Whole Foods' whole business model involves projecting a "green" image while selling premium groceries at a significant markup over regular grocery stores. Their calculations for whether to put up rooftop solar panels have a lot more to do with their green image than with whether the electricity is actually cheaper, which is why you see them doing it, but not Safeway (which doesn't have the same image concerns).
It isn't true that "oil deposits in US federal reserves forbidden for drilling could supply the entire world demand for close to 500 years". Total US oil reserves are less than 3% of the world oil reserves, and could supply just the U.S. needs, if somehow we magically extracted all of them instantly, for about 3-4 years. So maybe if you factor in a bunch of new discoveries 10, even 20 might be plausible, but hardly 500.
"Almost all" Wikipedia pageviews are cached static HTML served up by a squid proxy, because there are orders of magnitude more non-logged-in readers than logged-in users, and many orders of magnitude more reads than edits. Only a small minority of traffic hits the database at all.
MediaWiki doesn't literally generate static HTML pages because it doesn't need to, since it's designed to be used with the rest of the infrastructure. The "static pages" are the ones served by the squid clusters, which is simpler architecturally (and more distributed) than having the core software literally generate static HTML pages. And the vast majority of Wikipedia pageviews are these static pages served out of squids.
For important documents there may be more procedures, but a lot of faxes are sent in a pretty routine manner with no authentication.
For example, I just faxed a copyright-transfer form to a journal so they'll publish a paper of mine. How did I fax it? From an online fax service, which didn't even require me to create an account. I gave them a PDF, and they faxed it. The only "authentication" is the receipt of this PDF at the other fax machine, which will be filed away somewhere; there is no other protocol being followed. Now why couldn't I have just emailed that same PDF to them? How does routing it through a free online fax service increase security?
Yahoo's stock has ranged hugely in value over the past year, despite its outlook not changing a lot, so I'd say the market has difficulty figuring out what Yahoo actually is worth, and changes its mind with the winds.
Now if Microsoft had offered a large premium that was completely above the range of variation, I could see your argument that it would be by far the best offer stockholders could expect. But they didn't---the stock's 52-week range is $18.50 to $34.00, and Microsoft offered around $30, which is solidly inside the pre-offer range, albeit at the upper end.
The datacenter pollution problem is entirely one of cost---we have ways other than coal to power them, we just have to figure out how to set up the market and subsidies and regulation and whatnot so that the alternative methods actually get used.
Airlines have a bigger problem, because there actually is no other way for them to fly, using current technology.
If you model internet access as people lower on the totem pole buying bandwidth from people higher on the totem pole, then, absent an infinite regression, you eventually have to get to somebody at the top.
Originally, this was the ARPANET backbone, later replaced by the NSFNet backbone---once you got to the government infrastructure, you could get anywhere else.
This has since been replaced by a set of large ISPs, the "tier 1" providers, who interconnect with each other for free, and sell bandwidth to everyone who isn't a Tier 1 peer. They essentially "are" the internet, in the sense that anyone who isn't one of them has to buy from one of them, or from someone who buys from one of them (or from someone who buys from someone who buys from one of them, etc.). ISPs like AT&T and Verizon are among them.
Many Tier 2 ISPs, which includes a lot of other ISPs, are close to the same, because much of their bandwidth comes via peering agreements with other Tier 2s, even if technically they have to reach some parts of the internet on a for-money basis via a Tier 1 ISP.
The four prongs are actually somewhat different than summarized above, though. Roughly, they're:
1. Is the speech of a sort that would be protected in the first place? (i.e. fraudulent advertising doesn't count)
2. Does the regulation address a substantial government interest?
3. Does the regulation have the effect of actually advancing the asserted government interest?
4. Is the regulation the least restrictive way of achieving that government interest?
The fourth prong has been weakened a bit in subsequent cases to only require the regulation to be a narrowly tailored one that results in a good fit between the governmental interest and the means of advancing it, since proving something to be the "least" restrictive way of achieving something is tough in practice, and even if you could, sometimes a simple but fairly narrow rule might still be preferable to a slightly less restrictive but hugely complex rule riddled with special cases.
As C/C++ apps get more complex, I've found that many in effect start reimplementing garbage collectors, only in really buggy, inefficient ways.
In simple apps, you can exactly police your memory patterns, and everything is created and freed in carefully defined locations. The only real complication here is that you usually (for efficiency) want an allocation strategy other than literally getting memory from the OS and returning it every time you do something minor, which fortunately C++ allocators can let you do.
But things get complicated pretty quickly, when objects can be accessed from lots of places, and there's no obvious "ok now we're done, free it" point. The usual first step is to manage this with reference counting. Encapsulations like autopointers can help make this easier to use. This is in effect a really crappy garbage collector---not very fast, and prone to memory leaks.
