LIES! OS X is NOT UNIX. It's sorta-kinda-BSD-abomination-whatever-Mach. It's arguably a lot less UNIX than Linux, which is specifically not UNIX. The only UNIXy things about it are the GNU parts, and GNU's Not UNIX.
MacOS X is POSIX 2003 compliant, 100%. Linux isn't anywhere near. So yes, MacOS X is Unix.
Maybe because bugs that allow jail break are pretty damn serious as they allow for complete control of the phone. Other bugs that don't allow jail break aren't as critical.
There's also different levels of bugs that allow jail break. Any bug that allows a remote jail break (jail break by visiting a website) is absolutely security critical. If someone can make a website that does a jail break, they can make a website that takes over your phone completely. I think that happened once and was fixed very quickly.
A bug that allows a jail break by connecting your phone to your computer and running software on your computer that you voluntarily installed would be much less serious. If you can jail break your phone, but I can't hack into it, that's much less of a problem.
But then once a vulnerability is there, it's very hard to say how it could or could not be exploited. So while a vulnerability with a known remote use (jail break or malware) must be fixed immediately, _any_ vulnerability _must_ be fixed.
Citation needed. There's no requirement. In fact, most fonts have most of Unicode missing. It'd be insane to try to cover the entirety of Unicode with each new font released. The most complete font I've found is Arial Unicode MS, but even that has vast swaths of Unicode missing.
A font represents many characters in a consistent, unique style. Once you have characters that are different enough, trying to have a unique style doesn't make sense. Cyrillic and Latin, that makes sense. You can't really have lowercase greek characters in Arial style, and definitely not Arabic. Just doesn't make sense.
You don't lose any rights by false requests. Here's what's illegal: You have to state under threat of perjury that you are the copyright holder or represent the copyright holder of some work. If the work that you claim isn't actually the one that is uploaded (for example if NASCAR believed that you uploaded their official video, but you uploaded one you shot yourself), that isn't punished. If they claim that you uploaded a video made by some TV crew and they own the copyright, but in reality the TV station and not NASCAR owned the copyright, that would be perjury and punished accordingly.
If your achievements are always short of your goals, doesn't that make you a loser?
Of course not. If A has the goal to make a million, and makes only $900,000, and B has the goal to make $10,000 and makes $12,000, who is the loser?
The _proper_ use of goals is to use them as motivational tools. The best way to set a goal is to set it so high, that it looks just barely achievable. Such a goal will usually _not_ be achieved, but leads to the highest _actual_ achievement. Setting goals that are easily achieved leads to less actual achievement (while meeting the goal), while goals that are too hard make people not even try.
The claim (verified by the numbers in the summary!) is that crime wouldn't have risen if not for apple thefts.
The claim is actually that if we took the number of all crimes, and subtracted the number of iPhone thefts, the result would be less than the number of the previous years. And that claim is true, but meaningless.
There were many more iPhones around in 2012 than 2011. The percentage of iPhones among all phones is higher, so it would have to be expected that the number stolen would be higher. If Apple hadnl't sold any phones, almost everyone buying an iPhone would have bought some other phone, and all those iPhone thieves would have stolen some different phone. The crime rate would be exactly the same.
In 2000, the percentage of stolen cars built in the 21st century was zero (some people disagree, but they are clueless). In 2001, it grew slightly, and now it is close to 100%. If it wasn't for these pesky 21st century cars, car theft would be almost down to zero.
So the real owner gets blamed for the gas theft etc. - until it becomes obvious that there's two almost identical cars with the same license plate.
My thoughts on that: 1. There should be a register of suspected duplicated license plates, similar to stolen cars. You should be able to put yourself on that register, possibly adding what's your way to work. A car should be flagged when it's in the register like a stolen car, except a police officer would know that chances are at least 50% that the car is legit. 2. If person X legitimately owns a license plate, then they should be able to demand to be handed over any car using that license plate.
Ever since I first got involved in fighting the RIAA's litigation campaign, and blogging about it, in 2005 [that's almost 8 years ago] I've been arguing that it is not a sufficient basis to bring a lawsuit against someone that an internet access account for which he or she pays the bill was used by someone for a copyright infringement.
