I never understand why they required to pay extra again by some people.
Simple: Everybody agrees to raise taxes on demographics they do not belong to. Everybody wants the services of government, but nobody wants to pay for them.
The cause of the majority of leak problems seem to originate from the extensions engine. People with more extensions or more complex extensions will see their memory usage increase faster, and memory usage will typically never decline. Many people blame extensions coders for poor coding, but, honestly, if you're going to have a system which allows end-users with little to no development experience or resources to generate executable code in the browser you should have a pretty robust engine for it with some decent garbage collectors. It would be ridiculous if JavaScript-heavy sites caused this kind of memory-leak-like symptoms, or if certain HTML codes permanently consumed memory like extensions seem to. People would harshly and openly mock IE if it had this problem.
You should be able to duplicate the problem by installing Firefox, installing 5-6 popular extensions, and then browsing for about a day without ever closing the browser. If you maintain more than 10 tabs open at the same time, more's the better. It's much better since 3.0 days, which I believe had more severe problems, but it wouldn't surprise me if there were still problems today.
Overall the real problem with Firefox is that Mozilla has no idea what stage of development Firefox is in. They have too many devs working on new hotness features and not enough on old and busted maintenance.
It's not a new or unique problem. Talk to a.NET developer about System.IO. The.NET Framework 4 was released a year ago, and the damn system still can't handle paths deeper than 260 characters. You can't even use the Unicode handlers like \\?\ because those functions don't have a.NET provider. It's a system-wide problem in Windows, actually, but the fact that the original "backward compatibility" argument they used 10 years ago is the same one they're using now -- and still not providing a widely supported alternative -- is ridiculous and appalling.
I absolutely love Chrome, but the one feature I can't stand is the integrated PDF viewer. Why? You can't save the document to disk from the viewer, and you can't print the document from the viewer. Just what I want in a viewer: the lack of features which even a basic text document has. [Actually, I just checked Chrome 11 beta, and you can print now... not sure when that was added but it must have been recently.]
Who thought that was a good idea? I visit sites for work which generate PDF reports from javascript links. I can't right-click, save those. I sometimes need to print these reports once generated, or attach them to work orders or jobs in another web app. The built-in PDF viewer is completely useless for this. Nevermind if the PDF document has more complex features like forms, which don't work at all. Its one of the first things I turn off when I install the browser (and the inspiration for me tracking down the awesome ChromeAccess extension) now because it's so common for me to run into irritations with it.
I, too, am afraid these devices will fall into the wrong hands: Police officers who use them incorrectly, inappropriately, or abusively. The newspapers are already full of instances of tasers and pepper sprays being used in such ways.
To be clear, it's not targetting vulnerabilities in in IIS or MS-SQL. They're targetting Bobby Tables vulnerabilities in CMS and web apps. The same vulnerabilities exist regardless of what web server or database platform you're using. Once you've found your injection vulnerability you can just query the DB for the platform. Pretty much every platform has a built-in command for listing the attached databases. It's trivial to work back from there. Once you've established the specific CMS app (assuming they didn't brand it all over the normal output pages anyways) you can figure out how to return forwarders as desired.
PostgreSQL: select datname from pg_database; MySQL: show databases; MS SQL Server: exec sp_databases; Oracle: select * from user_tablespaces;
The issues that have been hitting our network (a public school district) have been the sites who only respond with malware infection when the HTTP Referer (sic) is from Google Image Search or Bing Image Search. Additionally, the software typically doesn't try to install anywhere outside of the user's home folders (which they're obviously going to have write access to). So even though only IT staff have rights above a standard User, users are able to get infected. The web-based security scans that Google and Bing do don't catch these infections at all, and so they're not removed from search results like they should be. It's quite frustrating that neither Google nor Microsoft has addressed this problem.
On the plus side, only teachers and elementary students are gullible enough to actually install the software. Older students know it's a scam.