So what do you do next? Usually the first step is noticing that you have a bunch of memory leaks due to your reference counting leaving mutually-referential garbage. So you write a some code to look for objects with non-zero reference counts that aren't referenced by any 'real' live objects (Firefox just recently did this part). Now you have a slightly better ad-hoc garbage collector: reference counting plus cycle collection.
In many cases, by the end you would've been better off in all respects, including efficiency, if you just used an off-the-shelf garbage collector.
Appellate courts can second-guess the trial judge, but they don't have the power to second-guess the trial jury. The only way these sorts of factual determinations could be reviewed on an appeal is if the trial judge gave incorrect instructions to the jury (i.e. didn't properly explain to them what standards of evidence they should be looking for), or if the judge improperly put to the jury something that was a matter of law instead of fact (i.e. if the allegation, even if true, wouldn't support the charge due to not meeting the definitions in the statute).
But if the trial was run improperly, and the jury just convicted someone on weak evidence, appeals courts don't have much they can do about it. (They also can't do much about it if the jury acquitted someone despite very strong evidence.)
Watching YouTube videos is not substantially different. In many cases it's worse for intellect than watching television.
I'm not sure where you found that San Francisco
on
eBay Sues Craigslist
·
· Score: 1
The non-commercial, non-capitalist San Francisco doesn't seem to be one I've ever encountered. Are you talking about the same one that has one of the highest per-capita incomes of any city in the United States? Where the most recognizable element of the skyline is named after an insurance company? And where there's currently a flurry of building activity for luxury condos?
First of all, they don't get to design whatever cool game they want and just get NASA endorsement and advertising. NASA has very specific requirements for the bids, which basically requires it to be "edutainment", targeted at a middle-school and high-school-aged audience. This drastically limits the chance it'll ever be profitable.
Second, making a high-quality MMO just takes a lot of money. A small startup game developer would have to be extremely well funded by venture capital to do so.
Third, I think you may be overestimating the cachet of a NASA endorsement.
Wikipedia brings many issues to the forefront that are latent in other sources.
The two biggest are:
1. Since you can't assume the author has any particular expertise, citations are extra important, where many "real" works you'd find in the library are often a bit hand-wavy about things.
2. When there are multiple viewpoints, if the subject is at least vaguely non-obscure, people holding all the viewpoints will tend to show up on at Wikipedia article, and there will be arguments trying to hash out some article that neutrally summarizes all the major viewpoints. With a "real" book, authors (even well-credentialed ones) will often give short shrift to other major viewpoints in the field, either summarizing them uncharitably, or even not acknowledging their existence.
If anything, I think Real Nonfiction Authors could do with a bit of the discipline that the better parts of Wikipedia enforce, when it comes to citing their sources and acknowledging and neutrally summarizing competing views.
I'm as up for a good continental/analytic rollick as anyone else, but I don't see what your reply really has to do with the point. You were pointing out a "problem" with "philosophy in the US" that isn't really responsible for this particular shoddy article, since analytic philosophy doesn't actually make the naive determinism-invalidates-free-will claims that the article seems to assume.
As for "profundity", part of the problem is that it's quite difficult to separate "profundity" from "bullshitting". See, for example, anything written by that armchair philosopher (and Debord recycler) extraordinaire, Baudrillard. You can basically get rigorous and inconsequential/boring on the one hand, or you can get high-minded and grandiose but essentially opinion on the other hand.
Analogous schisms happen in other fields, for what it's worth. In artificial intelligence (my field), you have on the one hand statistical machine learning, which tackles fairly narrow problems but in a rigorous way that gives us information about when it fails and when it succeeds. On the other hand, you have more classical AI, which tackles much larger and more interesting problems, but usually in a way that leaves it hard to determine if the problem was actually solved (or at least addressed interestingly), and if so what information we ought to glean from the result. The two camps tend not to like each other. (I'm not really in either one myself.)
The story assumes that everyone agrees determinism (or at least some sort of predictability coming close to determinism) implies lack of free will. This has been debated for centuries, and is far from agreed upon; if anything, the position that free will and determinism (or something like it) are compatible, termed compatibilism, is probably the more widespread position in contemporary philosophy.
See the Stanford Encyclopedia of Philosophy article for more detail than you could possibly want. Among writers likely to be familiar to many Slashdotters, this is the position Daniel Dennett takes.
Slashdot Mirror
Whole Foods' whole business model involves projecting a "green" image while selling premium groceries at a significant markup over regular grocery stores. Their calculations for whether to put up rooftop solar panels have a lot more to do with their green image than with whether the electricity is actually cheaper, which is why you see them doing it, but not Safeway (which doesn't have the same image concerns).
