Yes, desalination obviously requires more energy than you get out of this method. But the point of the desalination is not energy production, it's freshwater production. You get freshwater out of your desalination plant. That requires using some amount of energy X. Instead of dumping the waste product of the desalination plant (highly-concentrated brine) somewhere, you use it with one of these devices to produce some amount of energy Y where Y is less than X.
The net result is that you end up with freshwater, and instead of spending X energy to get it, you had to spend only (X - Y) energy.
I don't know what that's supposed to mean, but sure, whatever. Usually the thing you do when you're angry at the government has something to do with the thing you're angry about. What are you angry about that you're demonstrating by printing a gun?
What are you so upset about that you have to give the middle finger to the government? The ready availability of guns in the USA? The non-existence of regulation of 3D printers? The lack of DRM on gun designs?
Because meth is illegal, a person who's using it doesn't advertize the fact. So how would you recognize anyone who used it in moderation?
I wouldn't, of course. As I've acknowledge three times now, I'm no expert on meth usage. But we have all sorts of data on usage for all sorts of illegal drugs. Where is the data on casual meth usage? This is now the third time I've asked for it.
And please don't wikipedia-link me stupid crap. That's a Slashdot addiction that has got to stop.
Do you realize said opium dens were run by the British with the specific goal of selling as much opium as possible, and as such didn't have any of the safeguards I proposed?
So in your meth dens, the dens would not be run with the intent to sell as much meth as possible? Is that how a licensed liquor establishment (a bar) operates?
Start here [wikipedia.org].
That link has a chart that lists heroin and cocaine (two drugs I called out as bad candidates for legalization) as very harmful. It lists meth as more harmful than alcohol. It has no information on addiction rate. It does have information on how dependent an addict is on the drug.
Higher latency between the memory and the GPU which is KEY to a GPU performance.
Latency doesn't matter very much for GPU. A little, but not much. Bandwidth is what matters for GPU. (Latency matters for CPU tasks.) As long as the GPU has enough buffer depth to cover the latency to and from memory (which it certainly does), the memory bandwidth is what will keep the GPU pipelines completely full.
Just out of curiosity: have you ever actually seen a drunk person?
Yes, of course. But did you read the next couple sentences of my post? I've seen alcohol users drunk, but I've also seen many, many alcohol users NOT drunk, and as I stated, the vast majority of alcohol users I've seen have not been drunk. What I haven't heard much about is moderation in meth users. Every time I hear about meth (which, like I acknowledged in my post, is not much), I hear about meth heads being super high and totally messed up.
In any case, you could always restrict the sale of some substance to a licensed establishment, and have them require an agreement to stay on the premises while under influence as a condition of sale. Licensed opium dens, in other words. This would also allow monitoring of usage, and medical intervention in case of addiction.
Sure. You realize though that China figured out that even with legal opium dens, it was still bad to have a good fraction of the population addicated to it?
What I'm asking for is non-anecdotal studies on the addiction and abuse rates of the drugs everyone is proposing to legalize. If the addiction and/or abuse rates are too high, then I'm contending that legalizing them is still bad for society. You've trade one set of problems (black market drug trade, drug crime, etc.) for a different set (a good chunk of your population is whacked out on whatever drugs are now legal) and it's not an obvious call to me that the problems you introduce with legalization are actually superior.
Meth addicts don't only do stupid things just to get to the next fix. What they do while on meth is pretty stupid too. That's where the comparison between "hard drugs" and alcohol fails. Is there such a thing as non-binge/moderate meth usage? There is such a thing as non-binge alcohol usage, where someone has one or two drinks with dinner and feels a little relaxed and isn't totally messed up. I would say that's the vast majority of alcohol usage. I've never heard of that with meth (or cocaine, or heroin, for example), but please show me some evidence otherwise if it exists. I think it's because the effect is stronger and more addictive than alcohol, so users pretty instantly start going for the totally-fucked-up feeling rather than the moderate feeling.
