Right now, it is precisely the lawless nature of the Internet (in that it is unreasonably difficult to enforce accepted laws there, even when pretty much everyone agrees they are reasonable laws) that leads to problems like spam, defamation, phishing expeditions, and all the other bad stuff that I'm sure everyone except those benefiting personally could happily live without.
It is *precisely* the "lawless" state of the internet today that makes it useful as a tool for freedom (and flexible as a basis for building things).
Spam is a technical problem with the design of the SMTP protocol, and a really interesting social issue re: the appropriateness of push marketing in any medium designed for 1 to 1 personal communication. But, rather than trying to fix technical problems with laws, let's let SMTP as it is continue to die it's slow death.
Defamation is nothing new to the internet. You could always distribute anonymous pamphlets about people. Sure, more people can participate in both reading and writing, but the effect will go down as more people realize that talk is cheap. More importantly, Defamation is in no way an important enough issue to consider restraining the essential liberty that is freedom of communication.
Phishing and other scams are no more interesting to me than pickpockets in open air markets (where that sort of thing is common). Sure, it sucks when you aren't prepared and lose your wallet - but all the locals will correctly just laugh at you and tell you to be more alert next time. There will always be people out to scam you / take your stuff - one of the key skills to operate in human society is to avoid being the victim. I give the pickpocket example for a very good reason - this isn't a new class of problem, it's been solved, and it isn't the government's responsibility to protect you from everything.
Why should I give holocaust deniers and nazi propagandist the right to be heard?
You covered the most common argument. The second most common argument is this: If you let them make everyone familiar with their arguments, have the public discussion, and show everyone that they're wrong then many will accept their very well developed and sophisticated arguments when they make them in private.
My argument is much simpler. You can't impose censorship without necessarily censoring the meta-discussion about that censorship. If I want to argue that some hate speech isn't socially harmful, I can't start by giving an example of the hate speech that I'm talking about. If you want to argue that the law in the US where 17 year olds can consent to sex but not to being the subject of sexual photographs is absurd, you can't respond to the argument "17 year olds look too young, nude pictures of them will make perverts want to rape babies" with the most powerful response - a photographic counterexample.
Information isn't dangerous, uninformed people are. The only reason to censor discussion is to defend political positions that can't hold up to the close examination they would get if discussion were legal - or occasionally as a political tool to distract people from relevant political issues by reminding them of people/opinions they don't like.
Sure, documenting their hardware means they give operational details away for free. That's not something they do now, so apparently it means something to them. You can argue that they are wrongheaded all you want, but realistically, they don't see that opening things up would make them more money.
AMD doesn't provide clean avenues for contact on this sort of thing. There are only three types of action I could take to get them to listen:
I could try to get in a position where I was a direct customer, and then tell them that no hardware documentation would be a deal breaker.
As a shareholder, I could contact investor relations. But, my guess is that investor relations are the same people who do end-user PR - i.e. they don't get to talk to anyone who could influence policy.
I can try to reinforce the general feeling that "ATI's closed policy is ruining AMD" on sites like Slashdot. The whole point of their press releases is to create news articles, and if the news articles on major sites all have comments that say "Great new toy, but it's worthless without documentation", that's a PR nightmare. PR nightmares tend to actually get fixed. And, realistically, programmable hardware without programming documentation is absurd - trying to do that should be a PR nightmare.
Maybe if they believed it were sensible for them to give stuff away for free, they would.
Documenting their hardware has nothing to do with "giving stuff away for free". Their video cards cost quite a bit of money, and AMD is losing sales every day to Intel because they won't release the docs to let people use their hardware. That's both video card *and* processor sales, because Intel graphics only come with Intel processors.
Releasing programming documentation for hardware is pretty normal. In fact, high end 3D graphics is the only area where the "binary blob" driver thing has ever really come to be seen as "the norm". No one else is foolish enough to think that they can sell hardware without letting their customers actually use it.mm
AMD has gotten where it is today by releasing standard, well documented computer hardware. No one would consider for *one second* developing a product based on a CPU that wasn't documented, and the same is true for video cards - especially with the trend towards GPGPUs.