It isn't true that "oil deposits in US federal reserves forbidden for drilling could supply the entire world demand for close to 500 years". Total US oil reserves are less than 3% of the world oil reserves, and could supply just the U.S. needs, if somehow we magically extracted all of them instantly, for about 3-4 years. So maybe if you factor in a bunch of new discoveries 10, even 20 might be plausible, but hardly 500.
"Almost all" Wikipedia pageviews are cached static HTML served up by a squid proxy, because there are orders of magnitude more non-logged-in readers than logged-in users, and many orders of magnitude more reads than edits. Only a small minority of traffic hits the database at all.
MediaWiki doesn't literally generate static HTML pages because it doesn't need to, since it's designed to be used with the rest of the infrastructure. The "static pages" are the ones served by the squid clusters, which is simpler architecturally (and more distributed) than having the core software literally generate static HTML pages. And the vast majority of Wikipedia pageviews are these static pages served out of squids.
Certainly not fuel-efficient, which is the main concern in this discussion.
Why should a bike, that can do 15-25 easily, have to go 5-10 on the sidewalk? Why not make cars that could go 45 go 25 instead?
I can't really judge the matter without a sampling of pictures now can I?
For important documents there may be more procedures, but a lot of faxes are sent in a pretty routine manner with no authentication.
For example, I just faxed a copyright-transfer form to a journal so they'll publish a paper of mine. How did I fax it? From an online fax service, which didn't even require me to create an account. I gave them a PDF, and they faxed it. The only "authentication" is the receipt of this PDF at the other fax machine, which will be filed away somewhere; there is no other protocol being followed. Now why couldn't I have just emailed that same PDF to them? How does routing it through a free online fax service increase security?
On the announcement of the offer, the stock price went up to fairly close to the offer value. Anyone who wanted to cash out should have done so.
Yahoo's stock has ranged hugely in value over the past year, despite its outlook not changing a lot, so I'd say the market has difficulty figuring out what Yahoo actually is worth, and changes its mind with the winds.
Now if Microsoft had offered a large premium that was completely above the range of variation, I could see your argument that it would be by far the best offer stockholders could expect. But they didn't---the stock's 52-week range is $18.50 to $34.00, and Microsoft offered around $30, which is solidly inside the pre-offer range, albeit at the upper end.
The datacenter pollution problem is entirely one of cost---we have ways other than coal to power them, we just have to figure out how to set up the market and subsidies and regulation and whatnot so that the alternative methods actually get used.
Airlines have a bigger problem, because there actually is no other way for them to fly, using current technology.
If you model internet access as people lower on the totem pole buying bandwidth from people higher on the totem pole, then, absent an infinite regression, you eventually have to get to somebody at the top.
Originally, this was the ARPANET backbone, later replaced by the NSFNet backbone---once you got to the government infrastructure, you could get anywhere else.
This has since been replaced by a set of large ISPs, the "tier 1" providers, who interconnect with each other for free, and sell bandwidth to everyone who isn't a Tier 1 peer. They essentially "are" the internet, in the sense that anyone who isn't one of them has to buy from one of them, or from someone who buys from one of them (or from someone who buys from someone who buys from one of them, etc.). ISPs like AT&T and Verizon are among them.
Many Tier 2 ISPs, which includes a lot of other ISPs, are close to the same, because much of their bandwidth comes via peering agreements with other Tier 2s, even if technically they have to reach some parts of the internet on a for-money basis via a Tier 1 ISP.
PC gaming is down below 15% of the total gaming market, which is starting to approach irrelevance on comparative terms at least.
If Jobs knew anything, he'd stick to taking his advice from Slashdot UIDs that are three digits or less.
The four-part test for regulation of commercial speech was first outlined in the 1980 U.S. Supreme Court case Central Hudson Gas & Electric Corp. v. Public Service Commission .
The four prongs are actually somewhat different than summarized above, though. Roughly, they're:
1. Is the speech of a sort that would be protected in the first place? (i.e. fraudulent advertising doesn't count)
2. Does the regulation address a substantial government interest?
3. Does the regulation have the effect of actually advancing the asserted government interest?
4. Is the regulation the least restrictive way of achieving that government interest?
The fourth prong has been weakened a bit in subsequent cases to only require the regulation to be a narrowly tailored one that results in a good fit between the governmental interest and the means of advancing it, since proving something to be the "least" restrictive way of achieving something is tough in practice, and even if you could, sometimes a simple but fairly narrow rule might still be preferable to a slightly less restrictive but hugely complex rule riddled with special cases.
As C/C++ apps get more complex, I've found that many in effect start reimplementing garbage collectors, only in really buggy, inefficient ways.
In simple apps, you can exactly police your memory patterns, and everything is created and freed in carefully defined locations. The only real complication here is that you usually (for efficiency) want an allocation strategy other than literally getting memory from the OS and returning it every time you do something minor, which fortunately C++ allocators can let you do.
