Even with switching power supplies, 12v is not optimal because of the losses in the diodes. Even Schottky diodes have a voltage drop of 0.3v or so.
I think it would be a good idea to standardize on something in the 40-50 volt range for the DC grid in the house, with some leeway for adjusting the actual charging voltage to what is convenient for the battery. A 42-volt electrical system (http://en.wikipedia.org/wiki/42-volt_electrical_system#Choice_of_voltage) comes to mind. Even if it did not really take off the first time around.
The Open Source AMD drivers tend to lag a few years behind, especially in performance, so the latest generation of cards (HD 7xxx and its rebrandings) is hardly better under Linux than the HD 6xxx series, despite being more capable hardware in theory. Really new models may not work at all or only in 2D.
Closed source AMD drivers still have an iffy reputation. Personally, I'd avoid them unless I really need the performance or a specific feature.
Sounds interesting, but might be too expensive in the short run and not as urgent. "Normal" system RAM tends to be larger than typical VRAM sizes but not as bandwidth-critical. Also, classical DIMM modules allow upgrading if necessary.
Where HBM as the combined system RAM looks interesting are consoles. The PS4 in particular already has - an APU based design from AMD - fast but expensive GDDR5 RAM - a fixed, non - upgradeable memory size
That looks like a scenario which is just waiting for HBM. Though probably not in the current generation, as one of the main advantages (even higher bandwidth) might be irrelevant to a console with fixed performance requirements.
There is more in the works (Zen CPU cores), and I hope that will work out too. I have some AMD shares as well, but I'd also love to see serious competition again.
On the CPU side, AMD is even more behind Intel than it is on the GPU side behind Nvidia. If Zen can fix that, it will be even more important than catching up to Nvidia.
And considering APUs, one major drawback in the last years used to be memory bandwidth. Where discrete cards do reasonably well with GDDR5, APUs really get crippled by having to share DDR3 between CPU and GPU. A stack of HBM memory for the graphics RAM should do wonders here;-)
If you make really obvious mistakes at work, and more than once, you might eventually be fired for bad performance.
Which is exactly what should happen in this case: Even with only a high school education, cops should be able to understand the difference between arresting someone for assault and using extortion to get a pro-police statement on Facebook. If they don't get that or simply ignore Baton Bob's rights, they are unfit for duty. Get rid of them.
For me, a Science Fiction movie or book should fall under Mohs: Physics Plus or a higher degree of consistency to be considered good. Which tends to disqualify Star Wars and most of Star Trek, and thus eliminates a lot of the available Science Fiction flicks.
One of the reasons I usually read books rather than watching movies.
Ending the separation of the market by country is exactly what getting rid of geoblocking is about.
This said, it will be an "internal" EU affair as far as I understand it. There will still be geoblocking for content that crosses the outer EU borders.
I think French Movie Studio will likely license their movie to Netflix Europe (having separate licenses per country will make no more sense) once the theater run is up. And Netflix Europe will "lend" it everywhere in Europe.
After all, at this point it is a choice between getting no more revenue or some revenue.
Not in Germany. There is no law in Germany that would automatically nullify the copyright on works of a criminal.
Actually, the state of Bavaria still uses its claim to the copyright of Hitler's "Mein Kampf" to block re-publishing of that work. I'm not entirely sure if that claim is solid, but so far it works for them.
but unless the diaries are in the public domain, isn't this pretty cut and dry? If the diaries are in private hands, they're in private hands and you need permission to use their contents.
AFAIK, legally it is pretty cut and dry.
The question is: Should it be or should there be changes in the legislation? And if you change legislation, where do you draw the line?
Because there is currently a LOT of material that might be interesting for researchers, but is only accessible if you pay a fee. As in, most material controlled by scientific publishers such as Elsevier. With fees of some 10 dollars per article, buying access to one or two articles is affordable enough, but if you want to review some dozen to get an overview of your field of research it becomes expensive. Depending on your budget, perhaps prohibitively expensive.
The current case is no different: The estate of Joseph Goebbels does not want to forbid the use of the diaries but they want royalties for the extracts. So there is some financial obstacle to research.
