breaking a crypto law can't practically exceed the punishment they would get if they handed over proof of two decades of drug smuggling, contract murders, etc. Better to go to jail on that minor charge for a year or two than for the rest of your life. So there's absolutely no incentive for the bad guys to follow the law, which means they won't.
This isn't even one of those situations where you can justify it by secondary effects. Folks scream about gun control even though reducing the number of weapons in the hands of the good guys does reduce the number of weapons in the hands of bad guys by reducing the number of weapons out there in the world that can easily be stolen, de-serialed, and sold on the black market. This doesn't even have that advantage, because you don't have to steal crypto software. It costs nothing to make a copy of a piece of software (assuming it isn't commercial software), so the bad guys won't have any trouble getting real crypto even if they take away everyone else's access.
And even if somehow they could magically fix all of those problems with a crypto system based on rainbows and unicorn farts, breaking everyone's crypto still wouldn't buy them much. At best, in the hypothetical situation where someone committed a terrorist attack, they might be able to determine whether the people that person contacted were terrorists or not, instead of having to investigate all of them. So it would save a relatively small amount of investigative effort. And in exchange for that tiny savings by our government, they want us all to give up every shred of privacy—every shred of information security—and send us hurtling headlong towards the end of the world as we know it.
No, what they are proposing is approximately the single most stupid thing ever to come out of any branch of government. This tops the ban on carrying soft drinks through airport security. This tops the ban on pocketknives. This tops the California cities that limit the number of electric vehicle parking places at businesses in the hopes that somehow it will magically reduce road congestion by making people drive their gas guzzlers. It is completely unjustifiable through any logic, no matter how far you try to stretch it—completely and utterly bonkers. Sad.
Their idea is bad, and they should feel bad.
No, read that article again. Anonymity is just one of several factors that contributes to the effect, and it occurs even without anonymity. It turns out that actually seeing the look on someone's face when you hurt his/her feelings results in a lot more empathy than a text message sent ten minutes later, and not seeing that person in the flesh until the next day.
You aren't paying attention, then. Studies have shown that teenagers today are significantly more narcissistic, on average than twenty or thirty years ago. This may or may not be visible by looking at specific individuals in isolation, but in aggregate, the effect is very real and well documented. And that's precisely the effect that one would expect from a loss of empathy, which is precisely the effect that one would predict from people socializing online too much and in person too little. I mean, this isn't absolute proof, but it is about as close as you can get without a randomly selected control group.
There's actually a lot of objective evidence that people are a lot more mean online than they are in person. There's even a name for this difference—the online disinhibition effect. It should be self-evident, then, that doing most of your socializing online will lead to people not being as nice, and in aggregate, will cause significant societal harm.
This is not to say that parents need to micromanage their kids, but there definitely comes a point at which parents do need to actually parent, by telling their kids to put down the phone and actually talk to other people. That said, I'm not sure how Apple could address that—Facebook, Twitter, Snapchat, etc., sure, but not Apple. The problem isn't the hardware, and thus can't realistically be solved by the hardware, I don't think.
Funny. There was an NVIDIA driver security fix on Linux just a few days ago. So nice FUD, but you really need to be a lot more paranoid. GPU drivers are updated for security vulnerabilities often enough for it to have been the first thing I thought of when I saw this story, which is to say that those updates occur way more often than they reasonably should.
Defense in depth requires that you address holes at any level, regardless of whether they can be exploited by someone without breaking some other security. When applied to meatspace, you're basically arguing that there's no need to fix a stuck-open bank vault door, because the exterior doors of the bank have locks and alarms.
It obviously doesn't take. If it did, they wouldn't make so many utterly stupid and basic mistakes in their design. For example, the current Apple TV has a USB-C port and comes with a USB (non-C) to lightning cable to charge its remote control. Worse, even if you buy an adapter, the USB-C port on the device still won't provide power for charging the remote control. They literally provide a device that looks like it should be capable of being used standalone, but actually can't be used unless you either have a computer or a cell phone charger to charge up the remote control. (Well, it will work for a few weeks until the remote control's battery runs down. And I guess some TVs have a USB debug port, though not all.)
I mean, that's the sort of obvious, trivial oversight that would have caused Steve to throw the device across the room and say, "Come back when you have a product that people can use."
