There is actually a lot of evidence to support the premise that fracking not only pollutes the ground water, but also causes mild earth quakes.
Cite? Because the mechanism of the first is very non-obvious and the second seems like a good thing.
On ground water pollution, unless we're having a lot of trouble with surface-level spills of fracking fluid (which argues for higher standards and tighter regulation, not banning), it's very difficult to see how injecting fracking fluid thousands of feet below the groundwater can have any effect on it.
On the quake issue, it seems clear that fracking cannot cause earthquakes, it can only accelerate and reduce earthquakes, making them happen sooner and more incrementally by lubricating the substrata enough to allow them to move and relieve pressure that is already present. In the long run that's a good thing. It means having many smaller and therefore less damaging quakes instead of fewer, larger and more dangerous ones.
Of course, both of my points are based on reasoning about theory, and actual evidence always trumps theory... but in spite of your claim I haven't seen any actual evidence. Do you have some?
It's going to be like that stupid Inbox stuff Google tried pulling a few years back, isn't it? I don't need something to create rules and read my email for me to sort it out. I can do both of those tasks just fine. Doing that doesn't save me effort or mental expense; just the opposite.
Inbox is awesome.
The thing you have to understand about Inbox, though, is that it's an e-mail client focused on the needs of people dealing with enormous volumes of email, and people whose email inbox (or at least a subset of it) represents their to-do list. If you get 400+ emails per day, including lots of emails from mailing lists and various automated systems, and including many emails that you don't actually need to read but just scan quickly, Inbox is a lifesaver.
What makes it great?
1. Gmail's labels and filters. The basis for all Inbox goodness already exists in Gmail, and in fact Inbox uses exactly the same infrastructure. The interface to that infrastructure is simplified in Inbox so, for example, common filter creation tasks can be done with a couple of clicks. If you need a more sophisticated rule, or one that does something other than applying a label, you use the Gmail interface and write a filter rule that's as clever and complex as you like (no regexps, unfortunately -- though if you really, really need to you can do it with Apps Script).
2. Bundles. Gmail labels are (optionally) represented in Inbox as bundles, which collapse all of the emails inside them into a single entry in your inbox. When you have a good set of bundles (labels) defined, and good filter rules, hundreds or thousands of emails in your inbox should be represented in 20 or so entries. You can click on one of them and it expands to show all of the emails inside it. Depending on the nature of the bundle, you already have some mental context for the list you're looking at, and how much attention you need to pay. Some bundles I frequently (but not always) archive without looking at them. Some bundles I read every email. Most bundles I just scan the subject lines, click to "pin" the handful that are interesting to me and "sweep" the rest.
Note that it may seem that you can do the same thing with traditional labels/folders and you can, almost. There's a big difference (to me, at least) in having a bundle show up in the inbox, rather than having to look over at the "unread message count" in lists of folders. Also...
3. Deferred bundles. You can specify when bundles should show up. Low-priority mailing list? Set it to show once per week, and Inbox will quietly collect emails all week and not show the bundle at all until Monday morning. Important bundle? The default is to show the bundle the instant any email arrives. There's a once-per-day priority as well. I'd actually like a little more granularity. I'd like to have hourly and twice-per-day bundles (morning and noon).
4. Snooze. This is the feature that makes email into an effective to-do list. I've always used my email as a todo list, leaving emails that are related to tasks I need to do sitting in my inbox, but what happens is that it very quickly becomes cluttered, making it difficult to find the emails I need to keep and those that can go. Also, it become cumbersome to scan through the list of stuff which -- for whatever reason -- I can't work on yet to find the stuff that I can. Snooze fixes all that. It allows you to tell Inbox to remove a message until a particular date/time, with some quick options like "tomorrow", "next week" and "someday". I also often snooze to location. I may check my email when I'm not at the office, just to get through some of the backlog, but I rarely want to type extensive answers on my phone, so I "Snooze to work".
By snoozing away everything that I can't work on right now, it becomes not only feasible but relatively easy to get to "inbox zero", meaning that my inbox is completely empty. Every morning I have a pile of bundles that I can plow through quickly (thanks to t
But gun laws do work. It's not as if this is an untried experiment - there are many countries with strict gun laws
Yeah? Suppose we looked up the homicide and gun ownership rates of all countries in the world, plotted them against one another and computed the correlation. What do you think we would find? Is there a correlation between violent death and gun ownership? I could link you to the answer, but it's much more enlightening to do the research yourself.
Throw some more guns into the mix and you'd end up with people trying to be the "good guy with a gun" they hear so much about, and shooting the shit out of each other every time a car backfired or a champagne cork popped.
There are lots of places in the country where there are several armed civilians in virtually every crowd. So if what you say is inevitable, shouldn't it have happened? I can provide links to several counterexamples, where people with concealed carry permits withheld their fire because they didn't have a clear shot. Concealed carry instructors (like me) stress to their students that they do not have the same level of protection from the system that police officers do, and that they are fully and completely responsible for every bullet they fire. What we see in practice is that people with permits do understand that and do behave responsibly -- far more responsibly than police officers, in fact. Which makes sense because cops know that they'll get the benefit of the doubt and have an organization to stand behind them and fund their legal defense, if needed.