Personally, I'm in favor for easier access to scientific materials if they were created with public funds. Perhaps by putting a clause in the work contract of the researchers that they give their employers shared publishing rights. In case of some old Nazi diaries though, I'd prefer the simple approach of waiting another 8 1/2 months. After 2015, almost all of those will be in the public domain anyway:-)
It would also avoid accusations of anti-competitive behavior, as the requirement could in principle be satisfied by all vendors. I'm not sure if AMD can get into trouble at this point for anti-competitive behavior, given their shrunken market share, but it can't hurt to play it safe.
All that aside though, the submitter is wrong to characterize this spending as "pork". Pork is a localized project meant to benefit a specific representative's district, while this is clearly an expression of American foreign policy generally. We believe we have to be able to instantly thwart any threat conceivable at any time forever, even though that's impossible. These projects were specifically designed to stop a missile attack on US soil by North Korea, which is absurd
The claim that these systems were designed for stopping North Korean missiles are indeed absurd.
But North Korea is not that far from China. So the real objective was probably stopping Chinese missiles, while North Korea was just the pretext and bogeyman.
Don't underestimate the importance of the bargain bin. Recently the cheap stuff has become cheap enough to make it commercially interesting even without subsidies.
There is also a middle ground: Tests with pre-programmed equipment given to the patient for use at home for one night, as a screening test. If the results look suspicious, a night sleeping in a test facility follows.
I'm halfway through such a screening myself right now. The equipment included a pulse oxymeter, a microphone for detecting snoring and a programmable device that records the readings over one night. All of it probably lab grade or close enough for the screening.
Recently I discovered slightly oversize LED bulbs from Philips at my preferred hardware store, rated at 75 watts (equiv) or 100 watts. Cheap enough too, the 75 watts equivalent cost around 10 Euro, the 100 watts equivalent around 13 Euro.
I've already tried out the 75 watts equivalent in one of my lamps, and subjectively it is as bright as the 100 watt bulb in the identical lamp beside it. This may have to do with the fact that they emit their light over a hemisphere, part of the light that goes into the rear half of the lamp is lost with the incandescent bulbs. The LEDs mostly avoid that. Still, I find it pretty impressive.
The 100 watts equivalent has almost 50% more flux in its specs. In my little apartment, I expect it to come across like a floodlight.
Oh, finding a hacked copy is quite possible. Even if Microsoft try to make it difficult;-)
The real problem will come after January 14, 2020 when the extended support ends. Then running Windows 7 with net access will become increasingly risky, because no more patches.
Using a Samsung SyncMaster 204B here. In dim lighting conditions.
Looking at the picture from a "normal" angle (approximately from the front and with the screen slightly below eye level), the dress looks black and blue to me. But if I look down on the screen at an angle of maybe 70 degrees, the blue becomes sort of white and the black lace gets a bit of a golden hue. Still not clearly white and gold, but I see how one could get the idea.
Now the SyncMaster 204B is seven years old and probably not quite state of the art anymore. And it took me a rather unusual viewing angle to get anywhere near "white and gold". The people who saw the dress as white and gold must have pretty shitty screens;)
There is at least one game that may be of interest to me (Planetside 2), but before I touch that one, the former Sony Online Entertainment has to distance itself from its past.
Selling the company is a good start, but not sufficient. Once they have gotten rid of the old management too, I may take a chance on them. But not before.
Following the link to the study leads to this summary (excerpt):
Bioenergy is an inefficient use of land to generate energy.
Fast-growing sugarcane on highly fertile land in the tropics converts only around 0.5 percent of solar radiation into sugar, and only around 0.2 percent ultimately into ethanol. For maize ethanol grown in Iowa, the figures are around 0.3 percent into biomass and 0.15 percent into ethanol. Such low conversion efficiencies explain why it takes a large amount of productive land to yield a small amount of bioenergy, and why bioenergy can so greatly increase global competition for land.
It seems the study did not even consider any new approaches to making biofuels.
http://en.wikipedia.org/wiki/Second-generation_biofuels promise the ability to use material that would otherwise be waste, such as straw, thus lessening the competition between food and fuel. Any study that claims to make forecasts for the year 2050 (also in TFA) should take a serious look at these too.
The study in TFA only gives a cursory overview over second generation biofuels with an either-crop-or-cellulose point of view. It almost seems that the option of using crop residue was intentionally neglected...
