What helps improve code quality, is not having deadlines...
MS and other commercial organisations will ship code that isn't ready in order to meet deadlines, and put their developers under intense pressure to meet the deadlines causing them to cut corners.
Somewhat. I recently started working for Google, and I've been quite impressed at the intense focus on code quality here. I think a lot of it has to do with having management who understands the reality of engineering, but at Google it appears that when the time pressure is on, everyone understands that the right thing to do is to cut functionality, not quality.
At least in my little corner of Google. I can't speak for the rest of the company.
In any case, I don't think the dynamic you describe is inherent in any commercial development. Extremely common, yes, but not inherent. And what I said about review and scrutiny improving the quality of code is absolutely true, and independent of context.
It makes you wonder about code that MS doesn't have to let anyone else see i.e. their proprietary products.
It really doesn't. Bringing the code "up to the Linux" standard doesn't necessarily have anything to do with quality. I'm sure that the quality was improved by the process, but that's normal any time intense scrutiny is applied to a piece of code of any size. Another round of intense scrutiny would improve it some more. I'm sure much of it was also a matter of complying with Linux coding style standards. Changing from one style to another can produce a large volume of trivial changes.
I also would not be surprised if the quality of the code is Linux is somewhat higher than the quality level common at Microsoft. But I don't think this tells you one way or another.
While I agree about the peak in terms of free time being in college I suspect it's not entirely true that you are the busiest in your late forties or your fifties. You certainly can be but if you are some kind of "specialist"
All older programmers are aware of this fact. You younger programmers should find an older programmer of the relevant gender and experience this fact for yourselves.
For some I imagine this is necessary and allows the kid to actually get an education.
For a few, very few, it's really important. These are kids who in decades and centuries past would have just been written of as uneducable. For a lot, though, it's necessary and allows the kid to actually get an education from our education system. I think often 90% of the problem is the approach to education and 10% of it is the kid.
I don't really think you can compare Trazodone to Ambien, though. Trazodone tends to make you feel fuzzy after you wake up, while Ambien doesn't. Traz is also an anti-depressant, though, so for a lot of people it can do double duty.
What you say is true, but unrelated. Anti-psychotics aren't part of the drug path you mention. Among other things, they don't give you any kind of a "high", and many of them have some rather unpleasant side effects.
That should have been the original summary. Old people have more free time. Go figure.
It depends on what you mean by "older", but in general people have the most free time in their youth. College students have huge gobs of free time (though mostly don't realize it). Young, unmarried professionals have a little less. Married people with young children have even less, but still quite a bit. Married people with teenage children are typically the busiest of all, so the free-time "low" is generally around 45-50. After that, kids start to be less of a burden and free time returns to a bit better than the "young, unmarried professional" level (better because of greater earned vacation time and the established lifestyle).
This is all generalization, of course, but I think it's a sufficiently common trend to invalidate your explanation of the data.
I think the "older programmers know more stuff and can more easily provide useful answers to questions" is the simpler and more logical explanation.
So only one person in 10 billion is not "boring as hell"? That means there are only 0.7 people in the world the NSA finds interesting? I think you have a few too many nines there.
99.9999% of us are incapable of being weird enough to make their radar.
This is a bit better, but still... this means that only 300 people in the US are "weird enough to make their radar".
The mathematical equivalent of using five exclamation points aside, the real problem is that you don't know, and you can't know if you'll ever be "interesting" to the NSA, FBI, etc. Given the vast numbers of laws on the books -- including secret laws you're not allowed to know about -- you have no idea whether or not you're breaking even the written law, much less whether or not something you have done or will do may someday break some unwritten rule or just plain offend someone in power. And you certainly have no way of being sure that something you didn't really do, but only appeared to have done may someday offend someone in power.
We really need to bring our rogue government agencies under control. The NSA is actually not supposed to be allowed to spy on US citizens, you know. Not that it stops them.
I think it's odd that people are surprised by this at all. Eric Schmidt flat out stated that only people with something to hide care about privacy. That is your beloved Google, folks.
That's Schmidt's opinion. It's not the opinion of most (or even very many) of the people at Google, or of the people making the decisions at Google, nor is it the opinion of the current CEO. Google takes privacy very seriously. Yes, there have been a few mistakes, but everyone makes mistakes, and Google has come clean about its mistakes and tried to fix them.