If you live in a place where you feel you need to carry a gun, it's too late for you. That's not civilization, it's pathetic.
If you think you live in a place where there's no chance that you'll need a gun, you're delusional. No place is that safe.
Now, if what you believe is that the probability that you'll need a gun is very low, that other risks are greater, and that you, personally, wouldn't know what to do with a gun if you had one, that's a reasonable position. But it's not a counterargument for those of us who do know what we're doing. I agree that the probability that I'll ever need to draw my gun is very small, and I sincerely hope that it never happens. But I also believe that we're all safer if a significant fraction of law-abiding citizens is armed and prepared to respond to situations that may occur, because the police basically never arrive soon enough. They can't be everywhere.
You put way too much faith in doctors. Just because someone is "professionally trained" or has "years of experience" doesn't mean anything at all.
Well, it means they know some things that someone without training or experience does. If I'm hiring someone to write a program for me, I'm going to pick the guy with a reputable CS degree and a few years of successful experience over a guy who says he's really smart and "knows computers" but has never written a line of code or has any idea what a binary tree or a hash map are, or why he might care. Likewise, if I have a medical issue I'm going to hire someone who spent a few years studying the body and how to take care of it.
In both cases, I'm also going to look for the guy who stays current with the literature, reading books and journals in the relevant field, and the one who is able to communicate clearly with me about the issues involved and the rationales behind the decisions. So, yeah, there's are a lot of attributes required other than training and years of experience... but why in the world would you put any trust in someone who doesn't have them?
Otherwise I want to own copies of the music I want to listen to.
That's what I always thought, until I tried a "rental" service (I have a Google Music subscription), but the freedom of being able to listen to whatever I want without having to think about price is so great I can't imagine going back. I probably spend about the same on music as I did before, but now I listen to a lot more -- and a much wider variety -- of music than would have been possible by buying music.
You don't have a team of developers. You have a bunch of single-person teams. Don't try to manage it as a team.
I disagree. Even if they're working on different projects, if they're working on similar sorts of things -- similar platforms, tools, problems, etc. -- there's value in collaboration. Cross-pollination of design ideas, factoring out of common, reusable code components, sharing of ideas about what makes code better and more maintainable, collaborative problem solving -- there's a lot of value to be found in applying multiple brains and viewpoints.
There's also significant strategic value in ensuring that each of those one-person projects isn't known only by one person. Sharing knowledge around enables you to shift resources when it's useful, and helps to ensure that if something bad happens to one of your people someone else has a chance of picking the project up in a reasonable time frame.
In addition, most (not all, mind you) people like a little camaraderie at work. Building a sense of teamwork is good for morale. Even people who are pretty strongly introverted like a little interaction with like-minded people during their day.
First, if you truly have a bunch of single-person projects, you have an additional problem that you don't appear to have recognized: You have multiple single points of failure. I often use the notion of a "bus metric" which is "the number of people who have to be hit by a bus to cause serious problems for the organization". Your bus metric is 1, but it's a little worse than that, because you have several single individuals whose unavailability would cause problems.
This second problem actually points to part of my recommended solution: If you want to build team cohesion, you need to break down the silos between the projects and get your engineers working together more. At the same time, you don't want to impose a lot of process that will slow them down.
Some concrete suggestions (in no particular order):
1. Do code reviews. In fact, make them mandatory. In general, I'm against mandating much of anything but this is an exception because it is so extremely valuable. Mandate that every piece of code checked into the source repository must have been reviewed by someone other than the author. To make this effective and not an obstacle to getting work done, though, you need tool support. Something akin to Gerrit. Do not attempt to require code reviews until the infrastructure is in place, and once it is, ease into it and make sure everything is working well and everyone is comfortable with the tools before you mandate the reviews. However, be clear that the intention is to mandate them.
2. Do design reviews. Similar story to code reviews, except that design reviews should be attended by the entire team, or at least a largish subset. Again, good tooling is helpful. In this case using something like Google Docs or Office 360 for the design docs and then a good teleconference or, better yet, video conference solution for the review meeting. In terms of breaking down silos, cross-project design reviews will be hugely valuable.
3. Do standups via teleconference or (better yet) video conference. They don't need to be daily, and if your team is all working on different things they probably shouldn't be daily. Once a week for 15 minutes is a good place to start, just go around the "room" and have everyone briefly describe what they're working on and what challenges they're having. It's the discussion of challenges that make this valuable, because others will pipe up with their suggestions. Again, before you do this make sure that your infrastructure is in place and working *well*. Also, keep a close eye on how this is scaling. Beyond a certain number of participants it breaks down and you need to split it up.
4. Encourage lots of random, ad-hoc communication among the team. Again, good tools -- VC and shared docs, mostly, but you also need good chat and email systems -- are important for effective remote collaboration. As for how to encourage it, one good way is for you to get a solid understanding of what everyone is good at and when you have one-on-one meetings with them (which you should be doing regularly) look for opportunities to suggest collaborations.
5. Set up face-to-face get-togethers as frequently as you can, within the constraints of time and budget. Ensure that these contain equal parts business-related discussion of goals, plans, challenges, etc., and chances for people to do fun stuff together -- but make sure that you find activities that everyone can participate in and enjoy. IMO, you should avoid those obnoxious "team-building" programs.