Just a few days ago, there was this discussion here on Slashdot about some guy who can't manage to keep a leech of his network:
The copyright lobby group is now trying to say "no no, it's parody" but they've been lobbying hard to get parody removed from the list of exemptions and have recently succeeded.
Microsoft wouldn't be happy if you used a pirated copy of Microsoft Word to write a parody of Harry Potter. These guys try to create a parody of a website, not a parody of someone's CSS code.
Apple shares are in freefall falling from $705 a share to $450 and deservedly so. It is no longer the largest company in the world by market cap.
Actually, it is. For some strange reason, it was heavily reported when Apple fell behind Exxon, but nobody reported when Apple overtook Exxon again.
If you subtract cash from market caps, the enterprise value of Apple right now is only about 6 1/2 years of profits. That means the company is right now ridiculously undervalued.
No Apple has been growing slower than the industry...in fact its shrinking compared to the market.
Absolutely not. The percentage of iPhones in the phone market is growing. Actually, the phone market has been slightly shrinking in the last year. However, apparently it makes some people on Slashdot happy to declare "smartphones" to be _the_ market. Since $80 feature phones are being replaced with $80 "smartphones", the percentage of smartphones in the total phone market is growing. That doesn't help the sellers, when for every $80 smartphone they are selling they lose the sale of a $80 feature phone.
Apple on the other hand has been growing its phone market share and iPhone unit sales steadily year after year, and there is no reason why that should change in the future.
Can I just say that repeating the nonsense of "patenting shit like rounded corners" marks you as either a troll or an imbecile.
Apple didn't patent rounded corners. Apple has a _design patent_ for a design consisting of many items, one of them rounded corners. To infringe on this _design patent_, you have to copy the complete design, every single item listed in the design patent. You can have as many rounded corners as you like. As long as your design is in some way different from Apple's design patent.
Here's for your enjoyment an example of Samsung patenting rounded corners:
Yep. Judge people by their actions not their words.
Sometimes that happens. Sometimes it doesn't. As a result, Samsung is supposed to pay $1bn for violating Apple's patents, but a huge leak of hydrofluoric acid at Samsung's plant that killed one worker, injured four and according to police reports may have affected thousands, got them a one thousand dollar fine.
Non-commercial use: Any number of Macs that you own and control. Commercial use: Any number of computers used by the same single person, or one computer used by any number of persons.
One computer and can't move to a different computer? That's ridiculous. So if sell your computer and buy a better one, you have to re-buy the software? Or if your computer breaks? Or your computer is stolen? I wonder what your insurance company will say if your computer is stolen, they pay for a replacement, and then you say that instead of restoring your apps from your backup you want them to pay for new copies?
So if you're transferring a file over the network maxing it out and you know it will take 10 minutes but then your wife starts streaming video over the network, how do you predict that at the start of the transfer? It's pretty damn difficult to predict the future man.
I'll give you a different scenario: I start downloading a huge file in the browser, say 800MB. After the first 100MB is done, I start downloading a second huge file in the same browser, say 400MB. How well can you predict when each download will finish?
Another scenario that browsers usually get horribly wrong: I click on a download button, but it takes 10 seconds until the download starts. After 10.01 seconds, the browser calculates average number of bytes per seconds which is ridiculously low and tells me that the download will take 3 weeks. That number goes down rapidly to something reasonable. Why can't that be improved?
Another one that I saw: A download of a 500 MB file was interrupted after 200 MB. When the download is restarted, the browser figures out after 0.01 seconds that 200 MB are already there, so the whole file will take 0.025 seconds. Why can't that be improved?
Same defeatist attitude. There are people who would spend hours and hours arguing why something can't be done, instead of just doing it.
For your comparison between HD and SSD: Yes, on an SSD the time to access a file is much smaller relative to the time for reading bytes, compared to a hard drive. That's why you measure it. You don't need to know the technology, just the numbers. I'll admit that the maths is slightly above kindergarten level, and apparently above slashdot level, but it's not really hard.