Typically when there's a large gap between the initial public complaint and the filing of the lawsuit, the primary thing that's been going on are private negotiations or arbitration between the parties involved to work things out without the massive expense of going to court. There's no need to get courts involved -- courts decide things for strange, arbitrary reasons sometimes even without the added mystery of a jury. Courtroom rulings typically punish one party more than they'd agree to, and restore justice in a less appealing way to the other. Being able to bargain means you stand a good chance of getting a good part of what you want. Courts don't let that happen. In this case I would expect that negotiations simply broke down, or Tesla and the BBC made tentative agreements which one side or the other did not meet. Perhaps the BBC agreed not to air the program before making a public retraction or airing it with a notice that the problems were staged and then aired it unmodified, or perhaps Clarkson was supposed to write an article of explanation/apology and Tesla got tired of waiting.
Yes, Tesla could have made a media circus out of the issue as soon as they had the evidence they needed, but ideally you don't want to burn bridges over something like this. Unlike political elections, which are a zero-sum game, you don't make compromises by throwing egg on your opponent's face. If Tesla handled it wrong, they could very easily discourage any reviews of any kind of their product from any media outlet. Who wants to review something if you'll get sued for making a bad review? Who would trust reviews of a product from a company that sues bad reviewers? Not a good position for a company to be in.
You're equivocating. The options aren't "do not air for 22.3 years" and "air with indiscretion". There's a happy medium, probably in the neighborhood of 6 months at the earliest in this case, assuming the disaster doesn't significantly increase in severity. Considering the number of Simpsons episodes (nearly 500) they could air two episodes M-F and not repeat an episode until sometime next year without having a problem.
Shouting this behavior is somehow outrageous when it's entirely possible to do what they're suggesting in a totally reasonable manner is disingenuous. They're not being ridiculous and saying they should ban the program permanently because of the fuel rod in the opening sequence. Give them some credit and benefit of the doubt, and then when they screw up for doing something actually ridiculous you won't be considered a quack and a lunatic for raking them over the coals.
That's fine, but it would still have been insensitive and in bad taste to have aired that episode on 9/12. When bad things happen it's customary to be aware of the feelings of those who might have been impacted by it as a sign of respect. Stand-up comedians get away with it because they're supposed to be disrespectful and outrageous, but this is a TV station. If they're still banning the episodes next year at this time then I'd argue they're going to far.
This isn't "censorship" (a grossly misused term on this site). It's discretion.
I completely agree with you about everything -- particularly about developers who suffer from extreme 'not-coded-here' syndrome -- except the last line:
I'd say startups don't use.NET and Windows in general, because of licensing. Simple. They don't have to cash to do it. You might also find that the people who have worked at startups are used to dealing with this, because of their own monetary constraints.
I'd argue that a medium size startup isn't going to care about the cost of software licencing. It's a drop in the bucket compared to office space, computer hardware, office equipment, and personnel.
For a small developer start-up, I can't see them not using the MS Action Pack. It's free to be a MS Partner, and then $300/year to subscribe to the Action Pack. That gets you (among other things): Office 2010 Pro Plus x10 Project Pro 2010 x5 Visio Pro 2010 x10 Exchange Std 2010/w 10 CALs SQL Server 2008 Enterprise R2/w 10 CALs Windows 7 Pro upgrade x10 Server 2008 R2 Ent/w 10 CALs Web Server 2008 R2 SBS 2008 R2/w 10 CALs
All the above are real SKUs of the software. Not developer versions. But if you want developer versions, you get those, too, with the 3 MSDN Subcriptions. The three subscriptions alone would typically run you $2,000 apiece.
Because the major laws to bust existing monopolies, such as the Sherman Anti-Trust Act (and RICO if fraudulent practices are being performed), are Federal.
You're not selling it at a higher price. You're selling access to a software feature not normally included with the standard device [any longer]. It's the equivalent of the police package for an automobile.
