So according to this, if I had a competitor I don't want around I could just buy their app a lot and keep getting refunds until I bleed them dry? Sounds good to me.
If you don't like that competitor, why don't you go to his premises with a canister full of petrol and burn everything down? Simple: You don't want to go to jail for arson. And why don't people do what you suggest? Simple. They don't want to end up in court for tortious interference with a business.
So how does the developer of a pay application prevent someone from doing a DoS on the developer's bank account by asking readers of his blog to buy the app and get a refund?
So, let's say there's a game on your iPhone, what would be the expected total time to finish it and get bored with it? I bet less than 90 days. So you can basically have a free game: download, finish, refund. It's golden!
Except that you can't. The contract between Apple and the developer says that _if_ Apple refunds the money then they get some amount back from the developer. However, you are a store customer, and that contract has nothing to do with you. The contract between _you_ and Apple says: No refunds.
Earlier someone suggested sneezing/coughing, covering her mouth with a gloved hand, and then using the same gloved hand to pack swabs. Hard to avoid this unless she works in an enclosed suit (which is unlikely.)
I would assume that the police or their lab don't buy swabs used for DNA testing at the nearest supermarket, they will buy them from a supplier who guarantees uncontaminated swabs, for obvious reasons. To guarantee uncontaminated swabs, that supplier would have to train its employees making the swabs. Part of the training would be to tell them: If you cough or sneeze, and touch your mouth with your gloved hand, you then leave the production line immediately, taking any materials with you that could have been contaminated, destroy them, desinfect your hands, and put on fresh gloves. So either they didn't train their employees properly, or they put pressure on their employees so they couldn't work as needed, or they hired employees who give a shit. Of course this makes the swabs a bit more expensive, but I am sure the police pays more for "guaranteed uncontaminated" swabs then you pay when you want to clean your ears.
In any case, I would like to see what kind of contracts they had, and the company might be liable for a lot of damages.
. I'm curious to know the battery life difference if anyone has done an experiment with their player?
I think on hard-drive based iPods, the battery doesn't actually last for x hours, it lasts for x Megabytes. Use 256 Kbit AAC instead of 128 Kbit, half the battery life. Lossless is obviously worse.
For any operating system using a virtual address space (any modern operating system), this isn't an issue in the slightest. The memory given to a program or process can appear to be a single block to the process, but not have to be physically contiguous thanks to MMU magic.
Missing the point completely. A call to malloc (256 * 1024 * 1024) must return memory that is contiguous in virtual address space. And once that malloc call succeeded, the memory can't be moved in virtual address space. And virtual address space can easily get fragmented.
Assume 15 x 257 MB virtual address space. Use malloc to allocate 1MB, 256MB, 1MB, 256MB, etc. until memory is full. free () all the 256 MB blocks. You end up with fragmented memory with no available block greater than 256 MB. Try allocating 257 MB -> fail.
This also misses the obvious problem that an ID3 tagger might be written to read all the tags from my 10000 songs into memory, without any worries that it might not fit. And suddenly this software throws up on a few dozen files in the new format.
In this case, the MP3 signal serves as the prediction and the remaining data is the correction stream.
Complete bollocks. The first thing you do in MP3 compression is throwing away the phase information, which immediately removes half the data with no quality loss. That makes the compressed signal completely unusable for use as a prediction stream.
Think of replacing sin (omega times t + alpha) with sin (omega times t), where t is the time. There is no audible difference between the signals, but the values are absolutely completely different.
Those damage ratios are exactly what these filings are about. The judge has been asked to consider if such statutory damages may violate cruel and unusual punishment restriction by being so far out of line. Part of the judge's determination would be which cost if any to compare the damages with.
The damage ratio is not actually the problem. The problem is that the damages are so high that the average citizen cannot go to court, because losing the case would bankrupt them. If a railway company overcharges you buy 66 cents and is ordered to pay 10,000 times the amount in damages, that is surely very annoying for the railway company, and we might say that it is a harsh punishment. But the risk of having to pay $6,600 in damages won't keep the railway company from defending its case in court. If I am accused of having copied twenty songs and given the choice of going to court with a risk of being sentenced to pay $200,000, or paying $5000 to a blackmailer, going to court may be too risky for me, so the amount of damages prevents me from going to court.
Except that can't be so. Apple will sell you a Mac MIni for $599 with a full copy of OS X. There is several hundred dollars of value in the hardware plus the Apple insane profit margin. So the value of the software simply can't be $399.