If you managed to read the first file of 1MB after 10 milliseconds, and you read the second file of 0.5 MB after 7 milliseconds, how long will it take to read a third file of 2 MB? There are two stupid estimates that have been mentioned here again and again: 1. Based on the number of bytes - you read 1.5MB in 17ms, so 2MB take 22.67ms. 2. Based on the number of files - you read two files in 17ms, so the third file takes 8.5ms. Here is the mathematical bit: Assume that the overhead per file is the same as reading k bytes. We get the best match for the data if we assume 4ms overhead, and 6ms per MB, or k = 666,666 - we actually get a perfect match for the data so far. So we had 1.5MB data and 1.333MB of overhead in 17ms, we read 2MB data plus 0.667MB of overhead, and the calculation gives 16 ms. What are the bets that 16ms is a better prediction than either 8.5ms or 22.67ms? Especially since the 8.5ms estimate is less than the time for the 1MB file, and 22.67ms is more than twice the time for the 1MB file.
Network latency, seek times, sudden Windows Update that decides to start installing packages in the middle of the process.. there's a BILLION different things that could be happening, all the kinds of things that drastically alter the speeds and therefore throwing the estimate right off the table.
So why is it so hard to show a progress bar that is reasonably accurate if none of these things happen? Nobody is complaining that a progress bar is inaccurate if Windows starts installing updates. People complain when its inaccurate no matter what happens.
All these constants go out of the window if anything else starts accessing the drive as well, of course, and they're actually variable based on file system and a variety of other things.
What a totally defeatist attitude. Here's how it works currently: There are x MB to copy. So far, y MB have been copied in z seconds. We estimate the number of seconds per MB = z / y, and estimate the time we need as x * (z / y). That's repeated during the copying. Doesn't matter if someone else accesses the drive, doesn't matter how fast or slow the drive is.
Now here comes the ingenious change: There are n files of x MB to copy. So far, m files with y MB have been copied within z seconds. We assume that copying one file has an overhead equivalent to copying k bytes. Initially we assume k = 0 since we know nothing else, but as progress is made and files of different sizes are copied, we can find which k matches the available data best. Then we go through the original algorithm, but replacing x with x + kn, and y with y + km. Easy peasy if your IQ is just slightly above average. Your's apparently isn't.
To put it another way, it is impossible to make an accurate progress bar because it is impossible to predict the future. That's all.
A lot is just stupidity. When you copy lots of files on a hard drive, the time is proportional to some constant times total file size, plus some other constant times number of files. Lots of small files take longer than their size indicates. Most progress bars don't take that into account, so you make huge progress through a dozen 20 MB files, and then slow down through a lot of tiny files. Accounting for that alone would be a huge improvement.
MacOS X writes: "Size on disk: 4 KB" in that case, not bothering with decimals. For a bigger file, I found "240,831 bytes (242 KB on disk)". The 242 KB is 240,831 bytes, rounded up to a multiple of 4,096 bytes which is the size required on disk (241,664 bytes), then rounded to the nearest integer (242 KB) and not 236 KiB.
I was going to say, the physical world doesn't really care which base unit we used to count with, but computers are very particular about using base 2. When a computer is using integer math to calculate memory addresses in base two and you have a base 10 amount of memory, bad stuff will happen.
And what on earth does that have to do with displaying the correct amount of memory to the user? Is there something that will crash my computer if it displays "4.29 billion bytes of RAM"?
Slashdot Mirror
whos fuckin stopping me? apple? fuck them. ill do it anyway.
It's a DMCA violation, which means in any serious company your own company lawyers will eat you alive if you do it and they find out.
LIES! OS X is NOT UNIX. It's sorta-kinda-BSD-abomination-whatever-Mach. It's arguably a lot less UNIX than Linux, which is specifically not UNIX. The only UNIXy things about it are the GNU parts, and GNU's Not UNIX.
MacOS X is POSIX 2003 compliant, 100%. Linux isn't anywhere near. So yes, MacOS X is Unix.
