'' Isn't it a possibility that the big labels dictated as part of their agreement that *all* music sold via the iTunes Store has to have DRM? Otherwise, wouldn't it be way easy for other songs from smaller labels that don't demand DRM to sail to the top of the lists because they're DRM-free? ''
The story as I heard it is that Apple has signed identical contracts with the big record companies, and the other to the hundreds of small companies was to accept exactly the same deal. Advantage for Apple: Much less work for the lawyers and less chance of messing up a contract. Advantage for small labels: They won't get ripped off because they can't pay expensive lawyers (unless the big labels messed up and get ripped off as well, which is unlikely). Disadvantage: Everything is sold with DRM.
'' Does this mean KVMs won't work for DVI connections with HDCP?''
Not quite. Encryption stays turned on forever because there is no reason to turn it off. If you have say four computers connected to one monitor, three of them will think that encryption doesn't work and turn off HD output, which doesn't matter because you can't see a thing anyway. The fourth works. Now say you switch output from comp #2 to #3. #2 has been encrypting, #3 wanted to encrypt but didn't. Initially the monitor won't display correctly. Computer #3 should detect within two seconds that there is encryption capable hardware available on the other side, which wasn't there before. Then it should reset the monitor decryption, and turn encryption on again.
'' HDCP is the biggest crime in consumer history yet, let's hope this development kills it before it really takes of. ''
Every time I read a rant about HDCP, I conclude that customers (and content providers as well) have not the slightest clue what HDCP does.
At some point, after all the decryption, decoding, filtering and whatever else is done, your computer must send a signal to the monitor, which the monitor then translates into an image that you can see. This signal usually comes out of the DVI connector in your computer, goes into a cable, which feeds into the monitor or TV. Our paranoid friends at the MPAA or whatever abbreviation it is are afraid that you could catch the signal coming out of the video card, and record it.
Truth is, you can't. You just can't record a signal of 1920 x 1080 pixel times 12 bit per pixel times 60 frames per second on a harddisk. Well, I can't and no normal consumer can. There are people who could build stuff that could do it, but those people are probably happily building graphics cards for NVidia and ATI, or building DVD players.
Still, that signal had to be encrypted. So you have a chip just before the DVI chip (or integrated into it), and another chip in your TV, and they can negotiate to decide on a key for a cipher stream, and use that cipher stream to encrypt the signal on one end and decrypt it on the other end. Which means you can't record the signal coming out of your computer and turn it into a DVD. However, this has nothing to do with DRM whatsoever. Once this encryption is turned on, it stays turned on until the computer or the monitor are turned off. So if you read slashdot after watching a DVD, everything you see on the screen has gone through encryption and decryption. Doesn't matter, because you couldn't read the signal from the cable anyway.
Where the real effort is: First, the graphics driver has to check constantly that encryption works properly. That is not to make sure you don't steal the video signal (as long as encryption is turned on, you can't, and encryption doesn't turn itself off), it is because if the video card and monitor run out of sync then you will see nothing but snow on the monitor, and that makes for a very very unhappy customer. Second, all the commands from the OS to the driver are encrypted, and status reported by the driver is encrypted as well. Otherwise, a hacker could just pretend to be the OS and tell the graphics card to turn encryption off - and that's it! No, most of the work is not the encryption, but to make sure that the OS always knows whether encryption is turned on or off. And third, a DVD can request that high resolution is only used with encryption, so if the HDCP chip isn't there, the image is scaled down to lower resolution.
All in all, the whole HDCP stuff is complete nonsense. It prevents an attack from thieves in a place where you wouldn't attack. It costs money to add and implement. It doesn't hurt you as a consumer, except that you have to pay for the damned chips. It creates work for device driver writers. It doesn't protect contents. Anyone who can record 200 MB per second from a DVI output has invested some serious money, and a little bit more money will allow you to break into a monitor and get the signal from there.
Executive summary: If you can't record a signal coming from the DVI cable, HDCP doesn't matter. If you can record a signal coming from the DVI cable, HDCP doesn't matter much either.