However, when faced with court orders or -- perhaps, I don't know this true, but I'm supposing -- arm-twisting by the NSA or other powerful branches of the US government, Google has little choice but to comply. And, at least for Americans, criticizing Google for complying with government requests is rather stupid. The government is supposed to serve the interests of and be under the control of the American people. We should be fixing the government, not depending on corporations to exercise civil disobedience.
While I know more or less what the judge meant, and I think the ruling is good, I also think it's funny that, technically, there is no mathematical definition of randomness.
Our intuitive notion of randomness is, essentially, unpredictability. Usually when we look at trying to address unpredictability from a mathematical perspective, we (intially and naively) focus on trying to ensure uniform distribution. This makes sense, because we don't want any value to be more likely than any other value. But the output of a simple repeating counter is uniformly distributed but not at all unpredictable, so clearly uniform distribution isn't sufficient to ensure unpredictability. So we can apply a bunch of other mathematical tests, such as those in the Diehard suite, trying to find any correlations between different parts of the output stream. When we've verified that the stream passes every test we can think of then we've come as close as we can to a "mathematical" demonstration of randomness.
But that still doesn't give us any guarantee of unpredictability. Lots of deterministic PRNGs pass the Diehard suite -- indeed, that's the primary use and purpose of the Diehard tests, to evaluate the apparent randomness of PRNGs. But if you know the PRNG algorithm and any relevant state (e.g. the seed), then the result is no more unpredictable than a counter.
Indeed, it's probably more useful to look not for randomness not in mathematics, but in physics. Many quantum processes are believed to be inherently random, and therefore truly unpredictable.
Sort of.
The problem with physical processes is that although they may be inherently unpredictable at some level, they aren't necessarily uniformly distributed, which means that some values are more likely than others. This inherent bias towards some values and away from others means that although they aren't totally predictable, they are somewhat predictable, and in a way that can usually be exploited. Further, beyond the inherent biases, the measurement process often introduces additional biases.
We can get really good results by combining seed data from an inherently-random but biased source with a mathematical process that "smooths" the data into a nice uniform distribution without measurable biases. So I guess you could call that "mathematical randomness", but it depends as much on physics as it does mathematics.
I recently left IBM, after spending the last 15 years there. When I joined the company it was a great place to work, and I saw very little short-sighted decisionmaking. Over the years it went downhill, though, until I finally got fed up with all of it and jumped ship (and I was slow... virtually everyone I worked with wised up years before I did). You're right that IBM is still strong financially, but I maintain that they're just riding the wave of good decisions made in the past, and that the current crop of decisions are going to catch up with them. You're also right that IBM still spends a lot on R&D, but I saw firsthand the effects of the increased "profit focus" on the nature and depth of the R&D work being done by my friends in Somers and Zurich. I also have no doubt that unless there's a change in leadership, as soon as the financial situation starts to seriously tighten up, IBM will begin cutting R&D, because that's how SJP thinks.
HP is a great example of a company founded by engineers and later ran in to the ground by MBAs.
I think IBM is another case study in the making (again).
IBM began as a marketing-focused company (very early) and as it moved into the computer age acquired a strong (as in world-leading) engineering component while maintaining the market focus. Through the 70s and 80s, with the strong leadership of the Watsons gone, the market and engineering focus was replaced by bean counter optimization because IBM owned an unshakable monopoly and it was the most effective way to exploit that monopoly. The combination of (primarily) the PC revolution and (to a lesser degree) anti-trust litigation broke that monopoly and left IBM drowning in red ink.
Louis Gerstner came in and restored IBMs customer focus, increased investment in engineering and built a capable business & engineering services business. He put the company back on track and rebuilt its brand around a services-centric model that gave customers a one-stop shop for hardware, software and the services (business and technical) to make it go. When he left and Sam Palmisano took over, IBM began a long, gradual slide towards an MBA-driven company. Though IBM does still maintain one of the largest research divisions in the world, and employs a lot of brilliant people, researchers are increasingly facing demands that their work be quickly translatable into near-term profitable products, which (perhaps surprisingly to MBA types) is killing the profitability of their work. The accountants have been focused on achieving profits through cost reductions for a decade now, starting with reducing employee expenses through a wide variety of methods that have made IBM an increasingly unpleasant place to work.