The value of a copy of MacOS X running on a MacMini is not the same as the value of a copy running an an 8 core Mac Pro.
Err, They err, they do have UPGRADE written on the boxes.
Actually, they don't. And Apple's license doesn't mention "upgrade" anywhere. What it says is that you are allowed to install _one_ copy of MacOS X on _one_ Apple-labeled computer. Now in practice this means that the installation will be an upgrade, because you will have a very hard time finding a computer that is Apple-labeled, is capable of running MacOS X, and wasn't originally sold with a version of MacOS X installed.
On the other hand, this doesn't make much difference. Either customers have to follow the terms of the license, or they don't. In either case, whether the restriction is "upgrade only" or "Apple-labeled computer only", it is equally valid or equally invalid.
Except we know the value of the software, Apple sells it seperately for $130, or about what an OEM edition of Vista Business adds to a typical Dell.
It is astonishing how nonsense like this is modded up as "insightful".
$129 is Apple's special offer price for people who have (i) paid Apple good money for a previous version of the software and have (ii) paid Apple good money for Macintosh hardware. Apple doesn't at the moment sell a version for generic PCs, which many people find regrettable. There can be no doubt that Apple wouldn't sell such a version for $129. My estimate is that $399 would more correctly represent the value for a full version of MacOS X that can be installed on any computer.
He is a very experienced business man, so we can assume that he is right, an an Apple computer costs $500 more than a Windows PC with the same "spec". On the other hand, we know that Apple has increased its unit market share in the computer business year over year over year, so there must be something that convinces people to pay $500 more for the same spec, which means they believe that a Mac is worth $500 more than a PC with a same spec. Now money is somehow tighter at the moment, and some people will buy cheaper even though they know they get better value for more money. This doesn't change the fact that a Macintosh, by Ballmer's own admission, looks worth $500 more.
So where does this difference from? One part is design. If you just take the looks, I would pay more for a computer that looks like a white MacBook than for a computer that looks like a Dell, if they were otherwise identical. And I'd pay more again for a computer with the aluminium MacBook design. Then there is quality + service. I think a Mac is less likely to break down, and Apple's service is outstanding. That's worth money. Then there is the Apple logo. Yes, people will pay a little bit more just for having an "Apple" logo than a "Dell" logo. Obviously not $500, but maybe $20, maybe $50. Let's add this together: I think Apple could sell their computers with Windows pre-installed and no MacOS X license and charge $200 more than another brand with same spec and find customers to buy it.
But the difference is $500, not $200. Here is where Mr. Ballmer should take note: $300 of the difference in value between a Mac and a PC with the same spec is _the operating system_. Yes, Mr. Ballmer, people think that MacOS X is $300 worth more than Windows Vista. This is quite damning for Microsoft. Now maybe he is right. Maybe people will have to save money and go with the substandard Windows instead of MacOS X. Maybe he is wrong. What stays right: MacOS X is worth $300 more.
I think what we've learned is that there's a bug in the POSIX standard,...
It is not exactly a bug in the standard. There is a standard, and there is QOI (Quality of Implementation). When you write data, the Posix says that the data is vulnerable for a time interval of unknown length. A good implementation will replace "unknown length" with "length zero", or "length almost zero". ext4 decided that "unknown length" can mean "two minutes". QOI = zero.
I understand that doing writes immediately when requested leads to performance degradation but that is why business systems which defer writes to disk only do so when the hardware can guarantee it. In other words, we have a battery backed cache, if the battery is low or nearing end of life the cache is turned off and all writes are made when the data changes.
You don't even need to do this. The reported problem happened (I think) during some installation of five hundred files. The computer crashed just after the installation was finished, at a time when half the changes were written to disk. If the computer had crashed _before_ the installation started, everything would have been fine. If the computer had delayed _all_ writes by two minutes, and the computer crashed a minute after it said "installation finished", but before anything was actually written to disk, everything would have been fine (Ok, you would have to repeat the installation process, but that is no problem).
What the file system must do is do a bunch of changes together that belong together, and minimize the time interval where a crash would have bad results, preferably to zero.
If this is as described a small ISP with 400 customers whose bandwidth use is right at the limit most of the time, then throttling is already implemented. Automatically. By the ISPs upstream provider. So if customers would be unhappy because of throttling, then they are unhappy already. If there are contract problems because unlimited service was promised, then these contract problems are already there.