'' A couple of years ago I saw Stallman lecture about the dangers of software patents. A lot of his speech revolved around the busting of the myth of the "patent empowering the little guy", ie. the myth of the lone inventor walking down the street demanding money from the likes of Intel, IBM, Microsoft, because of his mighty patent. Stallman explained that if such a situation would arise, the large companies would simply find ways of countersuing for infringement of some of their numerous patents, thus forcing the smaller entity to give up it's claims, and possibly settle the countercase by giving up it's own patent. ''
But he is not correct. If I find that Microsoft is infringing on my one and only patent in every shipping copy of Windows XP, and Microsoft finds that some software of which I sold 200 copies infringes on 100 of their patents, once damages are compared, I still win.
'' I.e. what's 5 pounds added to 2 stone? I have absolutely no idea. But 500 grams added to 2 kilos is easy. ''
It is two stone five pounds. That's easy. Now converting into kilos is a bit of a problem:
One stone is fourteen pounds. Two stone five pounds = 2 * 14 + 5 pounds = 29 pounds. To convert to kilos, subtract ten percent and divide by two. Thats 26 divided by two = 13 kilos with tiny error.
'' You don't see a problem? The problem is How long does someone have to be ashamed for, and in front of how many people? You put something on the internet and potentially it's there forever and can be seen by millions, like with Star Wars Kid. I believe forgiveness is necessary in society - being allowed to learn from your mistakes and move on to become a better person - but we seem to have a culture where nobody forgives and nobody is allowed to forget. The people doing the uploading, who feel the need to shame and humliate someone this much, must be pretty unpleasant themselves. ''
You make some very very big assumptions there. First, that the punishment doesn't fit the crime. Do you think people get filmed when they steal their first newspaper or their first bottle of milk, or at the point when the neighbor gets sick of not getting the newspaper they paid for for the one hundredth time? Second, that the criminal has learned from their mistakes and become a better person - has he stopped stealing, has he apologized for his behavior, has he paid back the money for the stolen goods? I'd want to see some evidence of that first. Third, that videos cannot be removed and misdeeds are not forgiven when a change really happens - if it happens. Fourth, that taking legal and appropriate measures against constant asocial behavior makes you an unpleasant person yourself - in other words, nice people would be completely helpless against criminals?
'' A "perfect" DRM will have to adjust to changes in copyright status, which means it'll have to be able to do things like pick up the fact that a work has entered the public domain, or the copyright has been extended. ''
I could envision laws that limit any DRM to lets say ten years. If you buy music today, then after ten years, it may or may not be illegal to copy it, but the DRM mechanisms must allow it. 99% of the possible benefit of DRM for the copyright owner is gone by that time (do you think anyone will copy Briteny Spears' music in five years time? ). Since Microsoft DRM already caters for DRM that lets you play or copy music for say one year and not afterwards, it should be easy to modify this for music that cannot be copied for ten years, but can be copied afterwards.
'' Assuming that I could copy, watch, manipulate, change formats, watch on different media players or do whatever I wanted to with (it is mine, after all!) except for distribute illegally. Of course I would support it!
Of course, this is a pipe dream. Even if all players took a thumb-print to make sure I was the true owner, but allowed me to do all the stuff listed above, I would still need to buy multiple copies of whatever so that my wife and kids could enjoy it without me there to swipe my thumb! ''
And even when you're dead and gone, they would make sure they got hold of your thumb.
'' If Steve were forced out of the CEO slot by the Feds, it's likely that he would not be able to hold an officer position in any publicly-traded company. Nice try though. ''
Steve Jobs has created personal computers and created the most popular computer company, two of the best operating systems (or the best operating system twice, depending on how you count), the most popular music player of the decade, and Pixar managed to create five films that beat the top five Disney films in the "Top Hundred". So if he can't run any company in the US, what next?
He moves to Europe, and starts Open Source software seriously. Linux will overtake Microsoft in about ten years time. Tell the SEC to get rid of him at Apple.
'' Are there any studies that show what kind of a decrease in quality is realized by re-encoding, either with the same encoder/settings or with a different encoder? Frankly, I can't hear a damn bit of difference between a burnt & re-riped track and it's original, but my ears are admittedly scarred by that demon rock and roll... Still, I don't believe this quality drop is very significant. I'd be a bit surprised to find a well run study that shows that it is. ''
There were rumors that it would be quite audible, which might be true depending on the encoder. Normally, compression drops a huge percentage of the vast information coming from an uncompressed source. There might be some interference if the input signal has already vastly reduced information. On the other hand, it might be feasible, depending on the encoding standard and the decoder, to reconstruct the compressed signal exactly.