ATI was in a market niche where they didn't have much competition and all they really needed to do was support Microsoft's closed Direct X standard. AMD has to look at the whole market, and if they let the "closed hardware" ATI policy get on their CPUs (and they're coming together with Fusion), companies will stop buying their stuff damn fast. Right now it would be extremely valuable to their future business to open up their current graphics cards so that system developers don't shy away from Fusion simply based on ATI's reputation.
If AMD doesn't stay open, then everyone who has been ignoring the high power embedded / SFF / custom systems market (which is easily as large as the PC market) will get really confused when AMD and VIA swap places and the main competition to Intel is the (well documented) Via C7 and Unichrome Pro Graphics.
The situation is not quite so rosy in the 3D graphics chipset arena, as the review of the Radeon HD 2900 XT a few days ago highlighted... The Tech Report had to upgrade their PC's PSU to 750W to achieve stability.
High end graphics cards have never been terribly friendly to power supplies. I see no reason to think the 2600 or 2400 cards - which have significantly fewer stream processors and are fabbed at 65nm instead of 80nm - will have unreasonable power draw.
The much more important issue for DAAMIT at the moment is to actually release programming info for their hardware (or at least release some Open Source drivers). Even *Intel* can get this right - they have no excuse whatsoever.
but those companies can't afford to pay 80k per year for a TEM operator
Bullshit. If they need them, they can pay them. If they can't afford to pay them, that means they'd rather not have them (i.e. they don't really need them).
the company pays him 100k per year, but he spend one third of the time smoking outside and enjoy the sunshine. Why? Because they have to measure their products in SEM to know if they are good.
If the company is losing $33,000/year on just your friend, they can easily afford to pay $80,000 instead of $50,000 for microscope operators.
The US economy overall also loses out. A legitimate labor shortage will naturally fix itself as more people enter that field for the shortage-induced high pay. Adding H1B workers just means that the supply *never* corrects, thus maintaining the shortage and providing an excuse to demand more H1B's.
Perhaps by... training more operators? When demand goes up, price goes up. That drives supply up to what's needed, and price comes back down. Providing artificial supply just prevents real supply from being created - in your example screwing Americans out of 2,000 skilled jobs in the medium-term.
If it had a standard library like Windows Forms, Swing, AppKit, etc then it would be very useful for writing GUI programs, but right now there is only Tk or various roughly supported bindings.
There are a ton of good GUI apps written in Python. They use such GUI toolkits as WxWidgets and Gtk. Python is already a dependency, so including a second dependency is basically trivial. On Linux, packages can have dependencies that are installed automatically. On Windows, there's a tool that lets you blob together the python interpreter and any libraries into one big Windows binary.
Remind me again how sudo is supposed to be safer than having a root account?
Users occasionally need to perform administrative tasks. This happens frequently enough that if they needed to log out and log back in as root to do them, they would just stay logged in as root all the time for convenience. That's obviously not secure.
So, that means that there needs to be a way to get administrative privileges temporarily. In order to have any security at all from this, it has to require user interaction (otherwise programs could automatically get root, and there would be no security). That leaves the choice between the Windows Vista "press OK to perform admin task" solution and the Mac OS X / Ubuntu "type in your password to perform admin task" solution. Making the user type in their password at least proves that it's not someone else sitting down at their computer, and makes sure the user notices that something special happened.
That's not "programming", that's "construction". Once all the models and environments are constructed, you still have to program them - using a traditional programming language.
Perhaps a solid purely-visual programing environment could be built. People would still need to know how to program - pretty pictures don't make the problem of understanding data structures and algorithms go away.
I don't really want Dell selling Ubuntu-equipped PC's. It will be bad for distro diversity.
How is one desktop distribution becoming more popular, especially a community-maintained distribution like Ubuntu, a bad thing? It'll make Linux more popular overall, and totally kill the "too many choices" FUD. It's not like you have to stop using whatever your favorite distro is - hell, if you buy a Dell with Ubuntu you know the hardware works with any Linux-based distro. At worst, people might start releasing proprietary software in.deb format only - but that already happens with RPM and doesn't cause much of a problem.
Wait... how is being able to *sketch* a more useful skill than knowing how to make full and efficient use of a computer? Every job - all of them - can benefit from computer automation. Even just knowing enough programming to know what's feasible for a real programmer to do is priceless.