Seconded, because it matches the usage in mathematics. I don't think it makes a difference once you are used to the C syntax, but for someone new to programming the consistency with math textbooks is nice:-)
Maybe more importantly, AutoCAD licenses cost over 6000 Euros (AutoCAD Design Suite Standard 2015). For that much money, it becomes worthwhile to sue in court. While buyers of a 30 Euro game might just take the loss rather than sue.
Of course, that does not apply to outright stolen keys. But I consider it absolutely plausible that Ubisoft is making bogus claims about the "unauthorized" part and relies on people not suing over 50 pounds.
My personal solution is not buying games from publishers who require digital keys or other forms of DRM. Which means I have mostly older titles where the publisher has given up on DRM (bargain bin games often come "unlocked" so there is no more cost in maintaining the DRM). Tough luck for Ubisoft, EA and Valve;-)
To me, the critical part is noticing the difference in (for instance) a code review, not so much the understanding.
I'm assuming a moderately skilled programmer here, with enough brains to see that something is different and look it up in the online help. That guy would likely see the difference between 22 div 3 and 22/3, look it up and ultimately get it right.
While reliably seeing the difference between 22/3 and 22/3.0 almost requires someone who got burned before and has learned to look specifically for these differences. IMHO a higher level of experience...
What subset of "modern Pascal" do you have to restrict yourself to avoid those "problems"?
In practice, I guess you'd have to choose between Embarcadero's Delphi and Free Pascal.
AFAIK Delphi is the only platform that still has significant commercial usage, but too expensive for hobbyists. Free Pascal is probably the most popular open source Pascal variety, and the one I know of that seems to be actively maintained.
I think the rest of Pascal is thoroughly irrelevant these days;-)
Slashdot Mirror
Even with switching power supplies, 12v is not optimal because of the losses in the diodes. Even Schottky diodes have a voltage drop of 0.3v or so.
I think it would be a good idea to standardize on something in the 40-50 volt range for the DC grid in the house, with some leeway for adjusting the actual charging voltage to what is convenient for the battery.
A 42-volt electrical system (http://en.wikipedia.org/wiki/42-volt_electrical_system#Choice_of_voltage) comes to mind. Even if it did not really take off the first time around.
The Open Source AMD drivers tend to lag a few years behind, especially in performance, so the latest generation of cards (HD 7xxx and its rebrandings) is hardly better under Linux than the HD 6xxx series, despite being more capable hardware in theory. Really new models may not work at all or only in 2D.
Closed source AMD drivers still have an iffy reputation. Personally, I'd avoid them unless I really need the performance or a specific feature.
Sounds interesting, but might be too expensive in the short run and not as urgent. "Normal" system RAM tends to be larger than typical VRAM sizes but not as bandwidth-critical. Also, classical DIMM modules allow upgrading if necessary.
Where HBM as the combined system RAM looks interesting are consoles. The PS4 in particular already has
- an APU based design from AMD
- fast but expensive GDDR5 RAM
- a fixed, non - upgradeable memory size
That looks like a scenario which is just waiting for HBM. Though probably not in the current generation, as one of the main advantages (even higher bandwidth) might be irrelevant to a console with fixed performance requirements.
Well, I'm arguing for "only" sacking them. Unless they actually commited crimes, in which case any other employee would be financially liable too.
There is more in the works (Zen CPU cores), and I hope that will work out too. I have some AMD shares as well, but I'd also love to see serious competition again.
On the CPU side, AMD is even more behind Intel than it is on the GPU side behind Nvidia. If Zen can fix that, it will be even more important than catching up to Nvidia.
And considering APUs, one major drawback in the last years used to be memory bandwidth. Where discrete cards do reasonably well with GDDR5, APUs really get crippled by having to share DDR3 between CPU and GPU. A stack of HBM memory for the graphics RAM should do wonders here ;-)
If you make really obvious mistakes at work, and more than once, you might eventually be fired for bad performance.
Which is exactly what should happen in this case:
Even with only a high school education, cops should be able to understand the difference between arresting someone for assault and using extortion to get a pro-police statement on Facebook. If they don't get that or simply ignore Baton Bob's rights, they are unfit for duty. Get rid of them.
You, your wife and me are part of a minority. Some people just don't enjoy movies with threadbare logic, even if "everyone else" loves them.