On the other hand, an external 250 GB harddisk is cheap and can hold an awful lot of music.
'' 2) You still cannot buy a retail copy of OSX that will run on your shiny old MacIntel. You only get to buy either an upgrade or a retail pack for PPC. Can you think of any legitimate reason for this other than lockin? ''
God bless your little brain, but every MacIntel that you can buy comes with a copy of MacOS X included, so there is absolutely zero market for an upgrade or retail pack. This will change with MacOS X 10.5.
'' 3) Despite the fact that the MacIntel is a standard enough Intel box, Apple has gone to great lengths to lock OSX to only those Intel boxes that it has blessed with its logo. No technical reason, its pure lockin. ''
Which is fine because there are no non-pirated copies of MacOS X around that you could install legally.
'' 5) Jobs did say, to the NY Times, that you won't be able to run your own software on the iPhone. The laugable reason given was to protect you and the cellular network. But it fits with all the rest. Its just about control and lockin. As is the taboo on unlocking it and moving it to another network. ''
Guess what: They _really really really_ don't want anyone to mess around with the phone network. At the very least, the FCC would want proof that no user-written software can get to the phone network. More likely, the FCC would want proof that whatever software is on the iPhone, calling 911 will still work.
'' No mpeg2, divx or anything else... so it's clear they don't want you using videos from anywhere else. Pure h264 videos are rather hard to get at the moment. ''
'' Actually, I don't think you got me. My point is that if the German government can do broad searches to find who has made transactions on a child porn site, they could just as easily do the same kind of broad requests to find out who has been "purchasing musical illegally". ''
Actually, while purchasing child pornography is a crime, buying music is not a crime.
Other posters have mentioned things like servers in foreign countries, children in foreign countries etc., which were all absolutely irrelevant because the act of purchasing child porn is in itself criminal. For allofmp3.com, the situation is different. Purchasing music on the internet is legal - otherwise iTunes couldn't sell in Germany. Even if it is copyright infringement, that is no crime. So all these things that were irrelevant for child porn are now highly relevant. It is absolutely unclear whether downloading from allofmp3.com is illegal at all. I doubt very much that the RIAA can complain about downloading russian music from allofmp3.com. There is no chance they would get a warrant against _all_ transactions to allofmp3.com, when it isn't even clear at all that they would indicate anything illegal.
'' So, to date, they have 322 suspects out of 22 million scans...that's a hit rate of.00146 percent. That's a lot of people who had their personal data plowed through for a nearly zero yield. ''
I think you don't have a clue how a database works.
There is such a thing as an index. With an index, you can pick out _exactly_ those records that match and don't touch any other records. You _don't_ look at everyones records. You look at exactly 322 records.
If the police are told that a dangerous criminal lives at number 35 in your street, do you think the police wouldn't be allowed to get him because they can't find the house without looking at the house numbers of innocent citizens? You really crack me up.
'' it's not clear what "runs OS X" means. it would suck if programs need special compilation or settings or anything for the iPhone. but maybe the limited CPU/RAM will force optimised programs anyway. since it can multi-task several wifi and phone connections at once I don't see why skype shouldn't work. ''
It won't use an x86 and it is unlikely to use PowerPC. On the other hand, gcc supports compiling for ARM and the compiler produced quite decent code the last time I tried, and if you have code that runs on both PPC and x86 already then it will be no problem to make it run on an ARM chip as well. You'd have to compile and link against the right GUI libraries, and I am sure you don't want any fat binaries on your mobile phone; apart from that it should work fine.
I would assume that any legacy support has been rigorously cut out from the libraries. Possibly Cocoa only, no Carbon at all.
The second version seems more secure, because it doesn't compile. The first version seems to be quite secure as well, because it is likely to crash immediately, and obvious crashes will usually get fixed quickly.
Hint: What is the format string in >> sprintf("Enter something: "); and where will the output go?