If you go out and buy a PS3 today off the shelf, it's not going to have Other OS. If you want to do that, you've got to hack it. While you might argue it's the equivalent of jailbreaking a phone, the courts have yet to rule on it. Either way, you're not going to have vendor support for the hardware and it's going to void the warranty. That's a lot of risk to subsume.
No one built a HPC cluster from the PS3 believed the game was going to go the whole nine innings.
Bullshit.
It's a cluster of 1,700 PS3s which means 1,700 consumer grade hard drives. Let's assume they've thought of this and changed to, say, Western Digital Caviar Blacks (so-called enterprise grade). Those have a 1.2 million hours MTBF. Normal consumer grade hard drives are closer to 750,000 hours.
1,200,000 hours MTBF / 24 hours per day = 50,000 mean drive days between failures 50,000 mean drive days between failures / 1,700 drives = ~30 days between failure
Consumer grade drives would be about 60% of that (~18 days).
Therefore on a system this large you would expect to be replacing or repairing between one and two PS3s each month just due to normal hard drive failures under preferred conditions on average. Admittedly, hard drives are the most common failure type, but there will be systems failing due to bad RAM, bad power supply, defective system board, etc. Nobody builds a system this large without running these numbers. You don't build a system this large unless you can establish a steady supply of replacement hardware for the life of the project.
No, RAMBUS was a patent protected monopoly. They didn't license the technology cheaply (or broadly enough) and so prices were high for it due to low supply and high demand (it was Intel's only memory platform for a couple years). Rambus failed because they sucked at basic economics.
The DRAM collusion investigations involved Hynix, Infineon, Samsung, Micron, and Elpida. Rambus actually has lawsuits against those companies alleging that they colluded to drive the price of Synchronous DRAM down and thus drive Intel back to SDRAM. I'm not sure how that works, because the US DoJ fined the above companies ~$700 million for colluding to keep prices high.
Yes, that's the wise thing to do. Pick on the customer who not only has more lawyers, who not only has special laws which apply to their behavior as a defense organization, who not only has more money than Sony, but who also has more friends in Congress.
If you're the 9th grade bully you don't go picking on the 12th grade wrestling star who's the son of the Principal. Pick battles you can win.
The wise thing to do is to produce a new SKU of the PS3 designed for distributed computing and development which allows the Other OS option and has a special SDK but, for example, can't join PSN (and perhaps cannot even play PS3 games) or which uses a special PSN for this purpose. Then you no have a way to sell these devices to your customers and you can increase the price per unit because you can no longer expect to recoup your losses on game software purchases. Indeed, all you should need to do is put in an option that lets you enable a distributed computing mode. Perhaps entering a software key which the bootstrap firmware will recognize. Then it's just a matter of selling a site license software key. You don't even need a truly different SKU.
"But people will hack it!" Like they already have? This way you get paid for legitimate people to use your product as they wish. You do what you can to prevent loss from hacking and the like, but it's not a valid excuse for not selling what people are demanding from you. The secret of capitalism is to give people what they want at a price they will pay, not to punish them for doing something you didn't expect.
Indeed, the NYT has explicitly agreed that Twitter has the right to do what has been done. If they don't like having their tweets copied, processed, adapted, or published, they need to stop using Twitter.
Unfortunately, lack of personal responsibility is also the primary GOOD of corporations. It allows people to risk investing without risking personal bankruptcy.
The point is you can't call MS out about it without calling out all vendors. Red Hat does per-socket. I think most SlashDot readers would call them ethical. EnterpriseDB (the commercial support for PostgreSQL) does it. VMware does it. I argue Google *would* do it, but they don't have anything that they sell other than appliances and web services. Per-socket shouldn't discount them from a "top X ethical" list in any way since by definition it's a relative list of peers and all their peers generally do it.
I also question what's unethical about it. It prevents corporations from getting around per-install licencing by installing the software once on a huge monolithic server. Corporations are assholes. You know they'd do that and not feel any remorse about abusing their vendor's price structure for support. If it's unethical, what's their alternative to prevent asshole customers from abusing support contracts?