And as described, this is a small sideline of the companies business, so anything that will keep their lawyers busy, like contract changes, won't fly. Anything that is a major investment most likely won't fly. The only thing that could fly is anything that either makes money, or significantly improves the reputation of the company which could have other positive side effects.
Since Megabits are limited in this situation, his boss is absolutely right that the only thing he can do is to maximise the number of _happy_ customers. And that would be maximised by throttling the heavy users, giviing low bandwidth users fast access whenever they need it.
From the user's point of view: As a group, they pay 400x dollars per month to the ISP, who for that money gives them a total bandwidth with some limit. As a group, they don't want to include anyone who uses tons more than their fair 1/400th share.
Thus we arrive at what is without a doubt the single worst product that apple has ever released.
No, the puck mouse still has the nr. 1 place. The new iPod shuffle is at least usable, but it definately comes close though.
Nowhere near close. The puck mouse did exactly what it was supposed to do. This Apple product _may_ be their worst ever, but maybe someone knows something worse:
http://support.apple.com/kb/TA45469?viewlocale=en_US
This was a tape backup device with 38.5 MB storage capacity. The Macintosh II at the time shipped with a 40 MB hard drive, so the tape was too small. You couldn't backup your hard drive on a single tape. Except if you stored your backup as individual files, in which case the backup time was so bad, it wouldn't be finished in the morning if you started in the evening - it used a tape drive to simulate a direct access device, with seek times in minutes. I bet 99.9% of its users tried it once and gave up.
If you really care about music you'll take your ipod whatever back to the store and smash it to bits on the counter with your shoe.
Apple will probably not mind if you buy as many $79 iPods for $79 and smash them to bits on the counter, as long as you clean up the mess behind you and as long as your credit card is Ok.It's all revenue and therefore profit.
I see another problem with this and that is that it shoots a hole in the GPS licensing. What good is open source if you still have to pay royalties to patent trolls in order to use it?
What do patent trolls have to do with anything? Let's say you are young enough to still go to school, and there is a school bully who threatens to beat up anyone who uses a computer that doesn't run Windows. So what good is GPL licensing? (I assume you meant GPL and not GPS). What good is a license to MacOS X, when Apple can't protect you from getting beaten up?
This whole GPL angle on the TomTom case is nonsense. TomTom uses Linux under the GPL license. Linux is either infringing on Microsoft's patent, or it isn't. If it is, that is not TomTom's fault. So TomTom gets blackmailed. They either pay or they don't. Whether they pay or they don't doesn't affect whether Linux is infringing on Microsoft's patent or not. Payment doesn't mean that TomTom admits Microsoft's patents are valid, it means they want to avoid a court case. Even if TomTom admits Microsoft's patents are valid, that isn't binding on anyone.
As long as TomTom puts all the code on their website, and doesn't itself add restrictions to its use, I can't see how they would be violating the GPL. Sure, they can tell you about this bully boy who forced them to pay money, and the bully boy could go after you as well. But the patent infringement, if there is one, is there in all Linux versions.
That's all very well, but in most cases, the machine is a single-user desktop, and it's sitting *idle* during the 5-40 seconds in which the filesystem writes are being batched up. It seems to me that the writes should be sent to disk immediately, unless the disk is already busy, in which case they may be delayed for (at most) 40 seconds, while other reads take place ahead of them.
The write delay is not the problem. It is only a problem because the file system is fucked up and a crash at the wrong moment messes things up, and the write delay increases the interval that is critical.
You have to assume that there can be a crash. You have to assume that this crash can happen at any point in time (for example, just before a call to fsync for the fsync fanatics). Apparently KDE opened, truncated, wrote, and closed a few hundred files. The problem isn't that this happened 120 seconds delayed instead of happening immediately, the problem was that the stupid file system managed to do the truncates immediately and the writes 120 seconds later. If the _whole_ operation had been delayed, nothing bad would have happened.
The filesystem doesn't guarantee anything is written until you've called fsync and it has returned.
Which may be true, but is missing the point. Your computer can crash just before fsync is called, so relying on fsync doesn't really help if you want to ensure that the data has been written. It only helps if you need to _know_ that it is written (for example, if you want to delete the original after the copy has been safely written).
They are referring to the case when the system isn't shut down cleanly. This means a kernel crash or a power outage. What is your point exactly? Seriously, and I really am doing my best to hold back on the personal insults (even when you something as annoying as "And calm down !!"), what is so difficult that you fail to comprehend what the real issue being discussed here is?