'' I know I prefer manual transmissions. Is this really a trait of programmers? Does this mean anything? ''
I prefer automatic. I would prefer it even more if I could change its gear change algorithms to be a bit more sensible at times. (And no, I don't want to hack _my_ automatic transmission, but I would like to design algorithms that do-exactly-what-I-want more often).
'' So. How many people realise you can open consumption data by double-clicking in a location? Fucking NONE OF THEM! They all bitch and moan that they "don't have access to that now", purely on the basis that I haven't held their hand and double-clicked for them. ''
How responsive is your application? I have used apps that were so non-responsive that I would never dare using a double-click. Click. Wait until the system responds. Carefully do the next click. Double-click means too much risk that the first click is processed, changes the display, the second click is interpreted as something completely different, and the user is shafted.
'' Increment/decrement of a register is one of those 1-byte instructions. And don't underestimate those fancy instructions. It's very nice to be able to do things like a nondestructive multiply by 5 without using up the ALU. ''
Increment/decrement of a register is gone in 64 bit code and painfully slow anyway (it seemed a good idea in 1977, but partial condition code register updates are a pain). Lots of floating point instructions have three or four bytes in prefixes alone, that's before we even start with the instruction itself. Multiplication by 5 using LEA instruction on a P4 took five cycles; on an ARM it takes one...
I had to write a bit of decryption code a few months ago, and a 200 MHz ARM chip was about as fast as an 800 MHz P4.
'' The question was: "What effect exactly have these two had?" My answer was: "Apple shipped fixes." Good enough for me. ''
In that case, can we remove these two from the article, and replace their names with those of the unknown Apple engineers who went through the code and found whatever he found, to those who fixed the problem (probably the same ones), the unknown testers at Apple who made sure that the fix didn't break anything, and their manager who changed priorities to the wireless driver?
What effect exactly have these two had? Made serious security researchers ridiculous by showing a rigged demo of a supposed exploit that until today hasn't been reproduced by anyone?
Slashdot Mirror
'' Isn't it a possibility that the big labels dictated as part of their agreement that *all* music sold via the iTunes Store has to have DRM? Otherwise, wouldn't it be way easy for other songs from smaller labels that don't demand DRM to sail to the top of the lists because they're DRM-free? ''
The story as I heard it is that Apple has signed identical contracts with the big record companies, and the other to the hundreds of small companies was to accept exactly the same deal. Advantage for Apple: Much less work for the lawyers and less chance of messing up a contract. Advantage for small labels: They won't get ripped off because they can't pay expensive lawyers (unless the big labels messed up and get ripped off as well, which is unlikely). Disadvantage: Everything is sold with DRM.
'' Does this mean KVMs won't work for DVI connections with HDCP?''
Not quite. Encryption stays turned on forever because there is no reason to turn it off. If you have say four computers connected to one monitor, three of them will think that encryption doesn't work and turn off HD output, which doesn't matter because you can't see a thing anyway. The fourth works. Now say you switch output from comp #2 to #3. #2 has been encrypting, #3 wanted to encrypt but didn't. Initially the monitor won't display correctly. Computer #3 should detect within two seconds that there is encryption capable hardware available on the other side, which wasn't there before. Then it should reset the monitor decryption, and turn encryption on again.
Just hope that the drivers get it all right.
'' Every cent saved adds up to millions over a manufacturing run of 100,000 units. ''
Not in my book.
'' HDCP is the biggest crime in consumer history yet, let's hope this development kills it before it really takes of. ''
Every time I read a rant about HDCP, I conclude that customers (and content providers as well) have not the slightest clue what HDCP does.
At some point, after all the decryption, decoding, filtering and whatever else is done, your computer must send a signal to the monitor, which the monitor then translates into an image that you can see. This signal usually comes out of the DVI connector in your computer, goes into a cable, which feeds into the monitor or TV. Our paranoid friends at the MPAA or whatever abbreviation it is are afraid that you could catch the signal coming out of the video card, and record it.
Truth is, you can't. You just can't record a signal of 1920 x 1080 pixel times 12 bit per pixel times 60 frames per second on a harddisk. Well, I can't and no normal consumer can. There are people who could build stuff that could do it, but those people are probably happily building graphics cards for NVidia and ATI, or building DVD players.