It's not worth it to Microsoft to trade the benifits of being a monopoly away to avoid the costs of being a monopoly. A couple million a day in non-compliance fines is much cheaper than application developers porting to Linux because it has 25% market share would be - once that starts happening they're basically done.
Though the silicon may well clock 50-100% higher and blow away nVidia's 8800GTX, it turns out that it eats 600W PSUs for breakfast -- that's the real reason AMD couldn't release a high-end part: Few except for hardcore overclockers have 700W power supplies ready to feed this thing. That, and no one in their right mind wants a computer that uses 400+ watts idling on Microsoft Word. So AMD had to settle for the mid-market, with mediocre performance that's within an acceptable (albeit still very high) power envelope.
If they really release the 2950 XTX built on a 65nm process, that could solve the heat problems. The 2900 XT is good enough to prevent people from counting DAAMIT out of the high end graphics market entirely. Oh, and the 2600 and 2400 parts look like they could be really nice for power/performance - and those are what matter to most of the market anyway.
IBM's patent portfolio is sufficient to completely drain MSFT's entire operating profit, if they charged all the royalties that they're legally entitled to.
That'd be really funny. As the saying goes "Live by the sword... ZOMG! They've Got Nukes!"
Intel graphics are a bit slow for very recent games, but for the majority of Linux users who aren't hardcore gamers the Intel graphics are way, way better than anything that Nvidia or ATI offer. They provide both 2D and 3D acceleration, they don't have stupid bugs that the community can't fix, and they even work great for older games - Quake 3 and stuff like Wolfenstien: Enemy Territory should run great.
Let me point you to the complete lack of any actual Open Source drivers released by ATI as of this moment. ATI is known for promising a lot of stuff, but their "Commitment to Open Source" (first announced around 1999, repeated consistently since then) has resulted in nothing of any value at all.
Until they actually release Open Source 2D AND 3D drivers, or (even better) release programming docs for their hardware, a boycott is a damn good idea.
Be careful. If this wasn't ATI it might be OK to trust them, but ATI has had a "commitment to Open Source" since 1999, and as a result we have really, really bad proprietary drivers for Linux.
My response to this announcement is: I will try even harder to buy Intel graphics over ATI, until a source release happens. They will get *nothing* from me on an "announcement" this time - this isn't "trick me once" or "trick me twice", we're closer to "trick me 18 times" at this point...
Once they really release Open Source drivers, or (even better) full interface documentation for their cards, then I'll consider their products again. But, until they actually do the whole thing this just looks like more lies.
I've also purchased, with actual money, every one of the Linux game releases that I've been interested in over the past 8 years. That's only 14 major commercial game releases, but that's constrained by my selection rather than by my budget.
This year, I'm waiting impatiently for two major Linux game releases: ET: Quake Wars and UT3.
I do it all the time, and absolutely don't see any of the 2D performance differences you've listed.
This seems to strongly depend on exactly which card/driver combination you're using. Some of them work great with the Free drivers, and some of them work like crap. When they work fine, I mostly don't notice. The Radeon X1300 on one of my machines though - the Free drivers don't even do basic 2D acceleration, surfing the web is ass-slow because it can't redraw fast enough to scroll.
Slashdot Mirror
It is *precisely* the "lawless" state of the internet today that makes it useful as a tool for freedom (and flexible as a basis for building things).
Spam is a technical problem with the design of the SMTP protocol, and a really interesting social issue re: the appropriateness of push marketing in any medium designed for 1 to 1 personal communication. But, rather than trying to fix technical problems with laws, let's let SMTP as it is continue to die it's slow death.
Defamation is nothing new to the internet. You could always distribute anonymous pamphlets about people. Sure, more people can participate in both reading and writing, but the effect will go down as more people realize that talk is cheap. More importantly, Defamation is in no way an important enough issue to consider restraining the essential liberty that is freedom of communication.
Phishing and other scams are no more interesting to me than pickpockets in open air markets (where that sort of thing is common). Sure, it sucks when you aren't prepared and lose your wallet - but all the locals will correctly just laugh at you and tell you to be more alert next time. There will always be people out to scam you / take your stuff - one of the key skills to operate in human society is to avoid being the victim. I give the pickpocket example for a very good reason - this isn't a new class of problem, it's been solved, and it isn't the government's responsibility to protect you from everything.