US government & corporate groups are hacking too. I cant wait for the "blah blah no proof" bullshit next.
More likely that the Chinese count every spam email as "hack attack".
Maybe because bugs that allow jail break are pretty damn serious as they allow for complete control of the phone. Other bugs that don't allow jail break aren't as critical.
There's also different levels of bugs that allow jail break. Any bug that allows a remote jail break (jail break by visiting a website) is absolutely security critical. If someone can make a website that does a jail break, they can make a website that takes over your phone completely. I think that happened once and was fixed very quickly.
A bug that allows a jail break by connecting your phone to your computer and running software on your computer that you voluntarily installed would be much less serious. If you can jail break your phone, but I can't hack into it, that's much less of a problem.
But then once a vulnerability is there, it's very hard to say how it could or could not be exploited. So while a vulnerability with a known remote use (jail break or malware) must be fixed immediately, _any_ vulnerability _must_ be fixed.
Citation needed. There's no requirement. In fact, most fonts have most of Unicode missing. It'd be insane to try to cover the entirety of Unicode with each new font released. The most complete font I've found is Arial Unicode MS, but even that has vast swaths of Unicode missing.
A font represents many characters in a consistent, unique style. Once you have characters that are different enough, trying to have a unique style doesn't make sense. Cyrillic and Latin, that makes sense. You can't really have lowercase greek characters in Arial style, and definitely not Arabic. Just doesn't make sense.
You don't lose any rights by false requests. Here's what's illegal: You have to state under threat of perjury that you are the copyright holder or represent the copyright holder of some work. If the work that you claim isn't actually the one that is uploaded (for example if NASCAR believed that you uploaded their official video, but you uploaded one you shot yourself), that isn't punished. If they claim that you uploaded a video made by some TV crew and they own the copyright, but in reality the TV station and not NASCAR owned the copyright, that would be perjury and punished accordingly.
If your achievements are always short of your goals, doesn't that make you a loser?
Of course not. If A has the goal to make a million, and makes only $900,000, and B has the goal to make $10,000 and makes $12,000, who is the loser?
The _proper_ use of goals is to use them as motivational tools. The best way to set a goal is to set it so high, that it looks just barely achievable. Such a goal will usually _not_ be achieved, but leads to the highest _actual_ achievement. Setting goals that are easily achieved leads to less actual achievement (while meeting the goal), while goals that are too hard make people not even try.
The claim (verified by the numbers in the summary!) is that crime wouldn't have risen if not for apple thefts.
The claim is actually that if we took the number of all crimes, and subtracted the number of iPhone thefts, the result would be less than the number of the previous years. And that claim is true, but meaningless.
There were many more iPhones around in 2012 than 2011. The percentage of iPhones among all phones is higher, so it would have to be expected that the number stolen would be higher. If Apple hadnl't sold any phones, almost everyone buying an iPhone would have bought some other phone, and all those iPhone thieves would have stolen some different phone. The crime rate would be exactly the same.
In 2000, the percentage of stolen cars built in the 21st century was zero (some people disagree, but they are clueless). In 2001, it grew slightly, and now it is close to 100%. If it wasn't for these pesky 21st century cars, car theft would be almost down to zero.
So the real owner gets blamed for the gas theft etc. - until it becomes obvious that there's two almost identical cars with the same license plate.
My thoughts on that: 1. There should be a register of suspected duplicated license plates, similar to stolen cars. You should be able to put yourself on that register, possibly adding what's your way to work. A car should be flagged when it's in the register like a stolen car, except a police officer would know that chances are at least 50% that the car is legit. 2. If person X legitimately owns a license plate, then they should be able to demand to be handed over any car using that license plate.
Ever since I first got involved in fighting the RIAA's litigation campaign, and blogging about it, in 2005 [that's almost 8 years ago] I've been arguing that it is not a sufficient basis to bring a lawsuit against someone that an internet access account for which he or she pays the bill was used by someone for a copyright infringement.
