This starts with the RAM - I want ECC RAM for reliability. That alone disqualifies 99% of pre-built PCs. Add some specific requirements like a fanless midrange GPU (because the fans on those tend to be REALLY crappy), and there is probably not a single pre-built box in the market that really suits my taste in PCs.
You can 'vote with your dollars' by giving the money directly to an open source project which (usually) barely break even, if they actually do (few seem to), rather than corporations which earn billions of dollars per year and generally don't care what you do with your card.
It's kindof naive to spend money on AMD, and expect that the work will be done faster, rather than giving it to the people actually doing all of the hard work.
What I want those corporations to do is provide specifications on how to program their hardware. Many Linux developers claim that this is the most important part of supporting OSS, because reverse engineering is extremely time consuming work.
Right now, AMD/ATI does that while Nvidia doesn't. That they actually pay developers to help with the OSS drivers is a bonus.
Pick any which one and stick with it, period.
Here I don't agree. I'll buy whatever seems to be the best value for money at the time of purchase. Supporting OSS is not the only thing to consider here, but it plays a role. A few years ago, Nvidia was the winner here because of their decent binary drivers while ATI had almost nothing. Crappy binaries and no hardware documentation for OSS developers on the ATI side. But over the last 1-2 years AMD has made a solid effort on the documentation side.
If you got band new shiny AMD card next week 'to support open source' (which it doesn't), and Nvidia open sourced all of their drivers first-party a day later, you'd be red in rage at AMD.
No I wouldn't. And I'm not angry about buying a Nvidia card about two years ago. The offer was better than ATI's at the time. That ATI kept their promise about the docs (which I didn't really trust) does not make me regret that decision.
The question is: what else to use? Intel has a pretty good reputation for its OSS graphics drivers, but the hardware is only adequate for office PCs. For current 3D games their integrated graphics lack the necessary performance.
ATI and NVIDIA have much faster hardware. Binary Linux drivers from NVIDIA work pretty well while ATI's have a horrible reputation, but NVIDIA don't release hardware specifications like ATI does. That means I expect the development of open source drivers to progress much better for ATI. AFAIK the Nouveau open source drivers for NVIDIA rely on reverse engineering which is a rather slow undertaking.
When I built my current PC two years ago, I voted with my money for NVIDIA. ATI's release of hardware specs was promised at the time, but they had not actually released anything yet. My next PC will probably have an ATI card.
The "open source graphics card" (http://wiki.opengraphics.org/tiki-index.php) has reached the prototype stage, but I doubt it will see mass sales anytime soon. In terms of performance, that card seems to be similar to recent integrated graphics, and there we already have Intel chipsets with good OSS support.
Yes, it seems the binary is really crappy in this case and the OSS driver at least passable. Although I'm not familiar with those benchmarks and how they measure up against similar software on Windows. 3D still seems to lag behind, otherwise we could officially forget the Catalyst driver and use the OSS driver exclusively for Linux. But I think we'll get there.
Getting a bit off topic, but I like the direction ATI is taking recently with Open Source. Long term, I think they will be the better choice for Linux. In a recent test at Phoronix (http://www.phoronix.com/scan.php?page=article&item=amd_r600_r700_2d&num=1) the OS driver already offered better 2D performance over the binary one:-)
Maybe GM did not want to spend money on designing two different bumpers for the US and Europe.
I remember one case where I rammed a concrete pillar at maybe 3 mph, with only a tiny dent in the bumper. A few days later, I watched someone else hit the same pillar with a newer Opel model. That one had a weaker bumper. Which got bent, and the rear fender got bent too. With the rear light being another casualty. I guess that cost at least 1000 Euros where I did not bother to make repairs at all.
So give me a ugly big, goofy (and heavy!) 5 mph bumper any day;-)
"Minor" accidents are something like 5 mph. Around 1980, regulations (http://en.wikipedia.org/wiki/Bumper_(automobile)#Strengthening_standards actually required the bumper to handle that. I once had a 1989 Opel Vectra whose bumpers still seemed to match that requirement, and it has saved me a nice chunk of money.
But I agree that you should not expect the bumper to handle a 30 mph crash.
