It's slightly easier. Things like JITs are easier to port in the other direction (targeting ARM or Thumb-2[EE] from a typical bytecode is easier than targeting x86), but any high-level code just sees a relaxation in some restrictions. You have the same endian on both. Both incur a performance hit for unaligned loads and stores, and old ARM doesn't support them at all, so it's possible to do some pointer arithmetic and casting that will work on x86 and not ARM, but it's pretty hard and usually it will just be a bit slow on ARM (if the compiler knows it's unaligned it can work around it with some two loads and mask + shift).
Even after Microsoft dropped support for non-Intel architectures with Windows 2000
They dropped support for non-Intel architectures, but not for non-x86 architectures. Windows Server 2008 was the last version to have Itanium support, and it shares more or less the same kernel as the consumer editions (or, rather, they're all branched from the same development tree) so processor independence never really went away in Windows NT. ARM is a lot more similar to x86 than Itanium, in terms of parts of the machine model that are exposed in C/C++ level. It's pretty hard to write C/C++ code that will work on x86, x86-64 and Itanium, but not ARM.
If corporations weren't people you couldn't take them to court and sue them when they behaved poorly.
No, you'd have to take the officers of the corporation to court and sue them personally, just as you do with small businesses. With that level of personal responsibility, CEOs might actually deserve their massive salaries.
It's also worth noting that companies that do try to impose this kind of limit have a habit of dying. nVidia should know - they killed one of them. SGI wanted to keep cheap GPUs underpowered to keep their margins high at the top end. nVidia came along and destroyed the top end market by making GPUs that were almost as good at a tiny fraction of the price.
The Psion Series 3/3a/3c had similar specs to the HP palmtops, but they weren't crippled by running DOS. EPOC16 (which later evolved into Symbian) ran a multitasking GUI in 256KB of RAM, which was also used as a RAM disk. My 3 had a spreadsheet on an SSD, but the 3A and later came with it built in. There's a DOS-based 3A emulator that runs nicely in DOSBox floating around.
and all a serious writer ever really needs is a text editor to do real work
Well, that and the ability to store reasonable sized text files. A typical article can be 10-15KB of plain text. The 32KB of RAM in the TRS-80 doesn't leave much left over for the text editor. Having a spell check is also convenient, and that typically requires 1MB or so for the data.
We gave up on that game when we discovered that we didn't actually need lots of peasants. Now we just install corrupt governments and bribe them to let us take all of the natural resources. It's much cheaper and doesn't leave us with embarrassing colonies that we need to maintain.
Look at the population figures excluding immigration. For most of the EU, population growth is slightly negative if you discount immigration, which indicates that people are not breeding at below replacement rate.
The truth is that projects aren't jumping ships. No GPL projects are trying to change their license.
Projects aren't relicensing (it's really hard!), but GPL'd projects are seeing competition from more permissively licensed alternatives. It looks like in FreeBSD 10 we'll be able to replace all of the GPL'd components of the base system with permissively licensed alternatives. Doing that five years ago would have been impossible.
It's possible, but to make a concrete example, as a user, the only reason I could upgrade my phone firmware after its manufacturer stopped supporting it is that Android's kernel is under the GPL
The only reason I can upgrade my phone firmware after the manufacturer stopped supporting it is that the manufacturer (HTC) chose to release an updated (but unsupported) version of the blob that controls the GPU and a few other things to the community. As far as I know, there are no Android phones that have a completely open stack, or even a completely open source set of drivers.
This sounds like more of a reason to embrace community-friendly manufacturers than any particular license. The GPL in Linux hasn't prevented all Android phones from coming with binary-only drivers (if there are exceptions to this, please point them out to me), but being able to get long-term software support for their products (and therefore goodwill from users) without having to pay anything has made hardware manufacturers a lot more willing to cooperate willing to cooperates with groups like cyanogenmod.
Apple isn't the only company. The patent clauses and termination clauses in GPLv3 make a lot of companies nervous. There is a lot of FUD in the classic sense surrounding the GPLv3, but the most important is the U: uncertainty. When legal says 'we're 80% sure that there will be no problems with distributing this code in our products,' management hears 'there is a 20% chance that this will be really expensive' and opts for a more permissively license alternative.