6. If you can, consider getting budget for an additional engineer to focus on common tooling. The goal isn't to force all of your engineers into a common toolset against their will, but to build up some useful common tooling that all of them find valuable. Source control, code review system, continuous integration, automated testing, etc. This tool engineer will become another central point that everyone communicates with regularly (in addition to you).
What happens when 5 people stop off the sidewalk together and the only way to avoid that group of 5 is to run down another person who is on the sidewalk?
The issue with this problem is that the manufacturer is going to have to consider and program the car for this type of problem. In other words, a decision will have to be mostly made in advance.
That's not relevant to the question of whether the owner needs liability insurance. Whatever the decision is, and whatever liability accrues, it's on the maker of the self-driving system.
flufflemutter is right that the only insurance the owner of a purely automated car needs to care about is for protection against a tree falling on it, or similar.
If there is any criticism on my part in the post (and yes there is) is to this silly notion that " " 8. Given a large enough beta-tester and co-developer base, almost every problem will be characterized quickly and the fix obvious to someone.
Regardless of the development methodology essentially all significant programs will always contain bugs. Having lots of testers tends to find them faster, having a good QA team tends to find them faster yet... but there will always be bugs remaining.
Or, less formally, ``Given enough eyeballs, all bugs are shallow.'
Linus was saying that given enough people looking for the cause of a known bug, the cause will be obvious to someone. He didn't say anything about finding bugs.
The following is my idea of good taste (since the 1980s), whenever a comparison involves a constant term:
if ((options == (__WCLONE|__WALL)) && (0 = current->uid))
And when it doesn't involve a constant term, what then?
Better to make sure your compiler warns on assignments in conditionals (and maybe even turn warnings into errors), then write in comparisons in whatever order reads most naturally.
Disclaimer: I am a non-poor American, and I have chickens (six leghorn laying hens). Chickens are very easy to care for, and mine live mostly on table scraps, garden waste, and bugs.
This is interesting to me. I've been considering getting some chickens, but other family members who have them say they're actually quite expensive to feed, and argue that the resulting eggs, while good, are far more expensive than those from the grocery store. What's your take?
It's always about money, it's always about the economy. *ALWAYS*. VCs let you pretend otherwise for a little while, but ultimately you either a) start making money or b) go out of business.
Yes and no.
Yes, you have to make money or you'll be out of business. That is completely obvious... and it's a good thing in nearly all cases. Money is our proxy for utility. If what you're building isn't useful enough to enough people that it can generate revenue in excess of costs, it's probably something we shouldn't bother with. There are exceptions, of course, which is why there are charitable organizations, government funding, etc., but to a first order approximation, stuff that can't be profitable probably isn't worth the effort from a the perspective of society as a whole.
But that may or may not have anything to do with the motivations of the people actually doing the work. In many cases, people are motivated to do what they're doing because they think it's important and valuable, regardless of the profit potential. They recognize that it has to be profitable if they're going to be able to continue doing it, not least because they need to eat themselves, but that doesn't make money the motivation. For many people, including the sort I want to work with, profitability is a means, not an end.
Even if you get all of your electricity from dirty coal (like Colorado), it is still cleaner to drive an electric vehicle than a 35 mpg gas car
Also, if you really care you can often buy green power for a little more. When I lived in Colorado my Nissan Leaf ran primarily on wind and hydro power. Not because I cared so much but because I charged it primarily at work, and my employer cared enough to pay the premium for renewable energy. I could have done the same at home for a power bill about 20% higher.
Enriching Asians at the expense of Americans and Europeans is just robbing Peter to pay Paul.
This is your core fallacy. It's not a zero-sum game. Getting more of the world's population involved in producing and growing economically makes us all wealthier.
I refrained from insults, but you had to belittle patriotism as "ugly" and "tribalist". You are simply wrong. Being patriotic is simply respecting the values, ideals, and laws of the country you are either born in or immigrate to.
You're right about the definition of patriotism, you're wrong to label protectionist, xenophobic impulses as patriotic. Patriotism means respecting and valuing your own culture and people, but it in no way requires that you consider others to be less valuable.
They might even pull out their phone and verify that the claim to have "natural" ingredients even means anything
It doesn't. Oh, it may be true that the ingredients have been processed less, but it's still meaningless. Chemicals are chemicals. The only possible difference -- and only in some cases -- is that the "natural" chemical may be one that we have a bit more experience with. Maybe.
3D GPS does provide altitude, but it's significantly less precise than latitude and longitude. Your location is frequently accurate within a few feet, but altitude is always plus or minus at least a hundred feet.
Also, 3D GPS requires line of sight to several satellites, which is often difficult indoors.
But in this case it's a government that in one fell swoop changes it's mind and poof, everything changes.
No, that's not even slightly true. Coal mining has always been a horrible, polluting, dangerous business. Everyone involved with it has known -- or should have known -- for decades that it's unsustainable in the long run not only due to government recognition of its environmental impact (which itself has been a long time coming) but also the simple economics and the fact that mines are eventually depleted. These communities have had ample warning and opportunity to plan for this entirely expected and inevitable outcome!