But things get complicated pretty quickly, when objects can be accessed from lots of places, and there's no obvious "ok now we're done, free it" point. The usual first step is to manage this with reference counting. Encapsulations like autopointers can help make this easier to use. This is in effect a really crappy garbage collector---not very fast, and prone to memory leaks.
So what do you do next? Usually the first step is noticing that you have a bunch of memory leaks due to your reference counting leaving mutually-referential garbage. So you write a some code to look for objects with non-zero reference counts that aren't referenced by any 'real' live objects (Firefox just recently did this part). Now you have a slightly better ad-hoc garbage collector: reference counting plus cycle collection.
In many cases, by the end you would've been better off in all respects, including efficiency, if you just used an off-the-shelf garbage collector.
Appellate courts can second-guess the trial judge, but they don't have the power to second-guess the trial jury. The only way these sorts of factual determinations could be reviewed on an appeal is if the trial judge gave incorrect instructions to the jury (i.e. didn't properly explain to them what standards of evidence they should be looking for), or if the judge improperly put to the jury something that was a matter of law instead of fact (i.e. if the allegation, even if true, wouldn't support the charge due to not meeting the definitions in the statute).
But if the trial was run improperly, and the jury just convicted someone on weak evidence, appeals courts don't have much they can do about it. (They also can't do much about it if the jury acquitted someone despite very strong evidence.)
Unlike that shitty stuff on television, we produce only the finest art on youtube.com.
Watching YouTube videos is not substantially different. In many cases it's worse for intellect than watching television.
The non-commercial, non-capitalist San Francisco doesn't seem to be one I've ever encountered. Are you talking about the same one that has one of the highest per-capita incomes of any city in the United States? Where the most recognizable element of the skyline is named after an insurance company? And where there's currently a flurry of building activity for luxury condos?
If you wanted to find a language that'd make Java a speed demon, congratulations.
First of all, they don't get to design whatever cool game they want and just get NASA endorsement and advertising. NASA has very specific requirements for the bids, which basically requires it to be "edutainment", targeted at a middle-school and high-school-aged audience. This drastically limits the chance it'll ever be profitable.
Second, making a high-quality MMO just takes a lot of money. A small startup game developer would have to be extremely well funded by venture capital to do so.
Third, I think you may be overestimating the cachet of a NASA endorsement.
Wikipedia brings many issues to the forefront that are latent in other sources.
The two biggest are:
1. Since you can't assume the author has any particular expertise, citations are extra important, where many "real" works you'd find in the library are often a bit hand-wavy about things.
2. When there are multiple viewpoints, if the subject is at least vaguely non-obscure, people holding all the viewpoints will tend to show up on at Wikipedia article, and there will be arguments trying to hash out some article that neutrally summarizes all the major viewpoints. With a "real" book, authors (even well-credentialed ones) will often give short shrift to other major viewpoints in the field, either summarizing them uncharitably, or even not acknowledging their existence.
If anything, I think Real Nonfiction Authors could do with a bit of the discipline that the better parts of Wikipedia enforce, when it comes to citing their sources and acknowledging and neutrally summarizing competing views.
I'm as up for a good continental/analytic rollick as anyone else, but I don't see what your reply really has to do with the point. You were pointing out a "problem" with "philosophy in the US" that isn't really responsible for this particular shoddy article, since analytic philosophy doesn't actually make the naive determinism-invalidates-free-will claims that the article seems to assume.
As for "profundity", part of the problem is that it's quite difficult to separate "profundity" from "bullshitting". See, for example, anything written by that armchair philosopher (and Debord recycler) extraordinaire, Baudrillard. You can basically get rigorous and inconsequential/boring on the one hand, or you can get high-minded and grandiose but essentially opinion on the other hand.
Analogous schisms happen in other fields, for what it's worth. In artificial intelligence (my field), you have on the one hand statistical machine learning, which tackles fairly narrow problems but in a rigorous way that gives us information about when it fails and when it succeeds. On the other hand, you have more classical AI, which tackles much larger and more interesting problems, but usually in a way that leaves it hard to determine if the problem was actually solved (or at least addressed interestingly), and if so what information we ought to glean from the result. The two camps tend not to like each other. (I'm not really in either one myself.)
The story assumes that everyone agrees determinism (or at least some sort of predictability coming close to determinism) implies lack of free will. This has been debated for centuries, and is far from agreed upon; if anything, the position that free will and determinism (or something like it) are compatible, termed compatibilism, is probably the more widespread position in contemporary philosophy.
See the Stanford Encyclopedia of Philosophy article for more detail than you could possibly want. Among writers likely to be familiar to many Slashdotters, this is the position Daniel Dennett takes.