If you look at the video, that did not happen. The body stayed pretty much intact on the side away from the crash, no flying metal sheets. The big problem was that the passenger cell did crumble, and the driver would have been squashed like a bug.
A better use of tax money would be for the government to give it back to the people they took it from in the first place. Why should I be forced to "invest" (if you can even call it that given the track record of government spending) my hard-earned money in battery research through taxes? Has anyone else ever noticed that very often these sorts of "investments" are the very same ones that the private sector won't touch with a ten foot pole?
The private sector tends to ignore anything that won't yield a short term profit. It certainly ignores anything where the benefits are for the public instead of making the company rich. So now and then, it takes some tax money to get things underway that would happen too little too late otherwise.
Take renewable energy for instance. I happen to agree with the people who claim that global Peak Oil cannot be far off anymore. So we need something else, and we better have a lot of it when the next oil crisis arrives. Think summer 2008 again or worse.
In that case, Germany and to a lesser degree Spain have taken the lead by subsidizing various renewable energies, thus creating a market and making investments lucrative. As a result, the technology is a lot more mature than it would be without subsidies. Some companies even claim being close to the point where their PV cells can compete without the subsidies. In particular http://www.firstsolar.com/ and http://www.nanosolar.com/. Both are US companies so the money will not stay in Germany. But I still think it is a good idea, even being one of those Germans who pay for it.
BTW and slightly off topic: A real and much greater waste of money recently happened with the bank bailout. For THAT, I'm really pissed with our current government.
Much is being made of the US Govt is funding these cars that are to be built outside the US, but the fact is that the technology is going to be owned by a US company. Fisker is essentially outsourcing every aspect of their development but the resulting technology, and the profits, will accrue to the US business and be taxed in the US.
Unless the big profits just happen to be made by an offshore sub-contractor which just happens to be owned by the people behind Fisker;-) Seriously, giving money to multinational corporations is just asking for that kind of scam. And when it happens, it will probably use a loophole in the laws so these guys do not even risk jail time.
Besides, hybrid technology is not that new anymore. I have my doubts if it should be reason enough for governments to fund a new car maker. A better use of tax money would be battery research that is released under Open Access, with the patents going to the public domain.
A better approach would be to set up a test server first and run the new version on that. Upgrade or don't based on the results of the test. Now management may refuse that based on the cost for the extra sever. But then they share the fault if things go wrong.
I know one piece of medical equipment that is still being made and uses DOS. The reason is that it has some real time requirements where Windows will fail, and the company never got around to port the software to a real time capable OS.
In a libertarian society there are no laws that trump contract terms.
Yes, in a libertarian society corporations would have private jails (more likely work camps) for housing people who had violated contract terms or were indebted to them. After all, if those things were in the contract they should be legal and respected by government. Government jails would only be for criminals. But the government would probably sell criminals to the corporate work camps.
In practice, I see several problems with that approach.
-If a debtor has multiple creditors he does not pay, who gets to put him (or her) into a labor camp? -If the debtor is a company, who exactly gets put in jail? How do you ensure that the company's assets are actually used to pay the debt, instead of merely sending the management to jail while someone else gets to spend the remaining money in the company's account?
For the latter in particular, bankruptcy laws are necessary. If you eliminate the concept of corporations as persons and require that corporations are controlled only by natural persons, who are personally liable, things become easier. Come to think of it, that might be not such a bad idea.
The computer industry is (seems?) more open because it is in the interest of the dominant players not to have to make everything themselves. Consider the IBM-compatible PC an Windows: -APIs are documented because without them you would not have nearly as much applications to choose from. -Driver SDKs are available so Microsoft does not have to write all of these. Letting hardware vendors make their own means Microsoft saves money.
For a better analogy, imagine people want to repair bugs in Windows themselves. Here, the lack of source code is a big problem. You usually cannot repair Windows, at best you can choose another OS.
As an analogy for THAT, some new company could reverse engineer what the original software does by measuring it. For instance by measuring all the stuff like ignition timing, amount of injected fuel and so on, and then overwrite the original software with their own. To some extent that already happens, see http://en.wikipedia.org/wiki/Chip_tuning. But I have yet to hear of an attempt to make a universal, open Source ECU software.
A coal-burning (for instance) power plant converts only 30-40% of the energy in the fuel to electric energy. The rest is lost in form of waste heat at the power plant.