1) Petroleum is non-renewable. Plant and food waste is renewable.
2) If a large corporation adopts such a product, they must have found a way to make it economically viable because corporations are fundamentally selfish. Thus, we have an economically viable, renewable, alternative resource to produce PET.
Petroleum is finite. It's going away. We will need to find alternatives to all the uses for which we've found for petroleum. Primarily that's fuel, lubricants, and plastics. Petroleum-based fuel is fundamentally flawed and so different alternatives are important to find, but both lubricants and plastics are still very necessary. Because petroleum is largely composed of decomposed plant material, it makes sense to produce the replacement resource from contemporary plant materials. We just need to find a way to economically produce simple hydrocarbons.
How can that be "unethical" if it's what all vendors do? The only time MS got any criticism is when processor makers started producing multiple processors per socket and their software wasn't licensed with that idea in mind. How can you blame them for following the status quo for personal computers since their inception? Now MS, like all the vendors, has per-socket pricing.
I very well may be incorrect here, but it sounds to me like you're confusing clusters (aka, allocation units) of a file format and physical disk sectors. However, it's important to note that a cluster is a group of physical disk sectors, and at some point the hardware has to address every single sector of a disk. The OS can deal with clusters of sectors because it logically makes things much easier, but the actual hardware needs to work with sectors. Since historically magnetic disks have always had 512 byte sectors,
Yes, FAT32 supports 4K allocation units. So does NTFS. It's the NTFS default, in fact. Lots of disk formats support cluster sizes larger than 512 bytes.
Note that the FAT and FAT32 files systems impose the following restrictions on the number of clusters on a volume: FAT: Number of clusters <= 65526 FAT32: 65526 < Number of clusters < 4177918
You want a disk that has clusters larger than 512 bytes? Format it. Chances are it already had 4K clusters anyhow.
You want a disk with sectors larger than 512 bytes? You need an Advanced Format drive. The industry has been working on long sector format disks for the past 10 years precisely because they saw this problem coming, and only completed the specification about a year ago.
You want to boot a disk larger than 2TiB? Standard PC BIOS (one of the few remaining relics of the old IBM PC) can't do it. The MBR necessarily must end at 2TiB due to math limits. You need a disk with a GPT instead of an MBR and EFI instead of PC BIOS.
I'm not completely sure, but I *think* that you *should* be able to install and boot a 32-bit OS with native Advanced Format and EFI/GPT support (note that that excludes all Windows OSs currently), an Advanced Format drive, and a GPT/EFI setup all with a disk larger than 2TiB. Leastwise, in this scenario there shouldn't be any mathematical limits, but again I'm no expert and I've no idea if a given OS will support a given setup. Indeed, in this situation I think a 32-bit system would be able to work just fine on disks as large as 16TiB. The support has to be the complete stack from the sectors (Advanced Format) to the system firmware (EFI) to the disk partition format (GPT) to the file format to the operating system (you can suffer some performance issues if the OS is not AF-aware and doesn't align your clusters to the sectors correctly). I'm certainly no expert, but the reading I've done on the topic suggests this is the ultimate goal.
On Windows you can use Powershell to determine if your disk aligns to the 4KB sector boundaries pretty easily: Get-WmiObject Win32_DiskPartition | Select-Object Name, Index, BlockSize, StartingOffset | Format-Table -AutoSize
If StartingOffset is divisible by 4096 (that is, StartingOffset % 4096 = 0) you're fine. Windows 7 and Vista do this by default, I believe, although there is a hotfix for Windows PE to ensure that it does this. If you're still using disk-based imaging instead of file-based, though, you'll want to check your systems to be sure they're aligning correctly on Advanced Format disks. Such systems will experience performance degradation if clusters and sectors are misaligned. I do not know (as I've never used an Advanced Format disk yet) how BlockSize changes.