See also the "Mohs Scale Of Science Fiction Hardness" (http://tvtropes.org/pmwiki/pmwiki.php/Main/MohsScaleOfScienceFictionHardness).
For me, a Science Fiction movie or book should fall under Mohs: Physics Plus or a higher degree of consistency to be considered good. Which tends to disqualify Star Wars and most of Star Trek, and thus eliminates a lot of the available Science Fiction flicks.
One of the reasons I usually read books rather than watching movies.
Ending the separation of the market by country is exactly what getting rid of geoblocking is about.
This said, it will be an "internal" EU affair as far as I understand it. There will still be geoblocking for content that crosses the outer EU borders.
I think French Movie Studio will likely license their movie to Netflix Europe (having separate licenses per country will make no more sense) once the theater run is up. And Netflix Europe will "lend" it everywhere in Europe.
After all, at this point it is a choice between getting no more revenue or some revenue.
Not in Germany. There is no law in Germany that would automatically nullify the copyright on works of a criminal.
Actually, the state of Bavaria still uses its claim to the copyright of Hitler's "Mein Kampf" to block re-publishing of that work. I'm not entirely sure if that claim is solid, but so far it works for them.
but unless the diaries are in the public domain, isn't this pretty cut and dry? If the diaries are in private hands, they're in private hands and you need permission to use their contents.
AFAIK, legally it is pretty cut and dry.
The question is:
Should it be or should there be changes in the legislation? And if you change legislation, where do you draw the line?
Because there is currently a LOT of material that might be interesting for researchers, but is only accessible if you pay a fee. As in, most material controlled by scientific publishers such as Elsevier.
With fees of some 10 dollars per article, buying access to one or two articles is affordable enough, but if you want to review some dozen to get an overview of your field of research it becomes expensive. Depending on your budget, perhaps prohibitively expensive.
The current case is no different:
The estate of Joseph Goebbels does not want to forbid the use of the diaries but they want royalties for the extracts. So there is some financial obstacle to research.
Personally, I'm in favor for easier access to scientific materials if they were created with public funds. Perhaps by putting a clause in the work contract of the researchers that they give their employers shared publishing rights. :-)
In case of some old Nazi diaries though, I'd prefer the simple approach of waiting another 8 1/2 months. After 2015, almost all of those will be in the public domain anyway
This.
It would also avoid accusations of anti-competitive behavior, as the requirement could in principle be satisfied by all vendors. I'm not sure if AMD can get into trouble at this point for anti-competitive behavior, given their shrunken market share, but it can't hurt to play it safe.
All that aside though, the submitter is wrong to characterize this spending as "pork". Pork is a localized project meant to benefit a specific representative's district, while this is clearly an expression of American foreign policy generally. We believe we have to be able to instantly thwart any threat conceivable at any time forever, even though that's impossible. These projects were specifically designed to stop a missile attack on US soil by North Korea, which is absurd
The claim that these systems were designed for stopping North Korean missiles are indeed absurd.
But North Korea is not that far from China. So the real objective was probably stopping Chinese missiles, while North Korea was just the pretext and bogeyman.
Don't underestimate the importance of the bargain bin. Recently the cheap stuff has become cheap enough to make it commercially interesting even without subsidies.
According to Wikipedia, we already have grid parity in many scenarios: http://en.wikipedia.org/wiki/Grid_parity#Reaching_parity.
There is also a middle ground:
Tests with pre-programmed equipment given to the patient for use at home for one night, as a screening test. If the results look suspicious, a night sleeping in a test facility follows.
I'm halfway through such a screening myself right now. The equipment included a pulse oxymeter, a microphone for detecting snoring and a programmable device that records the readings over one night. All of it probably lab grade or close enough for the screening.
Recently I discovered slightly oversize LED bulbs from Philips at my preferred hardware store, rated at 75 watts (equiv) or 100 watts. Cheap enough too, the 75 watts equivalent cost around 10 Euro, the 100 watts equivalent around 13 Euro.
I've already tried out the 75 watts equivalent in one of my lamps, and subjectively it is as bright as the 100 watt bulb in the identical lamp beside it. This may have to do with the fact that they emit their light over a hemisphere, part of the light that goes into the rear half of the lamp is lost with the incandescent bulbs. The LEDs mostly avoid that. Still, I find it pretty impressive.