The GPL is about the user's freedom not the developer's
Is the user more free because he can't distribute a plugin that uses an LGPLv3 library with a program that uses GPLv2 while exercising the FSF's Freedom 2 (The freedom to redistribute copies so you can help your neighbor)? Is the user more free because he can't combine libraries using either the Apple Public Source license, the Mozilla Public License, the Common Development and Distribution License or the Apache Software Foundation License (all FSF-approved Free Software licenses) with a GPL'd application?
It guarantees that the user can modify the source in any way they see fit without the developer locking them out now or later.
Any Free or Open Source license guarantees this. If I have some MIT licensed code, for example, I have an irrevocable license (unlike the GPL, by the way, which has some revocation terms built into the license, and more aggressive ones in GPLv3) that allow me to continue to use it, distribute it, and distribute derived works of it in perpetuity. The only thing that the GPL prevents is someone other than the copyright holder from creating a new version and distributing it without granting these same rights (the GPL does not prevent the copyright holder from taking something closed source). In either case, the last open version is likely to be forked if enough people care about it.
It's misleading because the neither the Free Software Foundation's list of four freedoms required for something to be Free Software, nor the Open Source Definition requires something to be copyleft. I think there are some corner cases where something can be covered by one definition but not the other, but I've never seen one in the wild: projects tend to be either proprietary or both free software and open source. What the title meant was that open source project licenses are tending towards permissive, rather than copyleft.
Why? One of the characteristics shared by a lot of successful people is that they failed a lot before succeeding. They were the ones willing to back the crazy-sounding ideas. A lot of those ideas turned out not to be just crazy sounding, but actually crazy and failed, but the crazy idea that actually worked had a massive return on investment. This doesn't mean that the next crazy idea that they pick will work as well...
Plutonium is rare because it is the heaviest vaguely stable element (meaning that only trace amounts were created in the star exploded to create the elements that form our solar system and because its most stable form has a half life of 80 million years. The latter means that around 57 plutonium half lives have elapsed since the sun formed. That means that about 7x10^-16% of what was formed initially still exists. To put that in perspective, if an amount of plutonium equal to the total biomass on the Earth had been formed, there would be 1g left.
Or, to put it another way, if there had originally been a quantity of plutonium equal to the mass of jupiter formed, then there would now be somewhere around double the mass of the great pyramid in Giza left, scattered all over the solar system. That amount is a pretty optimistic estimate, especially if you exclude any that ended up in the Sun as irrelevant.
In fact, if you assume that all of the matter in the solar system except the Sun was originally plutonium, then that still only gives you three times the mass of the great pyramid in Giza (about 1.8x10^7 metric tons) of plutonium left, scattered all over the solar system. Imagine if you took the great pyramid, ground it up, and scattered it just over the Earth's surface - even if it had the energy density of antimatter it probably wouldn't be worthwhile to find and collect it. If it's scattered all over the solar system (meaning most of it will be inside large masses, and most of it inside Jupiter), it's not going to be even remotely energy positive to find it.
Or, for the TLDR version, even assuming that there is vastly more plutonium around than there is actually likely to be, it's still not even remotely worth harvesting.
In future, please try not to assume that just because people have an understanding of the workings of science and the limitations of (current) technology, and don't treat it as a magical solution to all possible problems, that they're anti-technology.
If we don't get off this rock, we will follow the path of the dinosaurs or even worse. Whether by interplanetary impact or a hugely destructive solar storm or a disastrous disease mutation etc.
Eventually, yes. The question is not whether we can survive on this planet forever, it's whether we can survive on it for a couple of hundred years more. Given that extinction events happen every few million years, the odds are that we probably can.
Proponents of space exploration are always keen to point out the spin offs from space research, but they tend to miss things going the other way. If we started the Apollo program from scratch now it would be a lot easier than in the '60s purely due to our improvements in simulation (from processors many orders of magnitude faster) and in materials science. Give us another twenty years of nanomaterials research - a field still very much in its infancy today - and we'll likely see some even bigger improvements. A thin and light material that's almost impervious to radiation, for example, would be a huge boon for manned space travel. The magnetic accelerators used in modern ion drives were developed for quite different reasons.