You covered the most common argument. The second most common argument is this: If you let them make everyone familiar with their arguments, have the public discussion, and show everyone that they're wrong then many will accept their very well developed and sophisticated arguments when they make them in private.
My argument is much simpler. You can't impose censorship without necessarily censoring the meta-discussion about that censorship. If I want to argue that some hate speech isn't socially harmful, I can't start by giving an example of the hate speech that I'm talking about. If you want to argue that the law in the US where 17 year olds can consent to sex but not to being the subject of sexual photographs is absurd, you can't respond to the argument "17 year olds look too young, nude pictures of them will make perverts want to rape babies" with the most powerful response - a photographic counterexample.
Information isn't dangerous, uninformed people are. The only reason to censor discussion is to defend political positions that can't hold up to the close examination they would get if discussion were legal - or occasionally as a political tool to distract people from relevant political issues by reminding them of people/opinions they don't like.
AMD doesn't provide clean avenues for contact on this sort of thing. There are only three types of action I could take to get them to listen:
Documenting their hardware has nothing to do with "giving stuff away for free". Their video cards cost quite a bit of money, and AMD is losing sales every day to Intel because they won't release the docs to let people use their hardware. That's both video card *and* processor sales, because Intel graphics only come with Intel processors.
Releasing programming documentation for hardware is pretty normal. In fact, high end 3D graphics is the only area where the "binary blob" driver thing has ever really come to be seen as "the norm". No one else is foolish enough to think that they can sell hardware without letting their customers actually use it.mm
AMD has gotten where it is today by releasing standard, well documented computer hardware. No one would consider for *one second* developing a product based on a CPU that wasn't documented, and the same is true for video cards - especially with the trend towards GPGPUs.
ATI was in a market niche where they didn't have much competition and all they really needed to do was support Microsoft's closed Direct X standard. AMD has to look at the whole market, and if they let the "closed hardware" ATI policy get on their CPUs (and they're coming together with Fusion), companies will stop buying their stuff damn fast. Right now it would be extremely valuable to their future business to open up their current graphics cards so that system developers don't shy away from Fusion simply based on ATI's reputation.
If AMD doesn't stay open, then everyone who has been ignoring the high power embedded / SFF / custom systems market (which is easily as large as the PC market) will get really confused when AMD and VIA swap places and the main competition to Intel is the (well documented) Via C7 and Unichrome Pro Graphics.
High end graphics cards have never been terribly friendly to power supplies. I see no reason to think the 2600 or 2400 cards - which have significantly fewer stream processors and are fabbed at 65nm instead of 80nm - will have unreasonable power draw.
The much more important issue for DAAMIT at the moment is to actually release programming info for their hardware (or at least release some Open Source drivers). Even *Intel* can get this right - they have no excuse whatsoever.
High quality 1080p video. Animated textures for video games. A massive sample database for a voice synthesizer.
I'm not actually sure what you would do with a 10,000 TB hard disk - but 10 TB is well within the "use it up with some video" range.
So spend money to train people. There's got to be a similar field that pays less than $100,000/year.
Bullshit. If they need them, they can pay them. If they can't afford to pay them, that means they'd rather not have them (i.e. they don't really need them).
If the company is losing $33,000/year on just your friend, they can easily afford to pay $80,000 instead of $50,000 for microscope operators.
The US economy overall also loses out. A legitimate labor shortage will naturally fix itself as more people enter that field for the shortage-induced high pay. Adding H1B workers just means that the supply *never* corrects, thus maintaining the shortage and providing an excuse to demand more H1B's.
Perhaps by... training more operators? When demand goes up, price goes up. That drives supply up to what's needed, and price comes back down. Providing artificial supply just prevents real supply from being created - in your example screwing Americans out of 2,000 skilled jobs in the medium-term.
Take a look at the "transmission" design in the Prius. It's sort of amusing - it never "shifts".
There are a ton of good GUI apps written in Python. They use such GUI toolkits as WxWidgets and Gtk. Python is already a dependency, so including a second dependency is basically trivial. On Linux, packages can have dependencies that are installed automatically. On Windows, there's a tool that lets you blob together the python interpreter and any libraries into one big Windows binary.