And everyone also knows that at some point in their life they will become unable to work and will need retirement savings, so there's no need for social security, right? People should plan and prepare. I could give a dozen similar examples, but you get the point.
Either you believe in social safety nets, or you don't. If you do, then it's completely reasonable to think that the government has a responsibility to help out people whose industry is being shut down, *especially* when it's the government that's shutting it down. If, on the other hand, you believe that people should take responsibility for themselves and their own futures, then you should apply that belief across the board which means expecting them to predict and plan for all sorts of life events.
This is the result of globalism - impoverishment of the working classes in the US and Western Europe
And the dramatically greater enrichment of the working classes in Asia and Eastern Europe. While globalization has increased economic inequality in the first world, if you look at humanity as a whole it has hugely decreased economic inequality. That's an unabashedly Good Thing, and it's one we need to continue -- and we need to include Africa as well (which depends mostly on Africa getting its ducks in order; the reason it's been left out is because most African countries are still too unstable for businesses to operate in).
Yes, equalization of extreme global inequality is painful for those at the top, but that pain will taper off as wages in the developing world rise and as increased trade and technological progress increase global wealth.
the rise of an aggressive world power in Asia
That world power was already aggressive. It's richer now, and able to do more, but it's also become economically interdependent with the rest of the world which effectively means that it cannot be militarily aggressive without seriously harming itself. That's a good thing. Trade prevents wars.
and instability in the Middle East and it's aftermath.
In what way is globalism the cause of instability in the Middle East? You're really reaching there.
And yet nationalist movements, which want to undo this mess, are condemned as somehow evil.
Not evil so much as continuing a long and ugly history of tribalist thinking, to the long-term detriment of all, including those who believe they're protecting their own.
...Despite these gains, household income has dropped by about 8% in the last 10 years. So in summary, since 1995 (ish) we doubled our GDP (both per person and in absolute terms), and household income right now is about the level it was at the start of the doubling...
Translation: The rich got richer. A LOT richer.
If anyone was looking to try and find out exactly when the "great divide" happened that started to establish the financial powerhouse that makes up the 1% today, there ya go.
And the future is painted quite clear as that chasm between the 99% and the 1% continues to grow, and billionaires turn into trillionaires by replacing the middle class with automation to maximize throughput while minimizing all those costs related to employing humans. In a weird twist of irony, it will be the software industry that ushers this era in. We've heard of H1-B visas being used to force IT workers to train their replacements. I often wonder if the AI developer realizes they're doing the same thing.
This is a normal process which we've seen with every big technological change. New technology enables massive wealth generation, but the wealth always accrues first to those who are in a position to exploit the change. Then, over time, competition erodes the ability of the wealthy to keep the benefits concentrated, and they spread to the wider population. That's not to say the rich stop being rich, but the degree of inequality decreases. We've seen this pattern over and over again throughout history.
Note that I'm not arguing that we shouldn't use government or other mechanisms to accelerate the equalization, but it's worth recognizing that it's a common pattern that we should expect to see -- and we should be careful not to stop economic progress merely because it creates temporary increases in inequality, because the rising tide does lift all boats, eventually.
Yes, that's part of why I said we should think seriously about revising the law regarding expectation of privacy in public places. I still don't think judges are the right people to be making those changes, though.
Police are careful to control evidence, but that's not because failing to do so automatically excludes it, it's just because it opens an avenue for the defense to question it.
and because the police are generally required to have legally obtained their evidence. I'm sure you know that, but worth stating explicitly. The exclusionary rule is a bitch. Although it also probably would not save Clinton if this made it to trial.
That's true, but it's an orthogonal issue.
Evidence can be legally obtained but with a poorly-managed chain of custody. That evidence will not be excluded, but the defense will probably be able to discredit it. The jury will see it, but won't trust it.
Evidence that is obtained illegally will be excluded, so chain of custody is completely irrelevant. The jury will never see it.
It's worth noting that the whole exclusionary rule is something invented out of whole cloth by the legal system. It isn't in the Constitution, wasn't part of the framers' plan and hasn't been created legislatively. The framers' theory was that evidence turned up by warrantless searches (for example) was perfectly valid, but in the absence of a warrant the searcher was subject to civil litigation for trespassing. A warrant was used to provide legal authority to trespass, so the fourth amendment specifies that warrants may not be issued except on probable cause. In that time the modern notion of police forces empowered to investigate crimes didn't really exist; investigations and collection of evidence were generally carried out by private citizens, and the threat of litigation was a good deterrent to excessive action.
With the rise of professional police investigators, courts wanted to be able to immunize the police from prosecution, but that meant that police didn't have a reason to care about warrants at all. The exclusionary rule was the compromise. Personally, I think it's a bad one and that police should be personally prosecuted for breaking laws, but it's what we have.
Slashdot Mirror
There is actually a lot of evidence to support the premise that fracking not only pollutes the ground water, but also causes mild earth quakes.
Cite? Because the mechanism of the first is very non-obvious and the second seems like a good thing.