Still, that signal had to be encrypted. So you have a chip just before the DVI chip (or integrated into it), and another chip in your TV, and they can negotiate to decide on a key for a cipher stream, and use that cipher stream to encrypt the signal on one end and decrypt it on the other end. Which means you can't record the signal coming out of your computer and turn it into a DVD. However, this has nothing to do with DRM whatsoever. Once this encryption is turned on, it stays turned on until the computer or the monitor are turned off. So if you read slashdot after watching a DVD, everything you see on the screen has gone through encryption and decryption. Doesn't matter, because you couldn't read the signal from the cable anyway.
Where the real effort is: First, the graphics driver has to check constantly that encryption works properly. That is not to make sure you don't steal the video signal (as long as encryption is turned on, you can't, and encryption doesn't turn itself off), it is because if the video card and monitor run out of sync then you will see nothing but snow on the monitor, and that makes for a very very unhappy customer. Second, all the commands from the OS to the driver are encrypted, and status reported by the driver is encrypted as well. Otherwise, a hacker could just pretend to be the OS and tell the graphics card to turn encryption off - and that's it! No, most of the work is not the encryption, but to make sure that the OS always knows whether encryption is turned on or off. And third, a DVD can request that high resolution is only used with encryption, so if the HDCP chip isn't there, the image is scaled down to lower resolution.
All in all, the whole HDCP stuff is complete nonsense. It prevents an attack from thieves in a place where you wouldn't attack. It costs money to add and implement. It doesn't hurt you as a consumer, except that you have to pay for the damned chips. It creates work for device driver writers. It doesn't protect contents. Anyone who can record 200 MB per second from a DVI output has invested some serious money, and a little bit more money will allow you to break into a monitor and get the signal from there.
Executive summary: If you can't record a signal coming from the DVI cable, HDCP doesn't matter. If you can record a signal coming from the DVI cable, HDCP doesn't matter much either.
'' A couple of years ago I saw Stallman lecture about the dangers of software patents. A lot of his speech revolved around the busting of the myth of the "patent empowering the little guy", ie. the myth of the lone inventor walking down the street demanding money from the likes of Intel, IBM, Microsoft, because of his mighty patent. Stallman explained that if such a situation would arise, the large companies would simply find ways of countersuing for infringement of some of their numerous patents, thus forcing the smaller entity to give up it's claims, and possibly settle the countercase by giving up it's own patent. ''
But he is not correct. If I find that Microsoft is infringing on my one and only patent in every shipping copy of Windows XP, and Microsoft finds that some software of which I sold 200 copies infringes on 100 of their patents, once damages are compared, I still win.
'' I.e. what's 5 pounds added to 2 stone? I have absolutely no idea. But 500 grams added to 2 kilos is easy. ''
It is two stone five pounds. That's easy. Now converting into kilos is a bit of a problem:
One stone is fourteen pounds. Two stone five pounds = 2 * 14 + 5 pounds = 29 pounds.
To convert to kilos, subtract ten percent and divide by two. Thats 26 divided by two = 13 kilos with tiny error.
'' But "A pint's a pound the world around". ''
Where I live, a pint of water weighs four thirds of a pound.
'' You don't see a problem? The problem is How long does someone have to be ashamed for, and in front of how many people? You put something on the internet and potentially it's there forever and can be seen by millions, like with Star Wars Kid. I believe forgiveness is necessary in society - being allowed to learn from your mistakes and move on to become a better person - but we seem to have a culture where nobody forgives and nobody is allowed to forget. The people doing the uploading, who feel the need to shame and humliate someone this much, must be pretty unpleasant themselves. ''
You make some very very big assumptions there. First, that the punishment doesn't fit the crime. Do you think people get filmed when they steal their first newspaper or their first bottle of milk, or at the point when the neighbor gets sick of not getting the newspaper they paid for for the one hundredth time? Second, that the criminal has learned from their mistakes and become a better person - has he stopped stealing, has he apologized for his behavior, has he paid back the money for the stolen goods? I'd want to see some evidence of that first. Third, that videos cannot be removed and misdeeds are not forgiven when a change really happens - if it happens. Fourth, that taking legal and appropriate measures against constant asocial behavior makes you an unpleasant person yourself - in other words, nice people would be completely helpless against criminals?