Users occasionally need to perform administrative tasks. This happens frequently enough that if they needed to log out and log back in as root to do them, they would just stay logged in as root all the time for convenience. That's obviously not secure.
So, that means that there needs to be a way to get administrative privileges temporarily. In order to have any security at all from this, it has to require user interaction (otherwise programs could automatically get root, and there would be no security). That leaves the choice between the Windows Vista "press OK to perform admin task" solution and the Mac OS X / Ubuntu "type in your password to perform admin task" solution. Making the user type in their password at least proves that it's not someone else sitting down at their computer, and makes sure the user notices that something special happened.
Why does it have to be standard? Why not just write something that meets your needs on top of PyGame?
That's not "programming", that's "construction". Once all the models and environments are constructed, you still have to program them - using a traditional programming language.
Perhaps a solid purely-visual programing environment could be built. People would still need to know how to program - pretty pictures don't make the problem of understanding data structures and algorithms go away.
How is one desktop distribution becoming more popular, especially a community-maintained distribution like Ubuntu, a bad thing? It'll make Linux more popular overall, and totally kill the "too many choices" FUD. It's not like you have to stop using whatever your favorite distro is - hell, if you buy a Dell with Ubuntu you know the hardware works with any Linux-based distro. At worst, people might start releasing proprietary software in .deb format only - but that already happens with RPM and doesn't cause much of a problem.
Wait... how is being able to *sketch* a more useful skill than knowing how to make full and efficient use of a computer? Every job - all of them - can benefit from computer automation. Even just knowing enough programming to know what's feasible for a real programmer to do is priceless.
It's not worth it to Microsoft to trade the benifits of being a monopoly away to avoid the costs of being a monopoly. A couple million a day in non-compliance fines is much cheaper than application developers porting to Linux because it has 25% market share would be - once that starts happening they're basically done.
If they really release the 2950 XTX built on a 65nm process, that could solve the heat problems. The 2900 XT is good enough to prevent people from counting DAAMIT out of the high end graphics market entirely. Oh, and the 2600 and 2400 parts look like they could be really nice for power/performance - and those are what matter to most of the market anyway.
That'd be really funny. As the saying goes "Live by the sword... ZOMG! They've Got Nukes!"
Intel graphics are a bit slow for very recent games, but for the majority of Linux users who aren't hardcore gamers the Intel graphics are way, way better than anything that Nvidia or ATI offer. They provide both 2D and 3D acceleration, they don't have stupid bugs that the community can't fix, and they even work great for older games - Quake 3 and stuff like Wolfenstien: Enemy Territory should run great.
Let me point you to the complete lack of any actual Open Source drivers released by ATI as of this moment. ATI is known for promising a lot of stuff, but their "Commitment to Open Source" (first announced around 1999, repeated consistently since then) has resulted in nothing of any value at all.
Until they actually release Open Source 2D AND 3D drivers, or (even better) release programming docs for their hardware, a boycott is a damn good idea.
Be careful. If this wasn't ATI it might be OK to trust them, but ATI has had a "commitment to Open Source" since 1999, and as a result we have really, really bad proprietary drivers for Linux.
My response to this announcement is: I will try even harder to buy Intel graphics over ATI, until a source release happens. They will get *nothing* from me on an "announcement" this time - this isn't "trick me once" or "trick me twice", we're closer to "trick me 18 times" at this point...
Once they really release Open Source drivers, or (even better) full interface documentation for their cards, then I'll consider their products again. But, until they actually do the whole thing this just looks like more lies.
I've also purchased, with actual money, every one of the Linux game releases that I've been interested in over the past 8 years. That's only 14 major commercial game releases, but that's constrained by my selection rather than by my budget.
This year, I'm waiting impatiently for two major Linux game releases: ET: Quake Wars and UT3.
This seems to strongly depend on exactly which card/driver combination you're using. Some of them work great with the Free drivers, and some of them work like crap. When they work fine, I mostly don't notice. The Radeon X1300 on one of my machines though - the Free drivers don't even do basic 2D acceleration, surfing the web is ass-slow because it can't redraw fast enough to scroll.