Quad core doesn't help you on single-threaded, lightly-threaded, or GPU benchmarks, which is most of the benchmarks that I saw in the article. That means you can't say that Krait sucks because it has four cores and barely beats a dual core, since the four cores aren't being utilized. The conclusion you can draw is that a quad core CPU isn't necessary for a good user experience on a phone.
And needing a faster clock to reach the same performance levels isn't a meaningful metric either, at least not in a phone. In a phone that is power-constrained, the metric is performance per watt. If both CPUs burn the same power and give the same performance, they're basically equivalent. How each chooses to provide that performance is immaterial in a phone which is a power-constrained environment. Maybe the Apple CPU has some performance headroom at higher power budgets if it could run at a faster frequency (thus providing higher perf at the same freq as the Krait core), but that doesn't help you if it has to run throttled at all times so as to not blow through the phone's battery life and/or burn your pants.
Most things being reported by mainstream newsmedia fit the description "stuff that matters". The question is, why does this matter to STEM nerds, which ostensibly is the intent of this website? (I say STEM nerds because just about anyone interested in a particular thing can be called a " nerd" nowadays; however this website started when nerd had a more specific meaning.)
The answer is, Slashdot cherry-picks clickbait, polarizing-to-STEM-nerds stories out of mainstream newsmedia. Rather than having any real STEM connection, they are merely topics about which STEM nerds tend to have strong opinions about. Common topics are almost always political: political debates climate change, political debates about alternative energy, political debates about privacy, American politics, and American foreign policy. (Note I did not say scientific or technical discussions of those topics. This website doens't really have those.) Even the Microsoft stories, which are at least about a tech company instead of about politics or mainstream news, tend to be non-tech pieces about how terrible Microsoft's business practices are, or yet another article about how Win8 Metro sucks.
Okay, cut one carrier group, fine. Why not two or three? Where is your line? Should the U.S. not have any carrier groups?
Let's say we cut defense spending by 80% and pour the money into infrastructure and services: roads and bridges, fiber internet, national healthcare, social security. Then DPRK nukes a bunch of people. Would you say that that money was better spent on the infrastructure and services?
Well, it may be multi-billion-dollar industry, but it spectacularly failed to meet its sales projections. My absolute favorite Itanium sales chart can be found here. Granted some of those initial projections were crazy stupid. But it fell short of even the much more modest, revised projections from 2002 and 2003.
No ARM CPU exists that provides performance that is comparable to an i3/i5/i7 x86 CPU. And if one did exist, it would have similar power characteristics. If the OP is willing to give up big-core CPU performance and would be willing to accept ARM-level performance, Intel's latest Atom, "Bay Trail", has as good or better performance and power characteristics to the low-power ARM stuff available today.
He acknowledged that it is an improvement, not a solution.
And then instead of saying any of the things that you said as to why it's not a complete solution (which are reasonable rebuttals), he just complained about underpowered console GPUs and compared them to high-end PC GPUs.
What he really did was read the first line of the response, ignore the rest of it, and assume that the argument that the response was making that games written for the new consoles would be better only because they would be targetting a GPU that is more powerful than the last-gen consoles. That was not the respondant's argument.
I certainly agree that there is effort involved to make a game originally written for a console look better on more powerful PC hardware. The respondant's point (which is probably wishful thinking, but still a point nonetheless) is that instead of spending developer effort and budget on simply porting the game to work on PC CPU+GPU instruction sets, their effort can be spent in all of those things that you named since the instruction sets of the consoles already match that which you'd find in a PC.
The gripe is not that consoles are less powerful than PCs. The gripe is that many games are designed around the limitations of consoles and don't take advantage of all of the power in a PC.
If you would read the very next post in the subthread, (here it is), it has a reasonable response to that gripe. Here, I'll quote it for you:
I'm looking forward to the optimizations and the fact that they'll be the same architecture as discrete cards. Hopefully that means game developers will allow their games to scale more since it shouldn't really be much work and they don't need to port them.
If that's hard to understand, I'll explain it. The console GPU architecture is basically PC GPU architecture, even though it's not quite as powerful as the best PC graphics cards. So the effort required by the game developers to use better PC hardware is hopefully low since it should be a pretty natural extension of what they're already doing for the consoles, as opposed to something totally different like it was for the last-gen consoles.