On ground water pollution, unless we're having a lot of trouble with surface-level spills of fracking fluid (which argues for higher standards and tighter regulation, not banning), it's very difficult to see how injecting fracking fluid thousands of feet below the groundwater can have any effect on it.
On the quake issue, it seems clear that fracking cannot cause earthquakes, it can only accelerate and reduce earthquakes, making them happen sooner and more incrementally by lubricating the substrata enough to allow them to move and relieve pressure that is already present. In the long run that's a good thing. It means having many smaller and therefore less damaging quakes instead of fewer, larger and more dangerous ones.
Of course, both of my points are based on reasoning about theory, and actual evidence always trumps theory... but in spite of your claim I haven't seen any actual evidence. Do you have some?
It's going to be like that stupid Inbox stuff Google tried pulling a few years back, isn't it? I don't need something to create rules and read my email for me to sort it out. I can do both of those tasks just fine. Doing that doesn't save me effort or mental expense; just the opposite.
Inbox is awesome.
The thing you have to understand about Inbox, though, is that it's an e-mail client focused on the needs of people dealing with enormous volumes of email, and people whose email inbox (or at least a subset of it) represents their to-do list. If you get 400+ emails per day, including lots of emails from mailing lists and various automated systems, and including many emails that you don't actually need to read but just scan quickly, Inbox is a lifesaver.
What makes it great?
1. Gmail's labels and filters. The basis for all Inbox goodness already exists in Gmail, and in fact Inbox uses exactly the same infrastructure. The interface to that infrastructure is simplified in Inbox so, for example, common filter creation tasks can be done with a couple of clicks. If you need a more sophisticated rule, or one that does something other than applying a label, you use the Gmail interface and write a filter rule that's as clever and complex as you like (no regexps, unfortunately -- though if you really, really need to you can do it with Apps Script).
2. Bundles. Gmail labels are (optionally) represented in Inbox as bundles, which collapse all of the emails inside them into a single entry in your inbox. When you have a good set of bundles (labels) defined, and good filter rules, hundreds or thousands of emails in your inbox should be represented in 20 or so entries. You can click on one of them and it expands to show all of the emails inside it. Depending on the nature of the bundle, you already have some mental context for the list you're looking at, and how much attention you need to pay. Some bundles I frequently (but not always) archive without looking at them. Some bundles I read every email. Most bundles I just scan the subject lines, click to "pin" the handful that are interesting to me and "sweep" the rest.
Note that it may seem that you can do the same thing with traditional labels/folders and you can, almost. There's a big difference (to me, at least) in having a bundle show up in the inbox, rather than having to look over at the "unread message count" in lists of folders. Also...
3. Deferred bundles. You can specify when bundles should show up. Low-priority mailing list? Set it to show once per week, and Inbox will quietly collect emails all week and not show the bundle at all until Monday morning. Important bundle? The default is to show the bundle the instant any email arrives. There's a once-per-day priority as well. I'd actually like a little more granularity. I'd like to have hourly and twice-per-day bundles (morning and noon).
4. Snooze. This is the feature that makes email into an effective to-do list. I've always used my email as a todo list, leaving emails that are related to tasks I need to do sitting in my inbox, but what happens is that it very quickly becomes cluttered, making it difficult to find the emails I need to keep and those that can go. Also, it become cumbersome to scan through the list of stuff which -- for whatever reason -- I can't work on yet to find the stuff that I can. Snooze fixes all that. It allows you to tell Inbox to remove a message until a particular date/time, with some quick options like "tomorrow", "next week" and "someday". I also often snooze to location. I may check my email when I'm not at the office, just to get through some of the backlog, but I rarely want to type extensive answers on my phone, so I "Snooze to work".
By snoozing away everything that I can't work on right now, it becomes not only feasible but relatively easy to get to "inbox zero", meaning that my inbox is completely empty. Every morning I have a pile of bundles that I can plow through quickly (thanks to t
I read "Steve Jobs" by Walter Isaacson and was blown away. I hadn't enjoyed a book that much in a long time, and I mainly read non-fiction.
So, you learned that you enjoy fiction more than you thought you did?
(j/k; I have no idea how accurate Isaacson's book is. Your comment motivated me to go grab the book on Audible.)
But gun laws do work. It's not as if this is an untried experiment - there are many countries with strict gun laws
Yeah? Suppose we looked up the homicide and gun ownership rates of all countries in the world, plotted them against one another and computed the correlation. What do you think we would find? Is there a correlation between violent death and gun ownership? I could link you to the answer, but it's much more enlightening to do the research yourself.
Throw some more guns into the mix and you'd end up with people trying to be the "good guy with a gun" they hear so much about, and shooting the shit out of each other every time a car backfired or a champagne cork popped.
There are lots of places in the country where there are several armed civilians in virtually every crowd. So if what you say is inevitable, shouldn't it have happened? I can provide links to several counterexamples, where people with concealed carry permits withheld their fire because they didn't have a clear shot. Concealed carry instructors (like me) stress to their students that they do not have the same level of protection from the system that police officers do, and that they are fully and completely responsible for every bullet they fire. What we see in practice is that people with permits do understand that and do behave responsibly -- far more responsibly than police officers, in fact. Which makes sense because cops know that they'll get the benefit of the doubt and have an organization to stand behind them and fund their legal defense, if needed.