Just a few days ago, there was this discussion here on Slashdot about some guy who can't manage to keep a leech of his network:
http://ask.slashdot.org/story/13/02/20/2058235/ask-slashdot-dealing-with-an-advanced-wi-fi-leech
In places like Florida, Stand Your Ground lets them legally shoot you dead for that.
Shoot them for leeching your WiFi? I prefer the punching bit.
The copyright lobby group is now trying to say "no no, it's parody" but they've been lobbying hard to get parody removed from the list of exemptions and have recently succeeded.
Microsoft wouldn't be happy if you used a pirated copy of Microsoft Word to write a parody of Harry Potter. These guys try to create a parody of a website, not a parody of someone's CSS code.
Apple shares are in freefall falling from $705 a share to $450 and deservedly so. It is no longer the largest company in the world by market cap.
Actually, it is. For some strange reason, it was heavily reported when Apple fell behind Exxon, but nobody reported when Apple overtook Exxon again.
If you subtract cash from market caps, the enterprise value of Apple right now is only about 6 1/2 years of profits. That means the company is right now ridiculously undervalued.
No Apple has been growing slower than the industry...in fact its shrinking compared to the market.
Absolutely not. The percentage of iPhones in the phone market is growing. Actually, the phone market has been slightly shrinking in the last year. However, apparently it makes some people on Slashdot happy to declare "smartphones" to be _the_ market. Since $80 feature phones are being replaced with $80 "smartphones", the percentage of smartphones in the total phone market is growing. That doesn't help the sellers, when for every $80 smartphone they are selling they lose the sale of a $80 feature phone.
Apple on the other hand has been growing its phone market share and iPhone unit sales steadily year after year, and there is no reason why that should change in the future.
Can I just say that repeating the nonsense of "patenting shit like rounded corners" marks you as either a troll or an imbecile.
Apple didn't patent rounded corners. Apple has a _design patent_ for a design consisting of many items, one of them rounded corners. To infringe on this _design patent_, you have to copy the complete design, every single item listed in the design patent. You can have as many rounded corners as you like. As long as your design is in some way different from Apple's design patent.
Here's for your enjoyment an example of Samsung patenting rounded corners:
http://www.patentbolt.com/2012/12/samsung-wins-a-design-patent-for-one-of-their-galaxy-phones.html
Yep. Judge people by their actions not their words.
Sometimes that happens. Sometimes it doesn't. As a result, Samsung is supposed to pay $1bn for violating Apple's patents, but a huge leak of hydrofluoric acid at Samsung's plant that killed one worker, injured four and according to police reports may have affected thousands, got them a one thousand dollar fine.
http://www.theinquirer.net/inquirer/news/2244389/police-contradict-samsungs-acid-discharge-claims
Non-commercial use: Any number of Macs that you own and control. Commercial use: Any number of computers used by the same single person, or one computer used by any number of persons.
One computer and can't move to a different computer? That's ridiculous. So if sell your computer and buy a better one, you have to re-buy the software? Or if your computer breaks? Or your computer is stolen? I wonder what your insurance company will say if your computer is stolen, they pay for a replacement, and then you say that instead of restoring your apps from your backup you want them to pay for new copies?
So if you're transferring a file over the network maxing it out and you know it will take 10 minutes but then your wife starts streaming video over the network, how do you predict that at the start of the transfer? It's pretty damn difficult to predict the future man.
I'll give you a different scenario: I start downloading a huge file in the browser, say 800MB. After the first 100MB is done, I start downloading a second huge file in the same browser, say 400MB. How well can you predict when each download will finish?
Another scenario that browsers usually get horribly wrong: I click on a download button, but it takes 10 seconds until the download starts. After 10.01 seconds, the browser calculates average number of bytes per seconds which is ridiculously low and tells me that the download will take 3 weeks. That number goes down rapidly to something reasonable. Why can't that be improved?
Another one that I saw: A download of a 500 MB file was interrupted after 200 MB. When the download is restarted, the browser figures out after 0.01 seconds that 200 MB are already there, so the whole file will take 0.025 seconds. Why can't that be improved?