The OP then acknowledged this point, but then again complained about the lack of GPU horsepower in the consoles and compared it to high-end PC GPUs. I didn't really understand this to be anything other than complaining about the consoles being underpowered, since it basically ignored the response. What else is there to do at that point besides acknowledging the OP's complaint at face value and offering an explanation for it?
Does that make sense now or do you want to have another try?
I could have chosen any of the other first twenty hits on Google search.
If you had done that, you would have had at least a 50% chance of discovering that the pedantic insult you were about to write was incorrect, and that that phrase has gained an "everyday" meaning that fits my definition perfectly.
Ah, of course. The all-powerful Wikipedia link, used to refute all manner of falseitude here on Slashdot. You know, there's more than one definition. Here's an equally-valid link showing an alternate definition. Thanks for your pedantry, and thanks for serving to reinforce the Slashdot written-idiocy stereotype, dinfinity!
And here you are reading Slashdot comments, which are known to be a bastion of rational thought and completely devoid of groupthink, cognitive dissonance, reductio ad absurdum, and many other forms of written idiocy.
You did not argue that people's constitutional rights were being violated. You argued that it is not government's job to attempt to provide public safety because that is not specified directly in the constitution. That is clearly wrong, and that's what I addressed with my post.
As I mentioned in my original post, if you want to question the manner in which the government is providing that service, i.e. that it is violating constitutional rights in doing so, that's a fine debate and argument to have. But that's not what you did.
When you had to do this clean build, were you in a critical code-writing/code-debugging loop? That was my main point. How often do you upgrade all the libraries in your project necessitating a clean build AND you're in a critical build loop?
Slashdot Mirror
Yes, desalination obviously requires more energy than you get out of this method. But the point of the desalination is not energy production, it's freshwater production. You get freshwater out of your desalination plant. That requires using some amount of energy X. Instead of dumping the waste product of the desalination plant (highly-concentrated brine) somewhere, you use it with one of these devices to produce some amount of energy Y where Y is less than X.
The net result is that you end up with freshwater, and instead of spending X energy to get it, you had to spend only (X - Y) energy.
He very clearly objected to the arbitrary price tag attached to the externality, not that there was no externality.
How does gun-printing convey that that's what you're angry about?
I don't know what that's supposed to mean, but sure, whatever. Usually the thing you do when you're angry at the government has something to do with the thing you're angry about. What are you angry about that you're demonstrating by printing a gun?
What are you so upset about that you have to give the middle finger to the government? The ready availability of guns in the USA? The non-existence of regulation of 3D printers? The lack of DRM on gun designs?
Because meth is illegal, a person who's using it doesn't advertize the fact. So how would you recognize anyone who used it in moderation?
I wouldn't, of course. As I've acknowledge three times now, I'm no expert on meth usage. But we have all sorts of data on usage for all sorts of illegal drugs. Where is the data on casual meth usage? This is now the third time I've asked for it.
And please don't wikipedia-link me stupid crap. That's a Slashdot addiction that has got to stop.
Do you realize said opium dens were run by the British with the specific goal of selling as much opium as possible, and as such didn't have any of the safeguards I proposed?
So in your meth dens, the dens would not be run with the intent to sell as much meth as possible? Is that how a licensed liquor establishment (a bar) operates?
Start here [wikipedia.org].
That link has a chart that lists heroin and cocaine (two drugs I called out as bad candidates for legalization) as very harmful. It lists meth as more harmful than alcohol. It has no information on addiction rate. It does have information on how dependent an addict is on the drug.
Higher latency between the memory and the GPU which is KEY to a GPU performance.
Latency doesn't matter very much for GPU. A little, but not much. Bandwidth is what matters for GPU. (Latency matters for CPU tasks.) As long as the GPU has enough buffer depth to cover the latency to and from memory (which it certainly does), the memory bandwidth is what will keep the GPU pipelines completely full.
Just out of curiosity: have you ever actually seen a drunk person?