Meanwhile, IBM's approach to growing the business is increasingly one of buying successful young software companies and marketing their products hard through IBM's extensive sales network while putting minimal, if any, effort into product enhancement. What little product improvement/development effort IBM continues is increasingly being pushed offshore. As various offshore organizations gradually build up the skills to be truly competent and useful, IBM has discovered their prices also rise so the MBA response has been to push out to other, lower-cost locations. With India now "too expensive", IBM is pushing more work off to South America, Eastern Europe and lower-cost areas of Asia -- and in doing so moving right back into the the problems that early offshoring efforts into India had with the added bonus of much more significant language barriers than existed in India.
Over and over again, the pattern of business in IBM today is micro-optimization of balance sheets, extreme focus on quarterly earnings and declining focus on the actions that are productive long-term: Developing customer relationships and creating compelling products.
Much of the discussion here has been on the conflict between marketing and development, but that misses the mark entirely. MBAs are not marketers. They don't know how to sell. They're all about business process optimization and financial optimization. Those are good things, but they're good for eking out marginal improvements in business effectiveness. They're a useful adjunct to the business components that REALLY matter -- making stuff to sell, and selling it. But when they begin to dominate, it's a problem.
Is IBM's current myopia going to cause it to fail? Almost certainly not. The micro-optimization and the buy-and-harvest approaches are currently working in the sense of growing revenues and profits. But the micro-optimization is just about played out, and customers are getting wise to the buy-and-harvest approach and becoming less willing to buy IBM. So IBM is heading towards a period of stagnation, at least, and serious decline at worst. The company is too big to die, though, and employes too many smart people, so what's going to happen is that
Beyond a certain scale, ACID compliance becomes the problem, not the solution. The mainframe DBs like IMS do provide high throughput from the standards of single organization workloads, but web-scale workloads managing petabytes of data, performing tens or hundreds of thousands of transactions per second and requiring extremely high uptimes are a different beast entirely. To address those, you start with sharding to split the load (and replicating the shards to provide for high availability), but when you don't have a nice shardable key then you have to look at distributed systems that simply cannot provide all of the ACID guarantees without very expensive commit protocols which destroy the performance you're hoping to gain.
To satisfy all of the requirements, then, you have to start giving up, or at least delaying, the ACID guarantees. The most common approach is to discard immediate consistency in favor of eventual consistency, and there are some nice protocols that make it work well... but not on those mainframe DBs.
It's fun to take curmudgeonly pleasure in pointing out that the problems young whippersnappers are scratching their heads over were solved 40 years ago, but it's really not the case here.
So hurray for MySQL. They saved 45-minutes during their installation on day one and now they'll spend a year or two plus millions of dollars to move away from their extremely dumb and uneducated decision. That's got to be one of the most expensive 45-minutes on earth - and yet its one of the single biggest decisions which MySQL users defend on a daily basis.
You're assuming that this random blogger has a clue, and that FB really is having problems due to their choice of MySQL. We have no evidence to support that.
I find the five-bladed razors are more comfortable to shave with and less likely to cut me. They're so good, in fact, that I can comfortably dry shave with them, though I only do that when I'm in a hurry because it significantly reduces the lifetime of the (expensive) blades.
As for shaving being gay... I actually wear a full beard and only shave my neck, but eventually plan to get my whole face lasered clean so I don't have to shave any more. I hate wasting time on that crap. I have a beard because it takes less time to trim it every couple of weeks than to shave it every day, not because I think I need facial hair to convince people I'm a man.
My outhouse is up at a camp site we use occasionally (and no one else uses; it's private land), so we decided to forgo the vent pipe, metal surfaces, etc. The floor has linoleum tile. I'd already planned to install a "urinal" the next time we go up. I made it out of a bleach bottle with the bottom cut off. We only use it during the summer so optimization for winter use isn't an issue.
So, anyone with half a brain has realized by now that there are ways to get around these backscatter devices, whether it's with surgically implanted devices, sticking devices up an orifice, or simply being in the other line at the checkpoint
Another option is to use plastic explosive and taper the edge so it blends smoothly into the skin. The backscatter machines can't distinguish between plastique and skin.