Also, Steve would never have removed the headphone jack, because he would have tried Bluetooth for maybe a week before throwing the phone at an engineer and saying, "This doesn't f**king work." My iPhone 6s won't reliably connect to either of my cars (different brands, different years). It just randomly fails, often requiring power-cycling both devices to get it going again. It takes one second to plug in a wire. When Bluetooth doesn't work, it takes three to five minutes to get it working again, and because it fails about once per week, on average, it is at least an order of magnitude slower than just using a wired connection. Given the poor reliability, I doubt Bluetooth will be Steve-ready for at least another five to ten years. It just sucks horribly.
And I'm pretty sure Steve would never have shipped the touchbar with an escape key (which is critical for proper keyboard navigation of dialog boxes in OS X, not to mention vi), nor would he have shipped the iPhone X without a fingerprint scanner. As somebody who has used a fingerprint scanner on the back, it works just fine. Their irrational fear of doing what the rest of the industry is doing resulted in a product that is considerably less usable than the previous generation. This is not how you lead.
I could continue ranting for hours like this. It isn't that Apple doesn't produce good products. It's just that the polish is no longer there. They don't spend the extra time to fix all the things that would have caused Steve to tell them that the product isn't ready. Their engineering teams are simultaneously too detail-oriented and not detail-oriented enough, missing the forest for the trees. And the only way to fix that is to bring in somebody with the same sort of eye for what details matter to real users that Steve had—to bring in a gatekeeper and champion of user-friendliness.
People do sometimes forget their phones at home, and not everybody has a phone that can access internal resources, so that's not necessarily true. However, every employee does have a badge (or else they couldn't get into their own building). It would not be that hard to have a battery-powered badge reader that, when you tap your badge on it, unlocks a locking hub, and re-locks it when you tap the same badge a second time. That approach has several benefits:
Of course, that device would have to get charged every day, but I think they pick up the bikes every night and redistribute them every morning anyway, so charging them up wouldn't be such a big deal. Alternatively, they could have badge readers on the bike racks, and require you to return the bikes to those racks. That would have all the same advantages without the need to charge the bikes, and would also have the advantage of guaranteeing that you can find the bikes in a consistent spot, rather than them being in any of a dozen different places around any given building. Of course, it would be a big disadvantage if any of those racks ever became full. :-)
Filing a false police report, depending on the seriousness of the report, can be considered a felony, so yes.
Which probably contains security holes that make it unfit for use in a data center. Sticking with an old version of GPU drivers is simply not a viable option, and anybody even suggesting otherwise should be stripped naked and dragged through the streets behind a Brinks truck on national television for all to see. It is the computer security equivalent of saying, "It doesn't matter if the fuel tank in that Ford Pinto is so thin and right in front of the rear bumper."
If this is legal (I'm pretty sure it isn't), then it's way past time for some serious changes to copyright law and contract law. No sane society can afford to allow a company to make arbitrary changes the license agreement on critical device drivers that are required for hardware to function properly and that must be kept up-to-date to keep a system secure. After all, if they can change these terms of sale retroactively, what's to stop them from deciding three years from now that the Tesla V100 drivers are no longer licensed for data center use, and you're required to upgrade to the Tesla V600 if you want to keep using it in a data center? One year from now? Six months?
Even if NVIDIA manages to find a way to avoid losing every lawsuit that arises from this suicidally stupid decision, I have to wonder why in h*** any data center purchaser in his/her right should mind even CONSIDER NVIDIA hardware in the future, knowing that NVIDIA might arbitrarily change their licensing terms in a way that forces them to sell all their hardware at a loss and replace it at any time?
This really should bankrupt NVIDIA in a just world. It's that heinous. And IMO, someone should be fired for even suggesting such an appalling change to their hardware licensing retroactively.
This guy repeatedly, intentionally, with malice aforethought, put people's lives in danger. He may not have pulled the trigger, but IMO, he should still be found guilty of felony murder (in the first degree) and tried accordingly. Watching him get a lethal injection sentence *might* be enough to deter others who still think it's fun....
Yeah, this is what I don't get. Anyone who would actually have something to hide would not carry it unencrypted across the border, because they would know that the border security people might decide to search it. So apart from catching the most incredibly stupid criminals (who would probably get caught for other reasons even without this search), the only thing this rather bizarre policy will do is cause Americans to become lackadaisical about our fourth amendment rights. Then again, maybe that's the point.