The real issue is that the same code and the same external event lead to data loss when using ext4, and don't lead to data loss when using ext3. Reported as a bug often enough that it is likely to affect any user eventually, and developers trying to explain why it's not their fault instead of trying to fix the problem.
I'd say the ext4 developers are 100 percent correct, and I would recommend to anyone not to use their product.
As an application developer, you are expected to know what the API does, in order to use it correctly. What Ext4 is doing is 100% respectful of the spec.
Respectful my arse.
So: Use ext4, and stuff tends to disappear. Use something else, and stuff doesn't tend to disappear. It doesn't matter what excuses ext4 has to put the blame somewhere else, fact is that stuff disappears. If people don't want stuff to disappear, they'll have to stay away from ext4.
It's very simple. Lots of application software is badly written. When this badly written software uses ext4, Bad Things Happen. As a user you have two choices: Don't use that software, or don't use ext4. To me, the decision is clear. There are other file systems that don't lead to these problems, so any reasonable user will avoid ext4 like the plague.
So the way I understand these comments, the file system has been written to be very fast by delaying certain operations, and it succeeds, except that in case of a crash your hard drive is in a very undesirable state. Programmers can do something about this, but the consequence is that performance drops down through the floor. So the file system is fast with unsafe applications, and dead slow for safe operations. Nice.
The layoffs that would occur as all developed software immediately entered the public domain because there was no legal recourse for copying binaries or code,
In Germany, as a concrete example, software could not be protected by copyright for many years. Guess what: Lots of software got developed and sold successfully by its creators, due to the power of competition laws. It is illegal to compete with me by copying my software at zero expense when I did the hard work of creating the software. On the other hand, the protection goes away the second that I stop selling, because once I leave the market voluntarily, you are not competing with me anymore.
Translated into music and books, Elvis and his record company would have the right to sell his records forever, and would be protected from competitors. As soon as they stop selling, others can take the works and sell them, because then they are not competing.
Slashdot Mirror
So according to this, if I had a competitor I don't want around I could just buy their app a lot and keep getting refunds until I bleed them dry? Sounds good to me.
If you don't like that competitor, why don't you go to his premises with a canister full of petrol and burn everything down? Simple: You don't want to go to jail for arson. And why don't people do what you suggest? Simple. They don't want to end up in court for tortious interference with a business.
So how does the developer of a pay application prevent someone from doing a DoS on the developer's bank account by asking readers of his blog to buy the app and get a refund?
By taking the blogger to court for tortious interference with his business. See Wikipedia: http://en.wikipedia.org/wiki/Tortious_interference or google for any other sources.
So, let's say there's a game on your iPhone, what would be the expected total time to finish it and get bored with it? I bet less than 90 days. So you can basically have a free game: download, finish, refund. It's golden!
Except that you can't. The contract between Apple and the developer says that _if_ Apple refunds the money then they get some amount back from the developer. However, you are a store customer, and that contract has nothing to do with you. The contract between _you_ and Apple says: No refunds.
I'm pretty sure the I in raid stands for "inexpensive". Isn't the whole point of RAID to avoid paying extra for high performance disks?
RAID is inexpensive compared to what you would have to pay for a single drive with the same performance.
Earlier someone suggested sneezing/coughing, covering her mouth with a gloved hand, and then using the same gloved hand to pack swabs. Hard to avoid this unless she works in an enclosed suit (which is unlikely.)
I would assume that the police or their lab don't buy swabs used for DNA testing at the nearest supermarket, they will buy them from a supplier who guarantees uncontaminated swabs, for obvious reasons. To guarantee uncontaminated swabs, that supplier would have to train its employees making the swabs. Part of the training would be to tell them: If you cough or sneeze, and touch your mouth with your gloved hand, you then leave the production line immediately, taking any materials with you that could have been contaminated, destroy them, desinfect your hands, and put on fresh gloves. So either they didn't train their employees properly, or they put pressure on their employees so they couldn't work as needed, or they hired employees who give a shit. Of course this makes the swabs a bit more expensive, but I am sure the police pays more for "guaranteed uncontaminated" swabs then you pay when you want to clean your ears.
In any case, I would like to see what kind of contracts they had, and the company might be liable for a lot of damages.
. I'm curious to know the battery life difference if anyone has done an experiment with their player?