If you live in a place where you feel you need to carry a gun, it's too late for you. That's not civilization, it's pathetic.
If you think you live in a place where there's no chance that you'll need a gun, you're delusional. No place is that safe.
Now, if what you believe is that the probability that you'll need a gun is very low, that other risks are greater, and that you, personally, wouldn't know what to do with a gun if you had one, that's a reasonable position. But it's not a counterargument for those of us who do know what we're doing. I agree that the probability that I'll ever need to draw my gun is very small, and I sincerely hope that it never happens. But I also believe that we're all safer if a significant fraction of law-abiding citizens is armed and prepared to respond to situations that may occur, because the police basically never arrive soon enough. They can't be everywhere.
You put way too much faith in doctors. Just because someone is "professionally trained" or has "years of experience" doesn't mean anything at all.
Well, it means they know some things that someone without training or experience does. If I'm hiring someone to write a program for me, I'm going to pick the guy with a reputable CS degree and a few years of successful experience over a guy who says he's really smart and "knows computers" but has never written a line of code or has any idea what a binary tree or a hash map are, or why he might care. Likewise, if I have a medical issue I'm going to hire someone who spent a few years studying the body and how to take care of it.
In both cases, I'm also going to look for the guy who stays current with the literature, reading books and journals in the relevant field, and the one who is able to communicate clearly with me about the issues involved and the rationales behind the decisions. So, yeah, there's are a lot of attributes required other than training and years of experience... but why in the world would you put any trust in someone who doesn't have them?
In other news: "IQ tests actually measure wealth"
Supposedly they corrected for wealth.
Otherwise I want to own copies of the music I want to listen to.
That's what I always thought, until I tried a "rental" service (I have a Google Music subscription), but the freedom of being able to listen to whatever I want without having to think about price is so great I can't imagine going back. I probably spend about the same on music as I did before, but now I listen to a lot more -- and a much wider variety -- of music than would have been possible by buying music.
You don't have a team of developers. You have a bunch of single-person teams. Don't try to manage it as a team.
I disagree. Even if they're working on different projects, if they're working on similar sorts of things -- similar platforms, tools, problems, etc. -- there's value in collaboration. Cross-pollination of design ideas, factoring out of common, reusable code components, sharing of ideas about what makes code better and more maintainable, collaborative problem solving -- there's a lot of value to be found in applying multiple brains and viewpoints.
There's also significant strategic value in ensuring that each of those one-person projects isn't known only by one person. Sharing knowledge around enables you to shift resources when it's useful, and helps to ensure that if something bad happens to one of your people someone else has a chance of picking the project up in a reasonable time frame.
In addition, most (not all, mind you) people like a little camaraderie at work. Building a sense of teamwork is good for morale. Even people who are pretty strongly introverted like a little interaction with like-minded people during their day.
First, if you truly have a bunch of single-person projects, you have an additional problem that you don't appear to have recognized: You have multiple single points of failure. I often use the notion of a "bus metric" which is "the number of people who have to be hit by a bus to cause serious problems for the organization". Your bus metric is 1, but it's a little worse than that, because you have several single individuals whose unavailability would cause problems.
This second problem actually points to part of my recommended solution: If you want to build team cohesion, you need to break down the silos between the projects and get your engineers working together more. At the same time, you don't want to impose a lot of process that will slow them down.
Some concrete suggestions (in no particular order):
1. Do code reviews. In fact, make them mandatory. In general, I'm against mandating much of anything but this is an exception because it is so extremely valuable. Mandate that every piece of code checked into the source repository must have been reviewed by someone other than the author. To make this effective and not an obstacle to getting work done, though, you need tool support. Something akin to Gerrit. Do not attempt to require code reviews until the infrastructure is in place, and once it is, ease into it and make sure everything is working well and everyone is comfortable with the tools before you mandate the reviews. However, be clear that the intention is to mandate them.
2. Do design reviews. Similar story to code reviews, except that design reviews should be attended by the entire team, or at least a largish subset. Again, good tooling is helpful. In this case using something like Google Docs or Office 360 for the design docs and then a good teleconference or, better yet, video conference solution for the review meeting. In terms of breaking down silos, cross-project design reviews will be hugely valuable.
3. Do standups via teleconference or (better yet) video conference. They don't need to be daily, and if your team is all working on different things they probably shouldn't be daily. Once a week for 15 minutes is a good place to start, just go around the "room" and have everyone briefly describe what they're working on and what challenges they're having. It's the discussion of challenges that make this valuable, because others will pipe up with their suggestions. Again, before you do this make sure that your infrastructure is in place and working *well*. Also, keep a close eye on how this is scaling. Beyond a certain number of participants it breaks down and you need to split it up.
4. Encourage lots of random, ad-hoc communication among the team. Again, good tools -- VC and shared docs, mostly, but you also need good chat and email systems -- are important for effective remote collaboration. As for how to encourage it, one good way is for you to get a solid understanding of what everyone is good at and when you have one-on-one meetings with them (which you should be doing regularly) look for opportunities to suggest collaborations.