In two hundred years, we're likely to have technology that makes space travel very easy whether we invest in space travel now or not. Or we're going to have rendered the Earth completely uninhabitable, in which case any space colonies we might have created probably wouldn't last more than a decade, if that. And the latter option seems a lot more likely if we're devoting funding and mindshare to the idea that we can just break this planet and move to a better one...
That's more to do with the fact that something else has already made a big hole in the ground for free than the relatively small lump of extraterrestrial rock in the middle...
It's a question of yield. The smaller the pixels, the lower the tolerance to fabrication errors, and so the higher the defect rate. This is why you initially see higher resolutions in smaller screens for TFTs. If you can get the defect ratio down to one in a million, then you can probably get three working 800x480 screens for every one with a dead / stuck pixel. There are also lots of customers for control panels in industrial equipment where a stuck pixel doesn't really matter, so you can also sell the defective ones. Try scaling this up to a desktop screen and you've now got 3-4 stuck pixels in every 27" monitor you make, and no one will buy them.
Unfortunately, it's only an advantage to the person buying the screen, not the person selling it. There was an article on Slashdot a few years ago with a test in a shop with glossy and non-glossy screens on otherwise identical laptops. The glossy one outsold the matte one about 2:1, yet when interviewed a month later the people buying the matte one were all happier with their purchase. In a bright shop environment, the glossy screens look better, it's only in normal use where they're worse. This is especially relevant for a company like Apple that sells a lot of computers and tablets based on demo units in stores...
Are you an ODM? If not, then you're complaining to the wrong people. Pixel Qi has been shipping devices to anyone who places an order with them, including a couple of large consumer device manufacturers and has even been selling replacement netbook screens to consumers for a couple of years. I've not owned one, but I played with one at FOSDEM a year ago and they seemed pretty nice. Contrast isn't great in the low power mode, but it's a lot more visible in direct sunlight than any laptop screen I've ever owned...
Slashdot Mirror
It's slightly easier. Things like JITs are easier to port in the other direction (targeting ARM or Thumb-2[EE] from a typical bytecode is easier than targeting x86), but any high-level code just sees a relaxation in some restrictions. You have the same endian on both. Both incur a performance hit for unaligned loads and stores, and old ARM doesn't support them at all, so it's possible to do some pointer arithmetic and casting that will work on x86 and not ARM, but it's pretty hard and usually it will just be a bit slow on ARM (if the compiler knows it's unaligned it can work around it with some two loads and mask + shift).
Even after Microsoft dropped support for non-Intel architectures with Windows 2000
They dropped support for non-Intel architectures, but not for non-x86 architectures. Windows Server 2008 was the last version to have Itanium support, and it shares more or less the same kernel as the consumer editions (or, rather, they're all branched from the same development tree) so processor independence never really went away in Windows NT. ARM is a lot more similar to x86 than Itanium, in terms of parts of the machine model that are exposed in C/C++ level. It's pretty hard to write C/C++ code that will work on x86, x86-64 and Itanium, but not ARM.
I quite a few tablets on the train recently, but the only one I got close enough to to read the the make was an Asus tablet of some kind.
If corporations weren't people you couldn't take them to court and sue them when they behaved poorly.
No, you'd have to take the officers of the corporation to court and sue them personally, just as you do with small businesses. With that level of personal responsibility, CEOs might actually deserve their massive salaries.
It's also worth noting that companies that do try to impose this kind of limit have a habit of dying. nVidia should know - they killed one of them. SGI wanted to keep cheap GPUs underpowered to keep their margins high at the top end. nVidia came along and destroyed the top end market by making GPUs that were almost as good at a tiny fraction of the price.
Hopefully now, after a string of modded down posts, you have learned that feeding trolls does not actually make them go away.
Seriously? This is the most evil thing Apple or Google has done? Let me guess, you're looking for a job in Silicon Valley...