I think on hard-drive based iPods, the battery doesn't actually last for x hours, it lasts for x Megabytes. Use 256 Kbit AAC instead of 128 Kbit, half the battery life. Lossless is obviously worse.
For any operating system using a virtual address space (any modern operating system), this isn't an issue in the slightest. The memory given to a program or process can appear to be a single block to the process, but not have to be physically contiguous thanks to MMU magic.
Missing the point completely. A call to malloc (256 * 1024 * 1024) must return memory that is contiguous in virtual address space. And once that malloc call succeeded, the memory can't be moved in virtual address space. And virtual address space can easily get fragmented.
Assume 15 x 257 MB virtual address space. Use malloc to allocate 1MB, 256MB, 1MB, 256MB, etc. until memory is full. free () all the 256 MB blocks. You end up with fragmented memory with no available block greater than 256 MB. Try allocating 257 MB -> fail.
This also misses the obvious problem that an ID3 tagger might be written to read all the tags from my 10000 songs into memory, without any worries that it might not fit. And suddenly this software throws up on a few dozen files in the new format.
In this case, the MP3 signal serves as the prediction and the remaining data is the correction stream.
Complete bollocks. The first thing you do in MP3 compression is throwing away the phase information, which immediately removes half the data with no quality loss. That makes the compressed signal completely unusable for use as a prediction stream.
Think of replacing sin (omega times t + alpha) with sin (omega times t), where t is the time. There is no audible difference between the signals, but the values are absolutely completely different.
Those damage ratios are exactly what these filings are about. The judge has been asked to consider if such statutory damages may violate cruel and unusual punishment restriction by being so far out of line. Part of the judge's determination would be which cost if any to compare the damages with.
The damage ratio is not actually the problem. The problem is that the damages are so high that the average citizen cannot go to court, because losing the case would bankrupt them. If a railway company overcharges you buy 66 cents and is ordered to pay 10,000 times the amount in damages, that is surely very annoying for the railway company, and we might say that it is a harsh punishment. But the risk of having to pay $6,600 in damages won't keep the railway company from defending its case in court. If I am accused of having copied twenty songs and given the choice of going to court with a risk of being sentenced to pay $200,000, or paying $5000 to a blackmailer, going to court may be too risky for me, so the amount of damages prevents me from going to court.
Except that can't be so. Apple will sell you a Mac MIni for $599 with a full copy of OS X. There is several hundred dollars of value in the hardware plus the Apple insane profit margin. So the value of the software simply can't be $399.
The value of a copy of MacOS X running on a MacMini is not the same as the value of a copy running an an 8 core Mac Pro.
Err, They err, they do have UPGRADE written on the boxes.
Actually, they don't. And Apple's license doesn't mention "upgrade" anywhere. What it says is that you are allowed to install _one_ copy of MacOS X on _one_ Apple-labeled computer. Now in practice this means that the installation will be an upgrade, because you will have a very hard time finding a computer that is Apple-labeled, is capable of running MacOS X, and wasn't originally sold with a version of MacOS X installed.
On the other hand, this doesn't make much difference. Either customers have to follow the terms of the license, or they don't. In either case, whether the restriction is "upgrade only" or "Apple-labeled computer only", it is equally valid or equally invalid.
Except we know the value of the software, Apple sells it seperately for $130, or about what an OEM edition of Vista Business adds to a typical Dell.
It is astonishing how nonsense like this is modded up as "insightful".
$129 is Apple's special offer price for people who have (i) paid Apple good money for a previous version of the software and have (ii) paid Apple good money for Macintosh hardware. Apple doesn't at the moment sell a version for generic PCs, which many people find regrettable. There can be no doubt that Apple wouldn't sell such a version for $129. My estimate is that $399 would more correctly represent the value for a full version of MacOS X that can be installed on any computer.
He is a very experienced business man, so we can assume that he is right, an an Apple computer costs $500 more than a Windows PC with the same "spec". On the other hand, we know that Apple has increased its unit market share in the computer business year over year over year, so there must be something that convinces people to pay $500 more for the same spec, which means they believe that a Mac is worth $500 more than a PC with a same spec. Now money is somehow tighter at the moment, and some people will buy cheaper even though they know they get better value for more money. This doesn't change the fact that a Macintosh, by Ballmer's own admission, looks worth $500 more.