'' A "perfect" DRM will have to adjust to changes in copyright status, which means it'll have to be able to do things like pick up the fact that a work has entered the public domain, or the copyright has been extended. ''
I could envision laws that limit any DRM to lets say ten years. If you buy music today, then after ten years, it may or may not be illegal to copy it, but the DRM mechanisms must allow it. 99% of the possible benefit of DRM for the copyright owner is gone by that time (do you think anyone will copy Briteny Spears' music in five years time? ). Since Microsoft DRM already caters for DRM that lets you play or copy music for say one year and not afterwards, it should be easy to modify this for music that cannot be copied for ten years, but can be copied afterwards.
'' Assuming that I could copy, watch, manipulate, change formats, watch on different media players or do whatever I wanted to with (it is mine, after all!) except for distribute illegally. Of course I would support it!
Of course, this is a pipe dream. Even if all players took a thumb-print to make sure I was the true owner, but allowed me to do all the stuff listed above, I would still need to buy multiple copies of whatever so that my wife and kids could enjoy it without me there to swipe my thumb! ''
And even when you're dead and gone, they would make sure they got hold of your thumb.
'' If Steve were forced out of the CEO slot by the Feds, it's likely that he would not be able to hold an officer position in any publicly-traded company. Nice try though. ''
Steve Jobs has created personal computers and created the most popular computer company, two of the best operating systems (or the best operating system twice, depending on how you count), the most popular music player of the decade, and Pixar managed to create five films that beat the top five Disney films in the "Top Hundred". So if he can't run any company in the US, what next?
He moves to Europe, and starts Open Source software seriously. Linux will overtake Microsoft in about ten years time. Tell the SEC to get rid of him at Apple.
'' Are there any studies that show what kind of a decrease in quality is realized by re-encoding, either with the same encoder/settings or with a different encoder?
Frankly, I can't hear a damn bit of difference between a burnt & re-riped track and it's original, but my ears are admittedly scarred by that demon rock and roll... Still, I don't believe this quality drop is very significant. I'd be a bit surprised to find a well run study that shows that it is. ''
There were rumors that it would be quite audible, which might be true depending on the encoder. Normally, compression drops a huge percentage of the vast information coming from an uncompressed source. There might be some interference if the input signal has already vastly reduced information. On the other hand, it might be feasible, depending on the encoding standard and the decoder, to reconstruct the compressed signal exactly.
On the other hand, an external 250 GB harddisk is cheap and can hold an awful lot of music.
'' 2) You still cannot buy a retail copy of OSX that will run on your shiny old MacIntel. You only get to buy either an upgrade or a retail pack for PPC. Can you think of any legitimate reason for this other than lockin? ''
God bless your little brain, but every MacIntel that you can buy comes with a copy of MacOS X included, so there is absolutely zero market for an upgrade or retail pack. This will change with MacOS X 10.5.
'' 3) Despite the fact that the MacIntel is a standard enough Intel box, Apple has gone to great lengths to lock OSX to only those Intel boxes that it has blessed with its logo. No technical reason, its pure lockin. ''
Which is fine because there are no non-pirated copies of MacOS X around that you could install legally.
'' 5) Jobs did say, to the NY Times, that you won't be able to run your own software on the iPhone. The laugable reason given was to protect you and the cellular network. But it fits with all the rest. Its just about control and lockin. As is the taboo on unlocking it and moving it to another network. ''
Guess what: They _really really really_ don't want anyone to mess around with the phone network. At the very least, the FCC would want proof that no user-written software can get to the phone network. More likely, the FCC would want proof that whatever software is on the iPhone, calling 911 will still work.
'' No mpeg2, divx or anything else... so it's clear they don't want you using videos from anywhere else. Pure h264 videos are rather hard to get at the moment. ''
DVD + Mac The Ripper + Handbrake.
'' Actually, I don't think you got me. My point is that if the German government can do broad searches to find who has made transactions on a child porn site, they could just as easily do the same kind of broad requests to find out who has been "purchasing musical illegally". ''
Actually, while purchasing child pornography is a crime, buying music is not a crime.