Okay, how DO you build an outhouse "correctly"? I built one a couple of weeks ago and it doesn't seem like there's very much to get wrong. Build a small shed with a bench at the back, cut a hole in the bench. Make sure you have decent ventilation -- some holes near the roof and floor covered with mosquito netting work well. Dig a good, deep hole and line the holes up. If you want to help the outhouse last, put it on some sort of a weatherproof base -- I used railroad ties.
Seems pretty hard to get wrong.
As for the rest of your comments, I suppose it depends on what sorts of things you think you need to survive through. If your concern is the complete and permanent collapse of civilization, then, yeah, you'd better be prepared to hold out for many years. But shorter-term disasters do occur -- and the average American family would be hungry inside of a week if the deliveries to the grocery stores stopped and the power went out (spoiling the food in the fridge/freezer). Most would be thirsty in a day or two if the water were cut off. And in a situation where all outside support failed, and without a supportive community, a rifle might well be a very useful tool by the end of two or three months. Not to hunt for food, but to keep your food from being taken.
My point is that a 90-120 day plan isn't silly. Perhaps it's not sufficient for all eventualities, but that doesn't mean it's not worthwhile. I figure I could feed, clothe, warm and protect my family for about a year based on what I have stored. Beyond that, we'd be in trouble. Perhaps we could make it to significant arable land and learn to raise food quickly enough, but probably not. That doesn't mean that my preparations are worthless, though. In fact, the most valuable preparations are those that will take care of your family for the first 72 hours in harsh conditions, because in most disasters help arrives within that time frame. As you extend your plan and materials further out you are preparing for less and less likely scenarios. Doesn't mean it's a bad idea, but I figure that beyond a year I'm well past the point of diminishing returns.
Slashdot Mirror
I think yours is a very dangerous attitude... and all the more dangerous because it's so common.
What helps improve code quality, is not having deadlines... MS and other commercial organisations will ship code that isn't ready in order to meet deadlines, and put their developers under intense pressure to meet the deadlines causing them to cut corners.
Somewhat. I recently started working for Google, and I've been quite impressed at the intense focus on code quality here. I think a lot of it has to do with having management who understands the reality of engineering, but at Google it appears that when the time pressure is on, everyone understands that the right thing to do is to cut functionality, not quality.
At least in my little corner of Google. I can't speak for the rest of the company.
In any case, I don't think the dynamic you describe is inherent in any commercial development. Extremely common, yes, but not inherent. And what I said about review and scrutiny improving the quality of code is absolutely true, and independent of context.
Keep in mind that this is news because it's unusual. All of the phone companies I've ever dealt with offer itemized bills by default.
Does their profit increase as costs increase?
No. Their profits decrease as costs increase, and they do care about minimizing costs.
Not all of them are particularly good at it, though. Never attribute to malice that which can be adequately explained by incompetence.
It makes you wonder about code that MS doesn't have to let anyone else see i.e. their proprietary products.
It really doesn't. Bringing the code "up to the Linux" standard doesn't necessarily have anything to do with quality. I'm sure that the quality was improved by the process, but that's normal any time intense scrutiny is applied to a piece of code of any size. Another round of intense scrutiny would improve it some more. I'm sure much of it was also a matter of complying with Linux coding style standards. Changing from one style to another can produce a large volume of trivial changes.
I also would not be surprised if the quality of the code is Linux is somewhat higher than the quality level common at Microsoft. But I don't think this tells you one way or another.
While I agree about the peak in terms of free time being in college I suspect it's not entirely true that you are the busiest in your late forties or your fifties. You certainly can be but if you are some kind of "specialist"
Or if you have kids.
... in bed.
All older programmers are aware of this fact. You younger programmers should find an older programmer of the relevant gender and experience this fact for yourselves.
For some I imagine this is necessary and allows the kid to actually get an education.
For a few, very few, it's really important. These are kids who in decades and centuries past would have just been written of as uneducable. For a lot, though, it's necessary and allows the kid to actually get an education from our education system. I think often 90% of the problem is the approach to education and 10% of it is the kid.
Melatonin is even cheaper.
I don't really think you can compare Trazodone to Ambien, though. Trazodone tends to make you feel fuzzy after you wake up, while Ambien doesn't. Traz is also an anti-depressant, though, so for a lot of people it can do double duty.