Same defeatist attitude. There are people who would spend hours and hours arguing why something can't be done, instead of just doing it.
For your comparison between HD and SSD: Yes, on an SSD the time to access a file is much smaller relative to the time for reading bytes, compared to a hard drive. That's why you measure it. You don't need to know the technology, just the numbers. I'll admit that the maths is slightly above kindergarten level, and apparently above slashdot level, but it's not really hard.
If you managed to read the first file of 1MB after 10 milliseconds, and you read the second file of 0.5 MB after 7 milliseconds, how long will it take to read a third file of 2 MB? There are two stupid estimates that have been mentioned here again and again: 1. Based on the number of bytes - you read 1.5MB in 17ms, so 2MB take 22.67ms. 2. Based on the number of files - you read two files in 17ms, so the third file takes 8.5ms. Here is the mathematical bit: Assume that the overhead per file is the same as reading k bytes. We get the best match for the data if we assume 4ms overhead, and 6ms per MB, or k = 666,666 - we actually get a perfect match for the data so far. So we had 1.5MB data and 1.333MB of overhead in 17ms, we read 2MB data plus 0.667MB of overhead, and the calculation gives 16 ms. What are the bets that 16ms is a better prediction than either 8.5ms or 22.67ms? Especially since the 8.5ms estimate is less than the time for the 1MB file, and 22.67ms is more than twice the time for the 1MB file.
Network latency, seek times, sudden Windows Update that decides to start installing packages in the middle of the process.. there's a BILLION different things that could be happening, all the kinds of things that drastically alter the speeds and therefore throwing the estimate right off the table.
So why is it so hard to show a progress bar that is reasonably accurate if none of these things happen? Nobody is complaining that a progress bar is inaccurate if Windows starts installing updates. People complain when its inaccurate no matter what happens.
In this century, updating progress won't slow down the actual progress, unless the code is written by a total moron.
All these constants go out of the window if anything else starts accessing the drive as well, of course, and they're actually variable based on file system and a variety of other things.
What a totally defeatist attitude. Here's how it works currently: There are x MB to copy. So far, y MB have been copied in z seconds. We estimate the number of seconds per MB = z / y, and estimate the time we need as x * (z / y). That's repeated during the copying. Doesn't matter if someone else accesses the drive, doesn't matter how fast or slow the drive is.
Now here comes the ingenious change: There are n files of x MB to copy. So far, m files with y MB have been copied within z seconds. We assume that copying one file has an overhead equivalent to copying k bytes. Initially we assume k = 0 since we know nothing else, but as progress is made and files of different sizes are copied, we can find which k matches the available data best. Then we go through the original algorithm, but replacing x with x + kn, and y with y + km. Easy peasy if your IQ is just slightly above average. Your's apparently isn't.
To put it another way, it is impossible to make an accurate progress bar because it is impossible to predict the future. That's all.
A lot is just stupidity. When you copy lots of files on a hard drive, the time is proportional to some constant times total file size, plus some other constant times number of files. Lots of small files take longer than their size indicates. Most progress bars don't take that into account, so you make huge progress through a dozen 20 MB files, and then slow down through a lot of tiny files. Accounting for that alone would be a huge improvement.
You write files in clusters, "1,301 byte file. Size on disk: 4.00 KB (4,096 bytes)" http://en.wikipedia.org/wiki/Data_cluster [wikipedia.org] (or allocation units)
MacOS X writes: "Size on disk: 4 KB" in that case, not bothering with decimals. For a bigger file, I found "240,831 bytes (242 KB on disk)". The 242 KB is 240,831 bytes, rounded up to a multiple of 4,096 bytes which is the size required on disk (241,664 bytes), then rounded to the nearest integer (242 KB) and not 236 KiB.
I was going to say, the physical world doesn't really care which base unit we used to count with, but computers are very particular about using base 2. When a computer is using integer math to calculate memory addresses in base two and you have a base 10 amount of memory, bad stuff will happen.
And what on earth does that have to do with displaying the correct amount of memory to the user? Is there something that will crash my computer if it displays "4.29 billion bytes of RAM"?