5. Set up face-to-face get-togethers as frequently as you can, within the constraints of time and budget. Ensure that these contain equal parts business-related discussion of goals, plans, challenges, etc., and chances for people to do fun stuff together -- but make sure that you find activities that everyone can participate in and enjoy. IMO, you should avoid those obnoxious "team-building" programs.
6. If you can, consider getting budget for an additional engineer to focus on common tooling. The goal isn't to force all of your engineers into a common toolset against their will, but to build up some useful common tooling that all of them find valuable. Source control, code review system, continuous integration, automated testing, etc. This tool engineer will become another central point that everyone communicates with regularly (in addition to you).
7
What happens when 5 people stop off the sidewalk together and the only way to avoid that group of 5 is to run down another person who is on the sidewalk?
The issue with this problem is that the manufacturer is going to have to consider and program the car for this type of problem. In other words, a decision will have to be mostly made in advance.
That's not relevant to the question of whether the owner needs liability insurance. Whatever the decision is, and whatever liability accrues, it's on the maker of the self-driving system.
flufflemutter is right that the only insurance the owner of a purely automated car needs to care about is for protection against a tree falling on it, or similar.
If there is any criticism on my part in the post (and yes there is) is to this silly notion that " " 8. Given a large enough beta-tester and co-developer base, almost every problem will be characterized quickly and the fix obvious to someone.
Regardless of the development methodology essentially all significant programs will always contain bugs. Having lots of testers tends to find them faster, having a good QA team tends to find them faster yet... but there will always be bugs remaining.
Or, less formally, ``Given enough eyeballs, all bugs are shallow.'
Linus was saying that given enough people looking for the cause of a known bug, the cause will be obvious to someone. He didn't say anything about finding bugs.
The following is my idea of good taste (since the 1980s), whenever a comparison involves a constant term: if ((options == (__WCLONE|__WALL)) && (0 = current->uid))
And when it doesn't involve a constant term, what then?
Better to make sure your compiler warns on assignments in conditionals (and maybe even turn warnings into errors), then write in comparisons in whatever order reads most naturally.
Disclaimer: I am a non-poor American, and I have chickens (six leghorn laying hens). Chickens are very easy to care for, and mine live mostly on table scraps, garden waste, and bugs.
This is interesting to me. I've been considering getting some chickens, but other family members who have them say they're actually quite expensive to feed, and argue that the resulting eggs, while good, are far more expensive than those from the grocery store. What's your take?
It's always about money, it's always about the economy. *ALWAYS*. VCs let you pretend otherwise for a little while, but ultimately you either a) start making money or b) go out of business.
Yes and no.
Yes, you have to make money or you'll be out of business. That is completely obvious... and it's a good thing in nearly all cases. Money is our proxy for utility. If what you're building isn't useful enough to enough people that it can generate revenue in excess of costs, it's probably something we shouldn't bother with. There are exceptions, of course, which is why there are charitable organizations, government funding, etc., but to a first order approximation, stuff that can't be profitable probably isn't worth the effort from a the perspective of society as a whole.
But that may or may not have anything to do with the motivations of the people actually doing the work. In many cases, people are motivated to do what they're doing because they think it's important and valuable, regardless of the profit potential. They recognize that it has to be profitable if they're going to be able to continue doing it, not least because they need to eat themselves, but that doesn't make money the motivation. For many people, including the sort I want to work with, profitability is a means, not an end.
Even if you get all of your electricity from dirty coal (like Colorado), it is still cleaner to drive an electric vehicle than a 35 mpg gas car
Also, if you really care you can often buy green power for a little more. When I lived in Colorado my Nissan Leaf ran primarily on wind and hydro power. Not because I cared so much but because I charged it primarily at work, and my employer cared enough to pay the premium for renewable energy. I could have done the same at home for a power bill about 20% higher.
Lo-behold, not long after "Google Voice" was born with technology similar to our own...
Google Voice was a rebranding of a company Google purchased, Grand Central.
https://techcrunch.com/2009/03/11/grand-central-to-finally-launch-as-google-voice-its-very-very-good/
Enriching Asians at the expense of Americans and Europeans is just robbing Peter to pay Paul.
This is your core fallacy. It's not a zero-sum game. Getting more of the world's population involved in producing and growing economically makes us all wealthier.
I refrained from insults, but you had to belittle patriotism as "ugly" and "tribalist". You are simply wrong. Being patriotic is simply respecting the values, ideals, and laws of the country you are either born in or immigrate to.
You're right about the definition of patriotism, you're wrong to label protectionist, xenophobic impulses as patriotic. Patriotism means respecting and valuing your own culture and people, but it in no way requires that you consider others to be less valuable.
They might even pull out their phone and verify that the claim to have "natural" ingredients even means anything
It doesn't. Oh, it may be true that the ingredients have been processed less, but it's still meaningless. Chemicals are chemicals. The only possible difference -- and only in some cases -- is that the "natural" chemical may be one that we have a bit more experience with. Maybe.
GPS returns altitude.
Not very well.
3D GPS does provide altitude, but it's significantly less precise than latitude and longitude. Your location is frequently accurate within a few feet, but altitude is always plus or minus at least a hundred feet.
Also, 3D GPS requires line of sight to several satellites, which is often difficult indoors.