Now your incandescent lights convert this to heat quite efficiently, but you still get only the above 30-40% of the heat that was originally in the coal. It would be far more efficient to throw an extra briquet into your stove at home, to stay with the coal example.
The above will become irrelevant once all power generation is changed to renewables. But at that point, I guess the price per kWh of electric power will be much higher too. So you will have an economic reason to get efficient devices.
Now and then, a mechanic will go bankrupt and his equipment gets auctioned off to reimburse the creditors. I'm sure there will be some jurisdictions where bankruptcy laws trump any repossession clause in the contract. At this point, a third party (who never signed the contract) might buy the equipment from the auction. Enter the hackers. They buy a few devices and reverse engineer them.
One mechanic violating the contract and putting the decryption keys anonymously on the internet might also work. Even if the manufacturer finds the culprit, the lack of IP laws protecting their trade secrets would mean they cannot go after the people who downloaded the information and keep redistributing it.
Medical insurance is to cover actual medical devices. There is a very good reason these things cost more than a smartphone ever would. They need to be safety tested with live humans, and that is not cheap.
True for things that can kill you if they fail. Imagine an implanted pacemaker blowing up like an iPhone;-)
But if you use a smartphone (with special software) in a way similar to how able-bodied people use it, special safety testing may be unnecessary. Hence I propose a new category "medical assistance device (non-hazardous)" that can be used without expensive special certification. It could cover things like general purpose computers that are loaded with special software, limited to applications where errors pose no significant health hazard.
That's all correct, but I don't really understand Microsoft's position on that.
XP is now reasonably mature, so the maintenance costs for this sort of bug fix should be way lower than when it was new. So why not keep selling it until sales fall off on their own (with Windows 7 being better than Vista that should happen anyway), and support it for a few years after that?
As it is, buyers of netbooks have more reason now to choose Linux. Just when it looked like Microsoft was relenting and keep XP available. Because an OS without support is not so good for connecting to the internet.
IANAL but I'm familiar with a few of the relevant regulations. One of them is that the customer has a claim against the dealer he got the software from. Not against Microsoft directly.
So in theory a pissed off customer in Germany could sue the dealer but not Microsoft. Maybe the dealer could sue Microsoft in turn, but I'm not sure about that. AFAIK contracts between companies allow a lot more exclusions of liabilities than contract with consumers, so Microsoft may have guarded against that.
Several separate, smaller devices combined into a RAID5 array, all inside one 3.5" case. That would take care of failures in one of the sub-devices. In case the "mainboard" that connects them all and holds the SAS interface fails, make the "mainboard" exchangeable. Swapping it will revive the drive.
Numbers matter a lot in EVE. With 4 or 5 ships, even frigates, you can bring enough jammers and firepower that a single opponent will be blind, crippled and soon shot down.
While I like EVE in general, I'd love to try something where a lone ship has somewhat better chances. Possible approaches are -Player skill based dogfights, where even a group of players with mediocre skills does not get the ace pilot -A game that allows stealth attacks, think submarine tactics in space (or acutally underwater, with a submarine MMO)
For the former, Jump Gate Evolution might be worth a try. But I don't expect beta this year anymore, much less release:-(
EVE is IMHO very well designed that way. At least better than the current "standard" MMO formula where you grind to gain levels, have to choose a class and end up with some cookie cutter build if you want to be competitive.
This said, newer MMOs already start being more flexible in that regard. Perfect World and Fallen Earth have a mixed level/skill system where you get some progress based on level, but you also get additional skill points for customizing your character's build. Fallen Earth in particular seems to give out enough of these to raise two specializations to a high level.
Now the The Secret World seems to have yet another approach with plenty of freedom to build your character. Sounds good but I'll wait how it turns out before praising it. The developers of Fallen Earth promised a similar degree of freedom but delivered only halfway.
Slashdot Mirror
This starts with the RAM - I want ECC RAM for reliability. That alone disqualifies 99% of pre-built PCs.
Add some specific requirements like a fanless midrange GPU (because the fans on those tend to be REALLY crappy), and there is probably not a single pre-built box in the market that really suits my taste in PCs.