Yes, of course. But did you read the next couple sentences of my post? I've seen alcohol users drunk, but I've also seen many, many alcohol users NOT drunk, and as I stated, the vast majority of alcohol users I've seen have not been drunk. What I haven't heard much about is moderation in meth users. Every time I hear about meth (which, like I acknowledged in my post, is not much), I hear about meth heads being super high and totally messed up.
In any case, you could always restrict the sale of some substance to a licensed establishment, and have them require an agreement to stay on the premises while under influence as a condition of sale. Licensed opium dens, in other words. This would also allow monitoring of usage, and medical intervention in case of addiction.
Sure. You realize though that China figured out that even with legal opium dens, it was still bad to have a good fraction of the population addicated to it?
What I'm asking for is non-anecdotal studies on the addiction and abuse rates of the drugs everyone is proposing to legalize. If the addiction and/or abuse rates are too high, then I'm contending that legalizing them is still bad for society. You've trade one set of problems (black market drug trade, drug crime, etc.) for a different set (a good chunk of your population is whacked out on whatever drugs are now legal) and it's not an obvious call to me that the problems you introduce with legalization are actually superior.
Meth addicts don't only do stupid things just to get to the next fix. What they do while on meth is pretty stupid too. That's where the comparison between "hard drugs" and alcohol fails. Is there such a thing as non-binge/moderate meth usage? There is such a thing as non-binge alcohol usage, where someone has one or two drinks with dinner and feels a little relaxed and isn't totally messed up. I would say that's the vast majority of alcohol usage. I've never heard of that with meth (or cocaine, or heroin, for example), but please show me some evidence otherwise if it exists. I think it's because the effect is stronger and more addictive than alcohol, so users pretty instantly start going for the totally-fucked-up feeling rather than the moderate feeling.
Quad core doesn't help you on single-threaded, lightly-threaded, or GPU benchmarks, which is most of the benchmarks that I saw in the article. That means you can't say that Krait sucks because it has four cores and barely beats a dual core, since the four cores aren't being utilized. The conclusion you can draw is that a quad core CPU isn't necessary for a good user experience on a phone.
And needing a faster clock to reach the same performance levels isn't a meaningful metric either, at least not in a phone. In a phone that is power-constrained, the metric is performance per watt. If both CPUs burn the same power and give the same performance, they're basically equivalent. How each chooses to provide that performance is immaterial in a phone which is a power-constrained environment. Maybe the Apple CPU has some performance headroom at higher power budgets if it could run at a faster frequency (thus providing higher perf at the same freq as the Krait core), but that doesn't help you if it has to run throttled at all times so as to not blow through the phone's battery life and/or burn your pants.
Most things being reported by mainstream newsmedia fit the description "stuff that matters". The question is, why does this matter to STEM nerds, which ostensibly is the intent of this website? (I say STEM nerds because just about anyone interested in a particular thing can be called a " nerd" nowadays; however this website started when nerd had a more specific meaning.)
The answer is, Slashdot cherry-picks clickbait, polarizing-to-STEM-nerds stories out of mainstream newsmedia. Rather than having any real STEM connection, they are merely topics about which STEM nerds tend to have strong opinions about. Common topics are almost always political: political debates climate change, political debates about alternative energy, political debates about privacy, American politics, and American foreign policy. (Note I did not say scientific or technical discussions of those topics. This website doens't really have those.) Even the Microsoft stories, which are at least about a tech company instead of about politics or mainstream news, tend to be non-tech pieces about how terrible Microsoft's business practices are, or yet another article about how Win8 Metro sucks.
Okay, cut one carrier group, fine. Why not two or three? Where is your line? Should the U.S. not have any carrier groups?
Let's say we cut defense spending by 80% and pour the money into infrastructure and services: roads and bridges, fiber internet, national healthcare, social security. Then DPRK nukes a bunch of people. Would you say that that money was better spent on the infrastructure and services?
Well, it may be multi-billion-dollar industry, but it spectacularly failed to meet its sales projections. My absolute favorite Itanium sales chart can be found here. Granted some of those initial projections were crazy stupid. But it fell short of even the much more modest, revised projections from 2002 and 2003.