What you say is true, but unrelated. Anti-psychotics aren't part of the drug path you mention. Among other things, they don't give you any kind of a "high", and many of them have some rather unpleasant side effects.
That should have been the original summary. Old people have more free time. Go figure.
It depends on what you mean by "older", but in general people have the most free time in their youth. College students have huge gobs of free time (though mostly don't realize it). Young, unmarried professionals have a little less. Married people with young children have even less, but still quite a bit. Married people with teenage children are typically the busiest of all, so the free-time "low" is generally around 45-50. After that, kids start to be less of a burden and free time returns to a bit better than the "young, unmarried professional" level (better because of greater earned vacation time and the established lifestyle).
This is all generalization, of course, but I think it's a sufficiently common trend to invalidate your explanation of the data.
I think the "older programmers know more stuff and can more easily provide useful answers to questions" is the simpler and more logical explanation.
99.99999999% of us are boring as hell
So only one person in 10 billion is not "boring as hell"? That means there are only 0.7 people in the world the NSA finds interesting? I think you have a few too many nines there.
99.9999% of us are incapable of being weird enough to make their radar.
This is a bit better, but still... this means that only 300 people in the US are "weird enough to make their radar".
The mathematical equivalent of using five exclamation points aside, the real problem is that you don't know, and you can't know if you'll ever be "interesting" to the NSA, FBI, etc. Given the vast numbers of laws on the books -- including secret laws you're not allowed to know about -- you have no idea whether or not you're breaking even the written law, much less whether or not something you have done or will do may someday break some unwritten rule or just plain offend someone in power. And you certainly have no way of being sure that something you didn't really do, but only appeared to have done may someday offend someone in power.
We really need to bring our rogue government agencies under control. The NSA is actually not supposed to be allowed to spy on US citizens, you know. Not that it stops them.
I think it's odd that people are surprised by this at all. Eric Schmidt flat out stated that only people with something to hide care about privacy. That is your beloved Google, folks.
That's Schmidt's opinion. It's not the opinion of most (or even very many) of the people at Google, or of the people making the decisions at Google, nor is it the opinion of the current CEO. Google takes privacy very seriously. Yes, there have been a few mistakes, but everyone makes mistakes, and Google has come clean about its mistakes and tried to fix them.
However, when faced with court orders or -- perhaps, I don't know this true, but I'm supposing -- arm-twisting by the NSA or other powerful branches of the US government, Google has little choice but to comply. And, at least for Americans, criticizing Google for complying with government requests is rather stupid. The government is supposed to serve the interests of and be under the control of the American people. We should be fixing the government, not depending on corporations to exercise civil disobedience.
While I know more or less what the judge meant, and I think the ruling is good, I also think it's funny that, technically, there is no mathematical definition of randomness.
Our intuitive notion of randomness is, essentially, unpredictability. Usually when we look at trying to address unpredictability from a mathematical perspective, we (intially and naively) focus on trying to ensure uniform distribution. This makes sense, because we don't want any value to be more likely than any other value. But the output of a simple repeating counter is uniformly distributed but not at all unpredictable, so clearly uniform distribution isn't sufficient to ensure unpredictability. So we can apply a bunch of other mathematical tests, such as those in the Diehard suite, trying to find any correlations between different parts of the output stream. When we've verified that the stream passes every test we can think of then we've come as close as we can to a "mathematical" demonstration of randomness.
But that still doesn't give us any guarantee of unpredictability. Lots of deterministic PRNGs pass the Diehard suite -- indeed, that's the primary use and purpose of the Diehard tests, to evaluate the apparent randomness of PRNGs. But if you know the PRNG algorithm and any relevant state (e.g. the seed), then the result is no more unpredictable than a counter.
Indeed, it's probably more useful to look not for randomness not in mathematics, but in physics. Many quantum processes are believed to be inherently random, and therefore truly unpredictable.
Sort of.
The problem with physical processes is that although they may be inherently unpredictable at some level, they aren't necessarily uniformly distributed, which means that some values are more likely than others. This inherent bias towards some values and away from others means that although they aren't totally predictable, they are somewhat predictable, and in a way that can usually be exploited. Further, beyond the inherent biases, the measurement process often introduces additional biases.