The Psion Series 3/3a/3c had similar specs to the HP palmtops, but they weren't crippled by running DOS. EPOC16 (which later evolved into Symbian) ran a multitasking GUI in 256KB of RAM, which was also used as a RAM disk. My 3 had a spreadsheet on an SSD, but the 3A and later came with it built in. There's a DOS-based 3A emulator that runs nicely in DOSBox floating around.
and all a serious writer ever really needs is a text editor to do real work
Well, that and the ability to store reasonable sized text files. A typical article can be 10-15KB of plain text. The 32KB of RAM in the TRS-80 doesn't leave much left over for the text editor. Having a spell check is also convenient, and that typically requires 1MB or so for the data.
We gave up on that game when we discovered that we didn't actually need lots of peasants. Now we just install corrupt governments and bribe them to let us take all of the natural resources. It's much cheaper and doesn't leave us with embarrassing colonies that we need to maintain.
Look at the population figures excluding immigration. For most of the EU, population growth is slightly negative if you discount immigration, which indicates that people are not breeding at below replacement rate.
The truth is that projects aren't jumping ships. No GPL projects are trying to change their license.
Projects aren't relicensing (it's really hard!), but GPL'd projects are seeing competition from more permissively licensed alternatives. It looks like in FreeBSD 10 we'll be able to replace all of the GPL'd components of the base system with permissively licensed alternatives. Doing that five years ago would have been impossible.
It's possible, but to make a concrete example, as a user, the only reason I could upgrade my phone firmware after its manufacturer stopped supporting it is that Android's kernel is under the GPL
The only reason I can upgrade my phone firmware after the manufacturer stopped supporting it is that the manufacturer (HTC) chose to release an updated (but unsupported) version of the blob that controls the GPU and a few other things to the community. As far as I know, there are no Android phones that have a completely open stack, or even a completely open source set of drivers.
This sounds like more of a reason to embrace community-friendly manufacturers than any particular license. The GPL in Linux hasn't prevented all Android phones from coming with binary-only drivers (if there are exceptions to this, please point them out to me), but being able to get long-term software support for their products (and therefore goodwill from users) without having to pay anything has made hardware manufacturers a lot more willing to cooperate willing to cooperates with groups like cyanogenmod.
Apple isn't the only company. The patent clauses and termination clauses in GPLv3 make a lot of companies nervous. There is a lot of FUD in the classic sense surrounding the GPLv3, but the most important is the U: uncertainty. When legal says 'we're 80% sure that there will be no problems with distributing this code in our products,' management hears 'there is a 20% chance that this will be really expensive' and opts for a more permissively license alternative.
The GPL is about the user's freedom not the developer's
Is the user more free because he can't distribute a plugin that uses an LGPLv3 library with a program that uses GPLv2 while exercising the FSF's Freedom 2 (The freedom to redistribute copies so you can help your neighbor)? Is the user more free because he can't combine libraries using either the Apple Public Source license, the Mozilla Public License, the Common Development and Distribution License or the Apache Software Foundation License (all FSF-approved Free Software licenses) with a GPL'd application?
It guarantees that the user can modify the source in any way they see fit without the developer locking them out now or later.
Any Free or Open Source license guarantees this. If I have some MIT licensed code, for example, I have an irrevocable license (unlike the GPL, by the way, which has some revocation terms built into the license, and more aggressive ones in GPLv3) that allow me to continue to use it, distribute it, and distribute derived works of it in perpetuity. The only thing that the GPL prevents is someone other than the copyright holder from creating a new version and distributing it without granting these same rights (the GPL does not prevent the copyright holder from taking something closed source). In either case, the last open version is likely to be forked if enough people care about it.
It's misleading because the neither the Free Software Foundation's list of four freedoms required for something to be Free Software, nor the Open Source Definition requires something to be copyleft. I think there are some corner cases where something can be covered by one definition but not the other, but I've never seen one in the wild: projects tend to be either proprietary or both free software and open source. What the title meant was that open source project licenses are tending towards permissive, rather than copyleft.
Why? One of the characteristics shared by a lot of successful people is that they failed a lot before succeeding. They were the ones willing to back the crazy-sounding ideas. A lot of those ideas turned out not to be just crazy sounding, but actually crazy and failed, but the crazy idea that actually worked had a massive return on investment. This doesn't mean that the next crazy idea that they pick will work as well...