Slashdot Mirror
Simple: Everybody agrees to raise taxes on demographics they do not belong to. Everybody wants the services of government, but nobody wants to pay for them.
Momentum? Ain't that Physics?
The cause of the majority of leak problems seem to originate from the extensions engine. People with more extensions or more complex extensions will see their memory usage increase faster, and memory usage will typically never decline. Many people blame extensions coders for poor coding, but, honestly, if you're going to have a system which allows end-users with little to no development experience or resources to generate executable code in the browser you should have a pretty robust engine for it with some decent garbage collectors. It would be ridiculous if JavaScript-heavy sites caused this kind of memory-leak-like symptoms, or if certain HTML codes permanently consumed memory like extensions seem to. People would harshly and openly mock IE if it had this problem.
You should be able to duplicate the problem by installing Firefox, installing 5-6 popular extensions, and then browsing for about a day without ever closing the browser. If you maintain more than 10 tabs open at the same time, more's the better. It's much better since 3.0 days, which I believe had more severe problems, but it wouldn't surprise me if there were still problems today.
Overall the real problem with Firefox is that Mozilla has no idea what stage of development Firefox is in. They have too many devs working on new hotness features and not enough on old and busted maintenance.
It's not a new or unique problem. Talk to a .NET developer about System.IO. The .NET Framework 4 was released a year ago, and the damn system still can't handle paths deeper than 260 characters. You can't even use the Unicode handlers like \\?\ because those functions don't have a .NET provider. It's a system-wide problem in Windows, actually, but the fact that the original "backward compatibility" argument they used 10 years ago is the same one they're using now -- and still not providing a widely supported alternative -- is ridiculous and appalling.
5 isn't so bad. Mozilla is planning to release both 6 and 7 this year, too!
I absolutely love Chrome, but the one feature I can't stand is the integrated PDF viewer. Why? You can't save the document to disk from the viewer, and you can't print the document from the viewer. Just what I want in a viewer: the lack of features which even a basic text document has. [Actually, I just checked Chrome 11 beta, and you can print now... not sure when that was added but it must have been recently.]
Who thought that was a good idea? I visit sites for work which generate PDF reports from javascript links. I can't right-click, save those. I sometimes need to print these reports once generated, or attach them to work orders or jobs in another web app. The built-in PDF viewer is completely useless for this. Nevermind if the PDF document has more complex features like forms, which don't work at all. Its one of the first things I turn off when I install the browser (and the inspiration for me tracking down the awesome ChromeAccess extension) now because it's so common for me to run into irritations with it.
I, too, am afraid these devices will fall into the wrong hands: Police officers who use them incorrectly, inappropriately, or abusively. The newspapers are already full of instances of tasers and pepper sprays being used in such ways.
To be clear, it's not targetting vulnerabilities in in IIS or MS-SQL. They're targetting Bobby Tables vulnerabilities in CMS and web apps. The same vulnerabilities exist regardless of what web server or database platform you're using. Once you've found your injection vulnerability you can just query the DB for the platform. Pretty much every platform has a built-in command for listing the attached databases. It's trivial to work back from there. Once you've established the specific CMS app (assuming they didn't brand it all over the normal output pages anyways) you can figure out how to return forwarders as desired.
PostgreSQL: select datname from pg_database;
MySQL: show databases;
MS SQL Server: exec sp_databases;
Oracle: select * from user_tablespaces;
The issues that have been hitting our network (a public school district) have been the sites who only respond with malware infection when the HTTP Referer (sic) is from Google Image Search or Bing Image Search. Additionally, the software typically doesn't try to install anywhere outside of the user's home folders (which they're obviously going to have write access to). So even though only IT staff have rights above a standard User, users are able to get infected. The web-based security scans that Google and Bing do don't catch these infections at all, and so they're not removed from search results like they should be. It's quite frustrating that neither Google nor Microsoft has addressed this problem.
On the plus side, only teachers and elementary students are gullible enough to actually install the software. Older students know it's a scam.