You can 'vote with your dollars' by giving the money directly to an open source project which (usually) barely break even, if they actually do (few seem to), rather than corporations which earn billions of dollars per year and generally don't care what you do with your card.
It's kindof naive to spend money on AMD, and expect that the work will be done faster, rather than giving it to the people actually doing all of the hard work.
What I want those corporations to do is provide specifications on how to program their hardware. Many Linux developers claim that this is the most important part of supporting OSS, because reverse engineering is extremely time consuming work.
Right now, AMD/ATI does that while Nvidia doesn't. That they actually pay developers to help with the OSS drivers is a bonus.
Pick any which one and stick with it, period.
Here I don't agree. I'll buy whatever seems to be the best value for money at the time of purchase. Supporting OSS is not the only thing to consider here, but it plays a role.
A few years ago, Nvidia was the winner here because of their decent binary drivers while ATI had almost nothing. Crappy binaries and no hardware documentation for OSS developers on the ATI side. But over the last 1-2 years AMD has made a solid effort on the documentation side.
If you got band new shiny AMD card next week 'to support open source' (which it doesn't), and Nvidia open sourced all of their drivers first-party a day later, you'd be red in rage at AMD.
No I wouldn't. And I'm not angry about buying a Nvidia card about two years ago. The offer was better than ATI's at the time. That ATI kept their promise about the docs (which I didn't really trust) does not make me regret that decision.
The question is: what else to use?
Intel has a pretty good reputation for its OSS graphics drivers, but the hardware is only adequate for office PCs. For current 3D games their integrated graphics lack the necessary performance.
ATI and NVIDIA have much faster hardware. Binary Linux drivers from NVIDIA work pretty well while ATI's have a horrible reputation, but NVIDIA don't release hardware specifications like ATI does. That means I expect the development of open source drivers to progress much better for ATI. AFAIK the Nouveau open source drivers for NVIDIA rely on reverse engineering which is a rather slow undertaking.
When I built my current PC two years ago, I voted with my money for NVIDIA. ATI's release of hardware specs was promised at the time, but they had not actually released anything yet. My next PC will probably have an ATI card.
The "open source graphics card" (http://wiki.opengraphics.org/tiki-index.php) has reached the prototype stage, but I doubt it will see mass sales anytime soon. In terms of performance, that card seems to be similar to recent integrated graphics, and there we already have Intel chipsets with good OSS support.
Read the f... Phoronix article ;-)
Yes, it seems the binary is really crappy in this case and the OSS driver at least passable. Although I'm not familiar with those benchmarks and how they measure up against similar software on Windows.
3D still seems to lag behind, otherwise we could officially forget the Catalyst driver and use the OSS driver exclusively for Linux. But I think we'll get there.
Getting a bit off topic, but I like the direction ATI is taking recently with Open Source. Long term, I think they will be the better choice for Linux. :-)
In a recent test at Phoronix (http://www.phoronix.com/scan.php?page=article&item=amd_r600_r700_2d&num=1) the OS driver already offered better 2D performance over the binary one
Maybe GM did not want to spend money on designing two different bumpers for the US and Europe.
I remember one case where I rammed a concrete pillar at maybe 3 mph, with only a tiny dent in the bumper. A few days later, I watched someone else hit the same pillar with a newer Opel model. That one had a weaker bumper. Which got bent, and the rear fender got bent too. With the rear light being another casualty. I guess that cost at least 1000 Euros where I did not bother to make repairs at all.
So give me a ugly big, goofy (and heavy!) 5 mph bumper any day ;-)
"Minor" accidents are something like 5 mph. Around 1980, regulations (http://en.wikipedia.org/wiki/Bumper_(automobile)#Strengthening_standards actually required the bumper to handle that. I once had a 1989 Opel Vectra whose bumpers still seemed to match that requirement, and it has saved me a nice chunk of money.
But I agree that you should not expect the bumper to handle a 30 mph crash.
If you look at the video, that did not happen. The body stayed pretty much intact on the side away from the crash, no flying metal sheets.
The big problem was that the passenger cell did crumble, and the driver would have been squashed like a bug.
A better use of tax money would be for the government to give it back to the people they took it from in the first place. Why should I be forced to "invest" (if you can even call it that given the track record of government spending) my hard-earned money in battery research through taxes? Has anyone else ever noticed that very often these sorts of "investments" are the very same ones that the private sector won't touch with a ten foot pole?