Lunchtime doubly so.
Or, to put it another way, if there had originally been a quantity of plutonium equal to the mass of jupiter formed, then there would now be somewhere around double the mass of the great pyramid in Giza left, scattered all over the solar system. That amount is a pretty optimistic estimate, especially if you exclude any that ended up in the Sun as irrelevant.
In fact, if you assume that all of the matter in the solar system except the Sun was originally plutonium, then that still only gives you three times the mass of the great pyramid in Giza (about 1.8x10^7 metric tons) of plutonium left, scattered all over the solar system. Imagine if you took the great pyramid, ground it up, and scattered it just over the Earth's surface - even if it had the energy density of antimatter it probably wouldn't be worthwhile to find and collect it. If it's scattered all over the solar system (meaning most of it will be inside large masses, and most of it inside Jupiter), it's not going to be even remotely energy positive to find it.
Or, for the TLDR version, even assuming that there is vastly more plutonium around than there is actually likely to be, it's still not even remotely worth harvesting.
In future, please try not to assume that just because people have an understanding of the workings of science and the limitations of (current) technology, and don't treat it as a magical solution to all possible problems, that they're anti-technology.
If we don't get off this rock, we will follow the path of the dinosaurs or even worse. Whether by interplanetary impact or a hugely destructive solar storm or a disastrous disease mutation etc.
Eventually, yes. The question is not whether we can survive on this planet forever, it's whether we can survive on it for a couple of hundred years more. Given that extinction events happen every few million years, the odds are that we probably can.
Proponents of space exploration are always keen to point out the spin offs from space research, but they tend to miss things going the other way. If we started the Apollo program from scratch now it would be a lot easier than in the '60s purely due to our improvements in simulation (from processors many orders of magnitude faster) and in materials science. Give us another twenty years of nanomaterials research - a field still very much in its infancy today - and we'll likely see some even bigger improvements. A thin and light material that's almost impervious to radiation, for example, would be a huge boon for manned space travel. The magnetic accelerators used in modern ion drives were developed for quite different reasons.
In two hundred years, we're likely to have technology that makes space travel very easy whether we invest in space travel now or not. Or we're going to have rendered the Earth completely uninhabitable, in which case any space colonies we might have created probably wouldn't last more than a decade, if that. And the latter option seems a lot more likely if we're devoting funding and mindshare to the idea that we can just break this planet and move to a better one...
That's more to do with the fact that something else has already made a big hole in the ground for free than the relatively small lump of extraterrestrial rock in the middle...
It's a question of yield. The smaller the pixels, the lower the tolerance to fabrication errors, and so the higher the defect rate. This is why you initially see higher resolutions in smaller screens for TFTs. If you can get the defect ratio down to one in a million, then you can probably get three working 800x480 screens for every one with a dead / stuck pixel. There are also lots of customers for control panels in industrial equipment where a stuck pixel doesn't really matter, so you can also sell the defective ones. Try scaling this up to a desktop screen and you've now got 3-4 stuck pixels in every 27" monitor you make, and no one will buy them.
Not glossy is a huge advantage if you ask me.
Unfortunately, it's only an advantage to the person buying the screen, not the person selling it. There was an article on Slashdot a few years ago with a test in a shop with glossy and non-glossy screens on otherwise identical laptops. The glossy one outsold the matte one about 2:1, yet when interviewed a month later the people buying the matte one were all happier with their purchase. In a bright shop environment, the glossy screens look better, it's only in normal use where they're worse. This is especially relevant for a company like Apple that sells a lot of computers and tablets based on demo units in stores...
Apple is outselling everyone in the tablet market, but in the (much bigger) laptop market they're still only at around 10%.
Are you an ODM? If not, then you're complaining to the wrong people. Pixel Qi has been shipping devices to anyone who places an order with them, including a couple of large consumer device manufacturers and has even been selling replacement netbook screens to consumers for a couple of years. I've not owned one, but I played with one at FOSDEM a year ago and they seemed pretty nice. Contrast isn't great in the low power mode, but it's a lot more visible in direct sunlight than any laptop screen I've ever owned...