The 100 watts equivalent has almost 50% more flux in its specs. In my little apartment, I expect it to come across like a floodlight.
Oh, finding a hacked copy is quite possible. Even if Microsoft try to make it difficult ;-)
The real problem will come after January 14, 2020 when the extended support ends. Then running Windows 7 with net access will become increasingly risky, because no more patches.
Using a Samsung SyncMaster 204B here. In dim lighting conditions.
Looking at the picture from a "normal" angle (approximately from the front and with the screen slightly below eye level), the dress looks black and blue to me. But if I look down on the screen at an angle of maybe 70 degrees, the blue becomes sort of white and the black lace gets a bit of a golden hue. Still not clearly white and gold, but I see how one could get the idea.
Now the SyncMaster 204B is seven years old and probably not quite state of the art anymore. And it took me a rather unusual viewing angle to get anywhere near "white and gold". The people who saw the dress as white and gold must have pretty shitty screens ;)
Agreed.
There is at least one game that may be of interest to me (Planetside 2), but before I touch that one, the former Sony Online Entertainment has to distance itself from its past.
Selling the company is a good start, but not sufficient. Once they have gotten rid of the old management too, I may take a chance on them. But not before.
Following the link to the study leads to this summary (excerpt):
Bioenergy is an inefficient use of land to generate energy.
Fast-growing sugarcane on highly fertile land in the tropics converts only around 0.5 percent of solar radiation into sugar, and only around 0.2 percent ultimately into ethanol. For maize ethanol grown in Iowa, the figures are around 0.3 percent into biomass and 0.15 percent into ethanol. Such low conversion efficiencies explain why it takes a large amount of productive land to yield a small amount of bioenergy, and why bioenergy can so greatly increase global competition for land.
It seems the study did not even consider any new approaches to making biofuels.
http://en.wikipedia.org/wiki/Second-generation_biofuels promise the ability to use material that would otherwise be waste, such as straw, thus lessening the competition between food and fuel. Any study that claims to make forecasts for the year 2050 (also in TFA) should take a serious look at these too.
The study in TFA only gives a cursory overview over second generation biofuels with an either-crop-or-cellulose point of view. It almost seems that the option of using crop residue was intentionally neglected...
Seconded, because it matches the usage in mathematics. :-)
I don't think it makes a difference once you are used to the C syntax, but for someone new to programming the consistency with math textbooks is nice
Maybe more importantly, AutoCAD licenses cost over 6000 Euros (AutoCAD Design Suite Standard 2015). For that much money, it becomes worthwhile to sue in court. While buyers of a 30 Euro game might just take the loss rather than sue.
Depending on where the buyer got the "unauthorized" key, exploiting differences in market prices may actually be legal. Compare http://www.olswang.com/articles/2012/09/2013/04/exhaustion-of-rights-in-the-download-to-own-market/. And I doubt that many people buy their keys outside the EU.
Of course, that does not apply to outright stolen keys. But I consider it absolutely plausible that Ubisoft is making bogus claims about the "unauthorized" part and relies on people not suing over 50 pounds.
My personal solution is not buying games from publishers who require digital keys or other forms of DRM. Which means I have mostly older titles where the publisher has given up on DRM (bargain bin games often come "unlocked" so there is no more cost in maintaining the DRM). Tough luck for Ubisoft, EA and Valve ;-)
To me, the critical part is noticing the difference in (for instance) a code review, not so much the understanding.
I'm assuming a moderately skilled programmer here, with enough brains to see that something is different and look it up in the online help. That guy would likely see the difference between 22 div 3 and 22/3, look it up and ultimately get it right.
While reliably seeing the difference between 22/3 and 22/3.0 almost requires someone who got burned before and has learned to look specifically for these differences. IMHO a higher level of experience...
What subset of "modern Pascal" do you have to restrict yourself to avoid those "problems"?
In practice, I guess you'd have to choose between Embarcadero's Delphi and Free Pascal.
AFAIK Delphi is the only platform that still has significant commercial usage, but too expensive for hobbyists.
Free Pascal is probably the most popular open source Pascal variety, and the one I know of that seems to be actively maintained.
I think the rest of Pascal is thoroughly irrelevant these days ;-)