We can get really good results by combining seed data from an inherently-random but biased source with a mathematical process that "smooths" the data into a nice uniform distribution without measurable biases. So I guess you could call that "mathematical randomness", but it depends as much on physics as it does mathematics.
I recently left IBM, after spending the last 15 years there. When I joined the company it was a great place to work, and I saw very little short-sighted decisionmaking. Over the years it went downhill, though, until I finally got fed up with all of it and jumped ship (and I was slow... virtually everyone I worked with wised up years before I did). You're right that IBM is still strong financially, but I maintain that they're just riding the wave of good decisions made in the past, and that the current crop of decisions are going to catch up with them. You're also right that IBM still spends a lot on R&D, but I saw firsthand the effects of the increased "profit focus" on the nature and depth of the R&D work being done by my friends in Somers and Zurich. I also have no doubt that unless there's a change in leadership, as soon as the financial situation starts to seriously tighten up, IBM will begin cutting R&D, because that's how SJP thinks.
HP is a great example of a company founded by engineers and later ran in to the ground by MBAs.
I think IBM is another case study in the making (again).
IBM began as a marketing-focused company (very early) and as it moved into the computer age acquired a strong (as in world-leading) engineering component while maintaining the market focus. Through the 70s and 80s, with the strong leadership of the Watsons gone, the market and engineering focus was replaced by bean counter optimization because IBM owned an unshakable monopoly and it was the most effective way to exploit that monopoly. The combination of (primarily) the PC revolution and (to a lesser degree) anti-trust litigation broke that monopoly and left IBM drowning in red ink.
Louis Gerstner came in and restored IBMs customer focus, increased investment in engineering and built a capable business & engineering services business. He put the company back on track and rebuilt its brand around a services-centric model that gave customers a one-stop shop for hardware, software and the services (business and technical) to make it go. When he left and Sam Palmisano took over, IBM began a long, gradual slide towards an MBA-driven company. Though IBM does still maintain one of the largest research divisions in the world, and employs a lot of brilliant people, researchers are increasingly facing demands that their work be quickly translatable into near-term profitable products, which (perhaps surprisingly to MBA types) is killing the profitability of their work. The accountants have been focused on achieving profits through cost reductions for a decade now, starting with reducing employee expenses through a wide variety of methods that have made IBM an increasingly unpleasant place to work.
Meanwhile, IBM's approach to growing the business is increasingly one of buying successful young software companies and marketing their products hard through IBM's extensive sales network while putting minimal, if any, effort into product enhancement. What little product improvement/development effort IBM continues is increasingly being pushed offshore. As various offshore organizations gradually build up the skills to be truly competent and useful, IBM has discovered their prices also rise so the MBA response has been to push out to other, lower-cost locations. With India now "too expensive", IBM is pushing more work off to South America, Eastern Europe and lower-cost areas of Asia -- and in doing so moving right back into the the problems that early offshoring efforts into India had with the added bonus of much more significant language barriers than existed in India.
Over and over again, the pattern of business in IBM today is micro-optimization of balance sheets, extreme focus on quarterly earnings and declining focus on the actions that are productive long-term: Developing customer relationships and creating compelling products.
Much of the discussion here has been on the conflict between marketing and development, but that misses the mark entirely. MBAs are not marketers. They don't know how to sell. They're all about business process optimization and financial optimization. Those are good things, but they're good for eking out marginal improvements in business effectiveness. They're a useful adjunct to the business components that REALLY matter -- making stuff to sell, and selling it. But when they begin to dominate, it's a problem.
Is IBM's current myopia going to cause it to fail? Almost certainly not. The micro-optimization and the buy-and-harvest approaches are currently working in the sense of growing revenues and profits. But the micro-optimization is just about played out, and customers are getting wise to the buy-and-harvest approach and becoming less willing to buy IBM. So IBM is heading towards a period of stagnation, at least, and serious decline at worst. The company is too big to die, though, and employes too many smart people, so what's going to happen is that
That video is content-free.
Doubtful.