The private sector tends to ignore anything that won't yield a short term profit. It certainly ignores anything where the benefits are for the public instead of making the company rich. So now and then, it takes some tax money to get things underway that would happen too little too late otherwise.
Take renewable energy for instance. I happen to agree with the people who claim that global Peak Oil cannot be far off anymore. So we need something else, and we better have a lot of it when the next oil crisis arrives. Think summer 2008 again or worse.
In that case, Germany and to a lesser degree Spain have taken the lead by subsidizing various renewable energies, thus creating a market and making investments lucrative. As a result, the technology is a lot more mature than it would be without subsidies. Some companies even claim being close to the point where their PV cells can compete without the subsidies. In particular http://www.firstsolar.com/ and http://www.nanosolar.com/. Both are US companies so the money will not stay in Germany. But I still think it is a good idea, even being one of those Germans who pay for it.
BTW and slightly off topic:
A real and much greater waste of money recently happened with the bank bailout. For THAT, I'm really pissed with our current government.
Much is being made of the US Govt is funding these cars that are to be built outside the US, but the fact is that the technology is going to be owned by a US company. Fisker is essentially outsourcing every aspect of their development but the resulting technology, and the profits, will accrue to the US business and be taxed in the US.
Unless the big profits just happen to be made by an offshore sub-contractor which just happens to be owned by the people behind Fisker ;-)
Seriously, giving money to multinational corporations is just asking for that kind of scam. And when it happens, it will probably use a loophole in the laws so these guys do not even risk jail time.
Besides, hybrid technology is not that new anymore. I have my doubts if it should be reason enough for governments to fund a new car maker. A better use of tax money would be battery research that is released under Open Access, with the patents going to the public domain.
It is a perfect reply to the invitation a few levels up to write something positive about Hyper-V. ;-)
No matter if it is really any good
Simply refusing is still pointless.
A better approach would be to set up a test server first and run the new version on that. Upgrade or don't based on the results of the test.
Now management may refuse that based on the cost for the extra sever. But then they share the fault if things go wrong.
Music and art: ;-)
I don't know about the US, but in Germany there are separate classes for these. At least in secondary school, can't remember primary
Baseball and football:
Phys ed. I think the US has that too.
I know one piece of medical equipment that is still being made and uses DOS. The reason is that it has some real time requirements where Windows will fail, and the company never got around to port the software to a real time capable OS.
In a libertarian society there are no laws that trump contract terms.
Yes, in a libertarian society corporations would have private jails (more likely work camps) for housing people who had violated contract terms or were indebted to them. After all, if those things were in the contract they should be legal and respected by government. Government jails would only be for criminals. But the government would probably sell criminals to the corporate work camps.
In practice, I see several problems with that approach.
-If a debtor has multiple creditors he does not pay, who gets to put him (or her) into a labor camp?
-If the debtor is a company, who exactly gets put in jail? How do you ensure that the company's assets are actually used to pay the debt, instead of merely sending the management to jail while someone else gets to spend the remaining money in the company's account?
For the latter in particular, bankruptcy laws are necessary.
If you eliminate the concept of corporations as persons and require that corporations are controlled only by natural persons, who are personally liable, things become easier. Come to think of it, that might be not such a bad idea.
The computer industry is (seems?) more open because it is in the interest of the dominant players not to have to make everything themselves. Consider the IBM-compatible PC an Windows:
-APIs are documented because without them you would not have nearly as much applications to choose from.
-Driver SDKs are available so Microsoft does not have to write all of these. Letting hardware vendors make their own means Microsoft saves money.
For a better analogy, imagine people want to repair bugs in Windows themselves. Here, the lack of source code is a big problem. You usually cannot repair Windows, at best you can choose another OS.
As an analogy for THAT, some new company could reverse engineer what the original software does by measuring it. For instance by measuring all the stuff like ignition timing, amount of injected fuel and so on, and then overwrite the original software with their own.
To some extent that already happens, see http://en.wikipedia.org/wiki/Chip_tuning. But I have yet to hear of an attempt to make a universal, open Source ECU software.