Typically when there's a large gap between the initial public complaint and the filing of the lawsuit, the primary thing that's been going on are private negotiations or arbitration between the parties involved to work things out without the massive expense of going to court. There's no need to get courts involved -- courts decide things for strange, arbitrary reasons sometimes even without the added mystery of a jury. Courtroom rulings typically punish one party more than they'd agree to, and restore justice in a less appealing way to the other. Being able to bargain means you stand a good chance of getting a good part of what you want. Courts don't let that happen. In this case I would expect that negotiations simply broke down, or Tesla and the BBC made tentative agreements which one side or the other did not meet. Perhaps the BBC agreed not to air the program before making a public retraction or airing it with a notice that the problems were staged and then aired it unmodified, or perhaps Clarkson was supposed to write an article of explanation/apology and Tesla got tired of waiting.
Yes, Tesla could have made a media circus out of the issue as soon as they had the evidence they needed, but ideally you don't want to burn bridges over something like this. Unlike political elections, which are a zero-sum game, you don't make compromises by throwing egg on your opponent's face. If Tesla handled it wrong, they could very easily discourage any reviews of any kind of their product from any media outlet. Who wants to review something if you'll get sued for making a bad review? Who would trust reviews of a product from a company that sues bad reviewers? Not a good position for a company to be in.
You're equivocating. The options aren't "do not air for 22.3 years" and "air with indiscretion". There's a happy medium, probably in the neighborhood of 6 months at the earliest in this case, assuming the disaster doesn't significantly increase in severity. Considering the number of Simpsons episodes (nearly 500) they could air two episodes M-F and not repeat an episode until sometime next year without having a problem.
Shouting this behavior is somehow outrageous when it's entirely possible to do what they're suggesting in a totally reasonable manner is disingenuous. They're not being ridiculous and saying they should ban the program permanently because of the fuel rod in the opening sequence. Give them some credit and benefit of the doubt, and then when they screw up for doing something actually ridiculous you won't be considered a quack and a lunatic for raking them over the coals.
Woo! Time to submit my patent to the PTO:
"Description of Selachimorpha-mounted Electrical Fire Suppression Systems"
No, this is SlashDot.
That's fine, but it would still have been insensitive and in bad taste to have aired that episode on 9/12. When bad things happen it's customary to be aware of the feelings of those who might have been impacted by it as a sign of respect. Stand-up comedians get away with it because they're supposed to be disrespectful and outrageous, but this is a TV station. If they're still banning the episodes next year at this time then I'd argue they're going to far.
This isn't "censorship" (a grossly misused term on this site). It's discretion.
I completely agree with you about everything -- particularly about developers who suffer from extreme 'not-coded-here' syndrome -- except the last line:
I'd argue that a medium size startup isn't going to care about the cost of software licencing. It's a drop in the bucket compared to office space, computer hardware, office equipment, and personnel.
For a small developer start-up, I can't see them not using the MS Action Pack. It's free to be a MS Partner, and then $300/year to subscribe to the Action Pack. That gets you (among other things): /w 10 CALs /w 10 CALs /w 10 CALs /w 10 CALs
Office 2010 Pro Plus x10
Project Pro 2010 x5
Visio Pro 2010 x10
Exchange Std 2010
SQL Server 2008 Enterprise R2
Windows 7 Pro upgrade x10
Server 2008 R2 Ent
Web Server 2008 R2
SBS 2008 R2
All the above are real SKUs of the software. Not developer versions. But if you want developer versions, you get those, too, with the 3 MSDN Subcriptions. The three subscriptions alone would typically run you $2,000 apiece.
https://partner.microsoft.com/40016455
Are you a small start up? Go out and buy your barebones servers, get half a dozen consumer-grade laptops, and get the Action Pack. Done.
Because the major laws to bust existing monopolies, such as the Sherman Anti-Trust Act (and RICO if fraudulent practices are being performed), are Federal.