So where does this difference from? One part is design. If you just take the looks, I would pay more for a computer that looks like a white MacBook than for a computer that looks like a Dell, if they were otherwise identical. And I'd pay more again for a computer with the aluminium MacBook design. Then there is quality + service. I think a Mac is less likely to break down, and Apple's service is outstanding. That's worth money. Then there is the Apple logo. Yes, people will pay a little bit more just for having an "Apple" logo than a "Dell" logo. Obviously not $500, but maybe $20, maybe $50. Let's add this together: I think Apple could sell their computers with Windows pre-installed and no MacOS X license and charge $200 more than another brand with same spec and find customers to buy it.
But the difference is $500, not $200. Here is where Mr. Ballmer should take note: $300 of the difference in value between a Mac and a PC with the same spec is _the operating system_. Yes, Mr. Ballmer, people think that MacOS X is $300 worth more than Windows Vista. This is quite damning for Microsoft. Now maybe he is right. Maybe people will have to save money and go with the substandard Windows instead of MacOS X. Maybe he is wrong. What stays right: MacOS X is worth $300 more.
I think what we've learned is that there's a bug in the POSIX standard, ...
It is not exactly a bug in the standard. There is a standard, and there is QOI (Quality of Implementation). When you write data, the Posix says that the data is vulnerable for a time interval of unknown length. A good implementation will replace "unknown length" with "length zero", or "length almost zero". ext4 decided that "unknown length" can mean "two minutes". QOI = zero.
I understand that doing writes immediately when requested leads to performance degradation but that is why business systems which defer writes to disk only do so when the hardware can guarantee it. In other words, we have a battery backed cache, if the battery is low or nearing end of life the cache is turned off and all writes are made when the data changes.
You don't even need to do this. The reported problem happened (I think) during some installation of five hundred files. The computer crashed just after the installation was finished, at a time when half the changes were written to disk. If the computer had crashed _before_ the installation started, everything would have been fine. If the computer had delayed _all_ writes by two minutes, and the computer crashed a minute after it said "installation finished", but before anything was actually written to disk, everything would have been fine (Ok, you would have to repeat the installation process, but that is no problem).
What the file system must do is do a bunch of changes together that belong together, and minimize the time interval where a crash would have bad results, preferably to zero.
If this is as described a small ISP with 400 customers whose bandwidth use is right at the limit most of the time, then throttling is already implemented. Automatically. By the ISPs upstream provider. So if customers would be unhappy because of throttling, then they are unhappy already. If there are contract problems because unlimited service was promised, then these contract problems are already there.
And as described, this is a small sideline of the companies business, so anything that will keep their lawyers busy, like contract changes, won't fly. Anything that is a major investment most likely won't fly. The only thing that could fly is anything that either makes money, or significantly improves the reputation of the company which could have other positive side effects.
Since Megabits are limited in this situation, his boss is absolutely right that the only thing he can do is to maximise the number of _happy_ customers. And that would be maximised by throttling the heavy users, giviing low bandwidth users fast access whenever they need it.
From the user's point of view: As a group, they pay 400x dollars per month to the ISP, who for that money gives them a total bandwidth with some limit. As a group, they don't want to include anyone who uses tons more than their fair 1/400th share.
Thus we arrive at what is without a doubt the single worst product that apple has ever released.
No, the puck mouse still has the nr. 1 place. The new iPod shuffle is at least usable, but it definately comes close though.
Nowhere near close. The puck mouse did exactly what it was supposed to do. This Apple product _may_ be their worst ever, but maybe someone knows something worse:
http://support.apple.com/kb/TA45469?viewlocale=en_US This was a tape backup device with 38.5 MB storage capacity. The Macintosh II at the time shipped with a 40 MB hard drive, so the tape was too small. You couldn't backup your hard drive on a single tape. Except if you stored your backup as individual files, in which case the backup time was so bad, it wouldn't be finished in the morning if you started in the evening - it used a tape drive to simulate a direct access device, with seek times in minutes. I bet 99.9% of its users tried it once and gave up.
If you really care about music you'll take your ipod whatever back to the store and smash it to bits on the counter with your shoe.
Apple will probably not mind if you buy as many $79 iPods for $79 and smash them to bits on the counter, as long as you clean up the mess behind you and as long as your credit card is Ok.It's all revenue and therefore profit.
I see another problem with this and that is that it shoots a hole in the GPS licensing. What good is open source if you still have to pay royalties to patent trolls in order to use it?