No, that's not even slightly true. Coal mining has always been a horrible, polluting, dangerous business. Everyone involved with it has known -- or should have known -- for decades that it's unsustainable in the long run not only due to government recognition of its environmental impact (which itself has been a long time coming) but also the simple economics and the fact that mines are eventually depleted. These communities have had ample warning and opportunity to plan for this entirely expected and inevitable outcome!
And everyone also knows that at some point in their life they will become unable to work and will need retirement savings, so there's no need for social security, right? People should plan and prepare. I could give a dozen similar examples, but you get the point.
Either you believe in social safety nets, or you don't. If you do, then it's completely reasonable to think that the government has a responsibility to help out people whose industry is being shut down, *especially* when it's the government that's shutting it down. If, on the other hand, you believe that people should take responsibility for themselves and their own futures, then you should apply that belief across the board which means expecting them to predict and plan for all sorts of life events.
This is the result of globalism - impoverishment of the working classes in the US and Western Europe
And the dramatically greater enrichment of the working classes in Asia and Eastern Europe. While globalization has increased economic inequality in the first world, if you look at humanity as a whole it has hugely decreased economic inequality. That's an unabashedly Good Thing, and it's one we need to continue -- and we need to include Africa as well (which depends mostly on Africa getting its ducks in order; the reason it's been left out is because most African countries are still too unstable for businesses to operate in).
Yes, equalization of extreme global inequality is painful for those at the top, but that pain will taper off as wages in the developing world rise and as increased trade and technological progress increase global wealth.
the rise of an aggressive world power in Asia
That world power was already aggressive. It's richer now, and able to do more, but it's also become economically interdependent with the rest of the world which effectively means that it cannot be militarily aggressive without seriously harming itself. That's a good thing. Trade prevents wars.
and instability in the Middle East and it's aftermath.
In what way is globalism the cause of instability in the Middle East? You're really reaching there.
And yet nationalist movements, which want to undo this mess, are condemned as somehow evil.
Not evil so much as continuing a long and ugly history of tribalist thinking, to the long-term detriment of all, including those who believe they're protecting their own.
...Despite these gains, household income has dropped by about 8% in the last 10 years. So in summary, since 1995 (ish) we doubled our GDP (both per person and in absolute terms), and household income right now is about the level it was at the start of the doubling...
Translation: The rich got richer. A LOT richer.
If anyone was looking to try and find out exactly when the "great divide" happened that started to establish the financial powerhouse that makes up the 1% today, there ya go.
And the future is painted quite clear as that chasm between the 99% and the 1% continues to grow, and billionaires turn into trillionaires by replacing the middle class with automation to maximize throughput while minimizing all those costs related to employing humans. In a weird twist of irony, it will be the software industry that ushers this era in. We've heard of H1-B visas being used to force IT workers to train their replacements. I often wonder if the AI developer realizes they're doing the same thing.
This is a normal process which we've seen with every big technological change. New technology enables massive wealth generation, but the wealth always accrues first to those who are in a position to exploit the change. Then, over time, competition erodes the ability of the wealthy to keep the benefits concentrated, and they spread to the wider population. That's not to say the rich stop being rich, but the degree of inequality decreases. We've seen this pattern over and over again throughout history.
Note that I'm not arguing that we shouldn't use government or other mechanisms to accelerate the equalization, but it's worth recognizing that it's a common pattern that we should expect to see -- and we should be careful not to stop economic progress merely because it creates temporary increases in inequality, because the rising tide does lift all boats, eventually.
Yes, that's part of why I said we should think seriously about revising the law regarding expectation of privacy in public places. I still don't think judges are the right people to be making those changes, though.
Police are careful to control evidence, but that's not because failing to do so automatically excludes it, it's just because it opens an avenue for the defense to question it.
and because the police are generally required to have legally obtained their evidence. I'm sure you know that, but worth stating explicitly. The exclusionary rule is a bitch. Although it also probably would not save Clinton if this made it to trial.
That's true, but it's an orthogonal issue.
Evidence can be legally obtained but with a poorly-managed chain of custody. That evidence will not be excluded, but the defense will probably be able to discredit it. The jury will see it, but won't trust it.
Evidence that is obtained illegally will be excluded, so chain of custody is completely irrelevant. The jury will never see it.
It's worth noting that the whole exclusionary rule is something invented out of whole cloth by the legal system. It isn't in the Constitution, wasn't part of the framers' plan and hasn't been created legislatively. The framers' theory was that evidence turned up by warrantless searches (for example) was perfectly valid, but in the absence of a warrant the searcher was subject to civil litigation for trespassing. A warrant was used to provide legal authority to trespass, so the fourth amendment specifies that warrants may not be issued except on probable cause. In that time the modern notion of police forces empowered to investigate crimes didn't really exist; investigations and collection of evidence were generally carried out by private citizens, and the threat of litigation was a good deterrent to excessive action.
With the rise of professional police investigators, courts wanted to be able to immunize the police from prosecution, but that meant that police didn't have a reason to care about warrants at all. The exclusionary rule was the compromise. Personally, I think it's a bad one and that police should be personally prosecuted for breaking laws, but it's what we have.