Other posters have mentioned things like servers in foreign countries, children in foreign countries etc., which were all absolutely irrelevant because the act of purchasing child porn is in itself criminal. For allofmp3.com, the situation is different. Purchasing music on the internet is legal - otherwise iTunes couldn't sell in Germany. Even if it is copyright infringement, that is no crime. So all these things that were irrelevant for child porn are now highly relevant. It is absolutely unclear whether downloading from allofmp3.com is illegal at all. I doubt very much that the RIAA can complain about downloading russian music from allofmp3.com. There is no chance they would get a warrant against _all_ transactions to allofmp3.com, when it isn't even clear at all that they would indicate anything illegal.
'' So, to date, they have 322 suspects out of 22 million scans...that's a hit rate of .00146 percent. That's a lot of people who had their personal data plowed through for a nearly zero yield. ''
I think you don't have a clue how a database works.
There is such a thing as an index. With an index, you can pick out _exactly_ those records that match and don't touch any other records. You _don't_ look at everyones records. You look at exactly 322 records.
If the police are told that a dangerous criminal lives at number 35 in your street, do you think the police wouldn't be allowed to get him because they can't find the house without looking at the house numbers of innocent citizens? You really crack me up.
'' The Apple iPhone will probably be the first phone with a dual core processor. ''
With two processors it would be much behind the average phone. Even an iPod has two ARM processors.
'' it's not clear what "runs OS X" means. it would suck if programs need special compilation or settings or anything for the iPhone. but maybe the limited CPU/RAM will force optimised programs anyway. since it can multi-task several wifi and phone connections at once I don't see why skype shouldn't work. ''
It won't use an x86 and it is unlikely to use PowerPC. On the other hand, gcc supports compiling for ARM and the compiler produced quite decent code the last time I tried, and if you have code that runs on both PPC and x86 already then it will be no problem to make it run on an ARM chip as well. You'd have to compile and link against the right GUI libraries, and I am sure you don't want any fat binaries on your mobile phone; apart from that it should work fine.
I would assume that any legacy support has been rigorously cut out from the libraries. Possibly Cocoa only, no Carbon at all.
'' Anyone know if it's a removable battery? ''
Of course it is removable. But I think it is not user-replacable.
The second version seems more secure, because it doesn't compile.
The first version seems to be quite secure as well, because it is likely to crash immediately, and obvious crashes will usually get fixed quickly.
Hint: What is the format string in >> sprintf("Enter something: "); and where will the output go?
'' I know I prefer manual transmissions. Is this really a trait of programmers? Does this mean anything? ''
I prefer automatic. I would prefer it even more if I could change its gear change algorithms to be a bit more sensible at times. (And no, I don't want to hack _my_ automatic transmission, but I would like to design algorithms that do-exactly-what-I-want more often).
'' So. How many people realise you can open consumption data by double-clicking in a location? Fucking NONE OF THEM! They all bitch and moan that they "don't have access to that now", purely on the basis that I haven't held their hand and double-clicked for them. ''
How responsive is your application? I have used apps that were so non-responsive that I would never dare using a double-click. Click. Wait until the system responds. Carefully do the next click. Double-click means too much risk that the first click is processed, changes the display, the second click is interpreted as something completely different, and the user is shafted.
'' Increment/decrement of a register is one of those 1-byte instructions. And don't underestimate those fancy instructions. It's very nice to be able to do things like a nondestructive multiply by 5 without using up the ALU. ''
Increment/decrement of a register is gone in 64 bit code and painfully slow anyway (it seemed a good idea in 1977, but partial condition code register updates are a pain). Lots of floating point instructions have three or four bytes in prefixes alone, that's before we even start with the instruction itself. Multiplication by 5 using LEA instruction on a P4 took five cycles; on an ARM it takes one...
I had to write a bit of decryption code a few months ago, and a 200 MHz ARM chip was about as fast as an 800 MHz P4.
'' The question was: "What effect exactly have these two had?" My answer was: "Apple shipped fixes." Good enough for me. ''
In that case, can we remove these two from the article, and replace their names with those of the unknown Apple engineers who went through the code and found whatever he found, to those who fixed the problem (probably the same ones), the unknown testers at Apple who made sure that the fix didn't break anything, and their manager who changed priorities to the wireless driver?
What effect exactly have these two had? Made serious security researchers ridiculous by showing a rigged demo of a supposed exploit that until today hasn't been reproduced by anyone?