What do patent trolls have to do with anything? Let's say you are young enough to still go to school, and there is a school bully who threatens to beat up anyone who uses a computer that doesn't run Windows. So what good is GPL licensing? (I assume you meant GPL and not GPS). What good is a license to MacOS X, when Apple can't protect you from getting beaten up?
This whole GPL angle on the TomTom case is nonsense. TomTom uses Linux under the GPL license. Linux is either infringing on Microsoft's patent, or it isn't. If it is, that is not TomTom's fault. So TomTom gets blackmailed. They either pay or they don't. Whether they pay or they don't doesn't affect whether Linux is infringing on Microsoft's patent or not. Payment doesn't mean that TomTom admits Microsoft's patents are valid, it means they want to avoid a court case. Even if TomTom admits Microsoft's patents are valid, that isn't binding on anyone.
As long as TomTom puts all the code on their website, and doesn't itself add restrictions to its use, I can't see how they would be violating the GPL. Sure, they can tell you about this bully boy who forced them to pay money, and the bully boy could go after you as well. But the patent infringement, if there is one, is there in all Linux versions.
That's all very well, but in most cases, the machine is a single-user desktop, and it's sitting *idle* during the 5-40 seconds in which the filesystem writes are being batched up. It seems to me that the writes should be sent to disk immediately, unless the disk is already busy, in which case they may be delayed for (at most) 40 seconds, while other reads take place ahead of them.
The write delay is not the problem. It is only a problem because the file system is fucked up and a crash at the wrong moment messes things up, and the write delay increases the interval that is critical.
You have to assume that there can be a crash. You have to assume that this crash can happen at any point in time (for example, just before a call to fsync for the fsync fanatics). Apparently KDE opened, truncated, wrote, and closed a few hundred files. The problem isn't that this happened 120 seconds delayed instead of happening immediately, the problem was that the stupid file system managed to do the truncates immediately and the writes 120 seconds later. If the _whole_ operation had been delayed, nothing bad would have happened.
The filesystem doesn't guarantee anything is written until you've called fsync and it has returned.
Which may be true, but is missing the point. Your computer can crash just before fsync is called, so relying on fsync doesn't really help if you want to ensure that the data has been written. It only helps if you need to _know_ that it is written (for example, if you want to delete the original after the copy has been safely written).
They are referring to the case when the system isn't shut down cleanly. This means a kernel crash or a power outage. What is your point exactly? Seriously, and I really am doing my best to hold back on the personal insults (even when you something as annoying as "And calm down !!"), what is so difficult that you fail to comprehend what the real issue being discussed here is? The real issue is that the same code and the same external event lead to data loss when using ext4, and don't lead to data loss when using ext3. Reported as a bug often enough that it is likely to affect any user eventually, and developers trying to explain why it's not their fault instead of trying to fix the problem.
I'd say the ext4 developers are 100 percent correct, and I would recommend to anyone not to use their product.
As an application developer, you are expected to know what the API does, in order to use it correctly. What Ext4 is doing is 100% respectful of the spec.
Respectful my arse.
So: Use ext4, and stuff tends to disappear. Use something else, and stuff doesn't tend to disappear. It doesn't matter what excuses ext4 has to put the blame somewhere else, fact is that stuff disappears. If people don't want stuff to disappear, they'll have to stay away from ext4.
It's very simple. Lots of application software is badly written. When this badly written software uses ext4, Bad Things Happen. As a user you have two choices: Don't use that software, or don't use ext4. To me, the decision is clear. There are other file systems that don't lead to these problems, so any reasonable user will avoid ext4 like the plague.
So the way I understand these comments, the file system has been written to be very fast by delaying certain operations, and it succeeds, except that in case of a crash your hard drive is in a very undesirable state. Programmers can do something about this, but the consequence is that performance drops down through the floor. So the file system is fast with unsafe applications, and dead slow for safe operations. Nice.
The layoffs that would occur as all developed software immediately entered the public domain because there was no legal recourse for copying binaries or code,
In Germany, as a concrete example, software could not be protected by copyright for many years. Guess what: Lots of software got developed and sold successfully by its creators, due to the power of competition laws. It is illegal to compete with me by copying my software at zero expense when I did the hard work of creating the software. On the other hand, the protection goes away the second that I stop selling, because once I leave the market voluntarily, you are not competing with me anymore.
Translated into music and books, Elvis and his record company would have the right to sell his records forever, and would be protected from competitors. As soon as they stop selling, others can take the works and sell them, because then they are not competing.