No ARM CPU exists that provides performance that is comparable to an i3/i5/i7 x86 CPU. And if one did exist, it would have similar power characteristics. If the OP is willing to give up big-core CPU performance and would be willing to accept ARM-level performance, Intel's latest Atom, "Bay Trail", has as good or better performance and power characteristics to the low-power ARM stuff available today.
but I really just want to see some ARM systems put out that are comparable to modern x86 machines in terms of specs.
Why do you care if it's ARM? There are plenty of actual x86 machines that would meet your requirements available today.
He acknowledged that it is an improvement, not a solution.
And then instead of saying any of the things that you said as to why it's not a complete solution (which are reasonable rebuttals), he just complained about underpowered console GPUs and compared them to high-end PC GPUs.
What he really did was read the first line of the response, ignore the rest of it, and assume that the argument that the response was making that games written for the new consoles would be better only because they would be targetting a GPU that is more powerful than the last-gen consoles. That was not the respondant's argument.
I certainly agree that there is effort involved to make a game originally written for a console look better on more powerful PC hardware. The respondant's point (which is probably wishful thinking, but still a point nonetheless) is that instead of spending developer effort and budget on simply porting the game to work on PC CPU+GPU instruction sets, their effort can be spent in all of those things that you named since the instruction sets of the consoles already match that which you'd find in a PC.
The gripe is not that consoles are less powerful than PCs. The gripe is that many games are designed around the limitations of consoles and don't take advantage of all of the power in a PC.
If you would read the very next post in the subthread, (here it is), it has a reasonable response to that gripe. Here, I'll quote it for you:
I'm looking forward to the optimizations and the fact that they'll be the same architecture as discrete cards. Hopefully that means game developers will allow their games to scale more since it shouldn't really be much work and they don't need to port them.
If that's hard to understand, I'll explain it. The console GPU architecture is basically PC GPU architecture, even though it's not quite as powerful as the best PC graphics cards. So the effort required by the game developers to use better PC hardware is hopefully low since it should be a pretty natural extension of what they're already doing for the consoles, as opposed to something totally different like it was for the last-gen consoles.
The OP then acknowledged this point, but then again complained about the lack of GPU horsepower in the consoles and compared it to high-end PC GPUs. I didn't really understand this to be anything other than complaining about the consoles being underpowered, since it basically ignored the response. What else is there to do at that point besides acknowledging the OP's complaint at face value and offering an explanation for it?
Does that make sense now or do you want to have another try?
I've read and reviewed the entire subthread again and can't find anything that I didn't comprehend. Please, englighten me!
You mean a machine that costs 4-5x what the console costs will be more powerful than the console? Shocking!
Sure! Is that better or worse than pretending that there is no alternate definition for reductio ad absurdum?
I could have chosen any of the other first twenty hits on Google search.
If you had done that, you would have had at least a 50% chance of discovering that the pedantic insult you were about to write was incorrect, and that that phrase has gained an "everyday" meaning that fits my definition perfectly.
Ah, of course. The all-powerful Wikipedia link, used to refute all manner of falseitude here on Slashdot. You know, there's more than one definition. Here's an equally-valid link showing an alternate definition. Thanks for your pedantry, and thanks for serving to reinforce the Slashdot written-idiocy stereotype, dinfinity!
And here you are reading Slashdot comments, which are known to be a bastion of rational thought and completely devoid of groupthink, cognitive dissonance, reductio ad absurdum, and many other forms of written idiocy.
You did not argue that people's constitutional rights were being violated. You argued that it is not government's job to attempt to provide public safety because that is not specified directly in the constitution. That is clearly wrong, and that's what I addressed with my post.
As I mentioned in my original post, if you want to question the manner in which the government is providing that service, i.e. that it is violating constitutional rights in doing so, that's a fine debate and argument to have. But that's not what you did.
When you had to do this clean build, were you in a critical code-writing/code-debugging loop? That was my main point. How often do you upgrade all the libraries in your project necessitating a clean build AND you're in a critical build loop?