A coal-burning (for instance) power plant converts only 30-40% of the energy in the fuel to electric energy. The rest is lost in form of waste heat at the power plant.
Now your incandescent lights convert this to heat quite efficiently, but you still get only the above 30-40% of the heat that was originally in the coal. It would be far more efficient to throw an extra briquet into your stove at home, to stay with the coal example.
The above will become irrelevant once all power generation is changed to renewables. But at that point, I guess the price per kWh of electric power will be much higher too. So you will have an economic reason to get efficient devices.
Now and then, a mechanic will go bankrupt and his equipment gets auctioned off to reimburse the creditors. I'm sure there will be some jurisdictions where bankruptcy laws trump any repossession clause in the contract.
At this point, a third party (who never signed the contract) might buy the equipment from the auction. Enter the hackers. They buy a few devices and reverse engineer them.
One mechanic violating the contract and putting the decryption keys anonymously on the internet might also work. Even if the manufacturer finds the culprit, the lack of IP laws protecting their trade secrets would mean they cannot go after the people who downloaded the information and keep redistributing it.
Medical insurance is to cover actual medical devices. There is a very good reason these things cost more than a smartphone ever would. They need to be safety tested with live humans, and that is not cheap.
True for things that can kill you if they fail. Imagine an implanted pacemaker blowing up like an iPhone ;-)
But if you use a smartphone (with special software) in a way similar to how able-bodied people use it, special safety testing may be unnecessary.
Hence I propose a new category "medical assistance device (non-hazardous)" that can be used without expensive special certification. It could cover things like general purpose computers that are loaded with special software, limited to applications where errors pose no significant health hazard.
That's all correct, but I don't really understand Microsoft's position on that.
XP is now reasonably mature, so the maintenance costs for this sort of bug fix should be way lower than when it was new. So why not keep selling it until sales fall off on their own (with Windows 7 being better than Vista that should happen anyway), and support it for a few years after that?
As it is, buyers of netbooks have more reason now to choose Linux. Just when it looked like Microsoft was relenting and keep XP available. Because an OS without support is not so good for connecting to the internet.
That will be interesting under German law.
IANAL but I'm familiar with a few of the relevant regulations. One of them is that the customer has a claim against the dealer he got the software from. Not against Microsoft directly.
So in theory a pissed off customer in Germany could sue the dealer but not Microsoft. Maybe the dealer could sue Microsoft in turn, but I'm not sure about that. AFAIK contracts between companies allow a lot more exclusions of liabilities than contract with consumers, so Microsoft may have guarded against that.
The same principle should be extendable to RAID5.
Several separate, smaller devices combined into a RAID5 array, all inside one 3.5" case. That would take care of failures in one of the sub-devices. In case the "mainboard" that connects them all and holds the SAS interface fails, make the "mainboard" exchangeable. Swapping it will revive the drive.
e) Sales of new handsets move to really cheap ones because the (otherwise nice) features of the high end ones are too crippled by DRM.
I agree that this will hurt Japanese handset makers. Independent importers who are not bound by the deal might profit (your case d).
Numbers matter a lot in EVE. With 4 or 5 ships, even frigates, you can bring enough jammers and firepower that a single opponent will be blind, crippled and soon shot down.
While I like EVE in general, I'd love to try something where a lone ship has somewhat better chances. Possible approaches are
-Player skill based dogfights, where even a group of players with mediocre skills does not get the ace pilot
-A game that allows stealth attacks, think submarine tactics in space (or acutally underwater, with a submarine MMO)
For the former, Jump Gate Evolution might be worth a try. But I don't expect beta this year anymore, much less release :-(
EVE is IMHO very well designed that way. At least better than the current "standard" MMO formula where you grind to gain levels, have to choose a class and end up with some cookie cutter build if you want to be competitive.
This said, newer MMOs already start being more flexible in that regard. Perfect World and Fallen Earth have a mixed level/skill system where you get some progress based on level, but you also get additional skill points for customizing your character's build. Fallen Earth in particular seems to give out enough of these to raise two specializations to a high level.
Now the The Secret World seems to have yet another approach with plenty of freedom to build your character. Sounds good but I'll wait how it turns out before praising it. The developers of Fallen Earth promised a similar degree of freedom but delivered only halfway.