You're not selling it at a higher price. You're selling access to a software feature not normally included with the standard device [any longer]. It's the equivalent of the police package for an automobile.
If you go out and buy a PS3 today off the shelf, it's not going to have Other OS. If you want to do that, you've got to hack it. While you might argue it's the equivalent of jailbreaking a phone, the courts have yet to rule on it. Either way, you're not going to have vendor support for the hardware and it's going to void the warranty. That's a lot of risk to subsume.
Bullshit.
It's a cluster of 1,700 PS3s which means 1,700 consumer grade hard drives. Let's assume they've thought of this and changed to, say, Western Digital Caviar Blacks (so-called enterprise grade). Those have a 1.2 million hours MTBF. Normal consumer grade hard drives are closer to 750,000 hours.
1,200,000 hours MTBF / 24 hours per day = 50,000 mean drive days between failures
50,000 mean drive days between failures / 1,700 drives = ~30 days between failure
Consumer grade drives would be about 60% of that (~18 days).
Therefore on a system this large you would expect to be replacing or repairing between one and two PS3s each month just due to normal hard drive failures under preferred conditions on average. Admittedly, hard drives are the most common failure type, but there will be systems failing due to bad RAM, bad power supply, defective system board, etc. Nobody builds a system this large without running these numbers. You don't build a system this large unless you can establish a steady supply of replacement hardware for the life of the project.
No, RAMBUS was a patent protected monopoly. They didn't license the technology cheaply (or broadly enough) and so prices were high for it due to low supply and high demand (it was Intel's only memory platform for a couple years). Rambus failed because they sucked at basic economics.
The DRAM collusion investigations involved Hynix, Infineon, Samsung, Micron, and Elpida. Rambus actually has lawsuits against those companies alleging that they colluded to drive the price of Synchronous DRAM down and thus drive Intel back to SDRAM. I'm not sure how that works, because the US DoJ fined the above companies ~$700 million for colluding to keep prices high.
Yes, that's the wise thing to do. Pick on the customer who not only has more lawyers, who not only has special laws which apply to their behavior as a defense organization, who not only has more money than Sony, but who also has more friends in Congress.
If you're the 9th grade bully you don't go picking on the 12th grade wrestling star who's the son of the Principal. Pick battles you can win.
The wise thing to do is to produce a new SKU of the PS3 designed for distributed computing and development which allows the Other OS option and has a special SDK but, for example, can't join PSN (and perhaps cannot even play PS3 games) or which uses a special PSN for this purpose. Then you no have a way to sell these devices to your customers and you can increase the price per unit because you can no longer expect to recoup your losses on game software purchases. Indeed, all you should need to do is put in an option that lets you enable a distributed computing mode. Perhaps entering a software key which the bootstrap firmware will recognize. Then it's just a matter of selling a site license software key. You don't even need a truly different SKU.
"But people will hack it!" Like they already have? This way you get paid for legitimate people to use your product as they wish. You do what you can to prevent loss from hacking and the like, but it's not a valid excuse for not selling what people are demanding from you. The secret of capitalism is to give people what they want at a price they will pay, not to punish them for doing something you didn't expect.
Indeed, the NYT has explicitly agreed that Twitter has the right to do what has been done. If they don't like having their tweets copied, processed, adapted, or published, they need to stop using Twitter.
Unfortunately, lack of personal responsibility is also the primary GOOD of corporations. It allows people to risk investing without risking personal bankruptcy.
Yes, yes, it's always a programmer's fault.
The point is you can't call MS out about it without calling out all vendors. Red Hat does per-socket. I think most SlashDot readers would call them ethical. EnterpriseDB (the commercial support for PostgreSQL) does it. VMware does it. I argue Google *would* do it, but they don't have anything that they sell other than appliances and web services. Per-socket shouldn't discount them from a "top X ethical" list in any way since by definition it's a relative list of peers and all their peers generally do it.