Beyond a certain scale, ACID compliance becomes the problem, not the solution. The mainframe DBs like IMS do provide high throughput from the standards of single organization workloads, but web-scale workloads managing petabytes of data, performing tens or hundreds of thousands of transactions per second and requiring extremely high uptimes are a different beast entirely. To address those, you start with sharding to split the load (and replicating the shards to provide for high availability), but when you don't have a nice shardable key then you have to look at distributed systems that simply cannot provide all of the ACID guarantees without very expensive commit protocols which destroy the performance you're hoping to gain.
To satisfy all of the requirements, then, you have to start giving up, or at least delaying, the ACID guarantees. The most common approach is to discard immediate consistency in favor of eventual consistency, and there are some nice protocols that make it work well... but not on those mainframe DBs.
It's fun to take curmudgeonly pleasure in pointing out that the problems young whippersnappers are scratching their heads over were solved 40 years ago, but it's really not the case here.
So hurray for MySQL. They saved 45-minutes during their installation on day one and now they'll spend a year or two plus millions of dollars to move away from their extremely dumb and uneducated decision. That's got to be one of the most expensive 45-minutes on earth - and yet its one of the single biggest decisions which MySQL users defend on a daily basis.
You're assuming that this random blogger has a clue, and that FB really is having problems due to their choice of MySQL. We have no evidence to support that.
I'm of British and Danish origin, but have very heavy, coarse whiskers, and a lot of them.
I find the five-bladed razors are more comfortable to shave with and less likely to cut me. They're so good, in fact, that I can comfortably dry shave with them, though I only do that when I'm in a hurry because it significantly reduces the lifetime of the (expensive) blades.
As for shaving being gay... I actually wear a full beard and only shave my neck, but eventually plan to get my whole face lasered clean so I don't have to shave any more. I hate wasting time on that crap. I have a beard because it takes less time to trim it every couple of weeks than to shave it every day, not because I think I need facial hair to convince people I'm a man.
Great tips!
My outhouse is up at a camp site we use occasionally (and no one else uses; it's private land), so we decided to forgo the vent pipe, metal surfaces, etc. The floor has linoleum tile. I'd already planned to install a "urinal" the next time we go up. I made it out of a bleach bottle with the bottom cut off. We only use it during the summer so optimization for winter use isn't an issue.
So, anyone with half a brain has realized by now that there are ways to get around these backscatter devices, whether it's with surgically implanted devices, sticking devices up an orifice, or simply being in the other line at the checkpoint
Another option is to use plastic explosive and taper the edge so it blends smoothly into the skin. The backscatter machines can't distinguish between plastique and skin.
how to build a outhouse correctly
Okay, how DO you build an outhouse "correctly"? I built one a couple of weeks ago and it doesn't seem like there's very much to get wrong. Build a small shed with a bench at the back, cut a hole in the bench. Make sure you have decent ventilation -- some holes near the roof and floor covered with mosquito netting work well. Dig a good, deep hole and line the holes up. If you want to help the outhouse last, put it on some sort of a weatherproof base -- I used railroad ties.
Seems pretty hard to get wrong.
As for the rest of your comments, I suppose it depends on what sorts of things you think you need to survive through. If your concern is the complete and permanent collapse of civilization, then, yeah, you'd better be prepared to hold out for many years. But shorter-term disasters do occur -- and the average American family would be hungry inside of a week if the deliveries to the grocery stores stopped and the power went out (spoiling the food in the fridge/freezer). Most would be thirsty in a day or two if the water were cut off. And in a situation where all outside support failed, and without a supportive community, a rifle might well be a very useful tool by the end of two or three months. Not to hunt for food, but to keep your food from being taken.
My point is that a 90-120 day plan isn't silly. Perhaps it's not sufficient for all eventualities, but that doesn't mean it's not worthwhile. I figure I could feed, clothe, warm and protect my family for about a year based on what I have stored. Beyond that, we'd be in trouble. Perhaps we could make it to significant arable land and learn to raise food quickly enough, but probably not. That doesn't mean that my preparations are worthless, though. In fact, the most valuable preparations are those that will take care of your family for the first 72 hours in harsh conditions, because in most disasters help arrives within that time frame. As you extend your plan and materials further out you are preparing for less and less likely scenarios. Doesn't mean it's a bad idea, but I figure that beyond a year I'm well past the point of diminishing returns.
Wouldn't make any difference in this case, since no one knows what the patents are.