I also question what's unethical about it. It prevents corporations from getting around per-install licencing by installing the software once on a huge monolithic server. Corporations are assholes. You know they'd do that and not feel any remorse about abusing their vendor's price structure for support. If it's unethical, what's their alternative to prevent asshole customers from abusing support contracts?
The major improvement is two-fold:
1) Petroleum is non-renewable. Plant and food waste is renewable.
2) If a large corporation adopts such a product, they must have found a way to make it economically viable because corporations are fundamentally selfish. Thus, we have an economically viable, renewable, alternative resource to produce PET.
Petroleum is finite. It's going away. We will need to find alternatives to all the uses for which we've found for petroleum. Primarily that's fuel, lubricants, and plastics. Petroleum-based fuel is fundamentally flawed and so different alternatives are important to find, but both lubricants and plastics are still very necessary. Because petroleum is largely composed of decomposed plant material, it makes sense to produce the replacement resource from contemporary plant materials. We just need to find a way to economically produce simple hydrocarbons.
How can that be "unethical" if it's what all vendors do? The only time MS got any criticism is when processor makers started producing multiple processors per socket and their software wasn't licensed with that idea in mind. How can you blame them for following the status quo for personal computers since their inception? Now MS, like all the vendors, has per-socket pricing.
I very well may be incorrect here, but it sounds to me like you're confusing clusters (aka, allocation units) of a file format and physical disk sectors. However, it's important to note that a cluster is a group of physical disk sectors, and at some point the hardware has to address every single sector of a disk. The OS can deal with clusters of sectors because it logically makes things much easier, but the actual hardware needs to work with sectors. Since historically magnetic disks have always had 512 byte sectors,
Yes, FAT32 supports 4K allocation units. So does NTFS. It's the NTFS default, in fact. Lots of disk formats support cluster sizes larger than 512 bytes.
Partial output of 'format /?' on Windows 7:
You want a disk that has clusters larger than 512 bytes? Format it. Chances are it already had 4K clusters anyhow.
You want a disk with sectors larger than 512 bytes? You need an Advanced Format drive. The industry has been working on long sector format disks for the past 10 years precisely because they saw this problem coming, and only completed the specification about a year ago.
You want to boot a disk larger than 2TiB? Standard PC BIOS (one of the few remaining relics of the old IBM PC) can't do it. The MBR necessarily must end at 2TiB due to math limits. You need a disk with a GPT instead of an MBR and EFI instead of PC BIOS.
I'm not completely sure, but I *think* that you *should* be able to install and boot a 32-bit OS with native Advanced Format and EFI/GPT support (note that that excludes all Windows OSs currently), an Advanced Format drive, and a GPT/EFI setup all with a disk larger than 2TiB. Leastwise, in this scenario there shouldn't be any mathematical limits, but again I'm no expert and I've no idea if a given OS will support a given setup. Indeed, in this situation I think a 32-bit system would be able to work just fine on disks as large as 16TiB. The support has to be the complete stack from the sectors (Advanced Format) to the system firmware (EFI) to the disk partition format (GPT) to the file format to the operating system (you can suffer some performance issues if the OS is not AF-aware and doesn't align your clusters to the sectors correctly). I'm certainly no expert, but the reading I've done on the topic suggests this is the ultimate goal.
On Windows you can use Powershell to determine if your disk aligns to the 4KB sector boundaries pretty easily:
Get-WmiObject Win32_DiskPartition | Select-Object Name, Index, BlockSize, StartingOffset | Format-Table -AutoSize
If StartingOffset is divisible by 4096 (that is, StartingOffset % 4096 = 0) you're fine. Windows 7 and Vista do this by default, I believe, although there is a hotfix for Windows PE to ensure that it does this. If you're still using disk-based imaging instead of file-based, though, you'll want to check your systems to be sure they're aligning correctly on Advanced Format disks. Such systems will experience performance degradation if clusters and sectors are misaligned. I do not know (as I've never used an Advanced Format disk yet) how BlockSize changes.