You've never seen an infinite loop inside malloc? I guess I have been unlucky enough to debug more memory corruptions than you then. They're not as common as crashes but they do happen on every platform I've worked with.
Nothing I said is false. Objective C requires you to manually manage memory. Yes, the Cocoa libraries provide some (weak) abstractions to help you with that, not that auto-release pools are very helpful for managing non-event scoped state. Yet it is based on C and does not bounds check, it does not prevent memory errors and it uses a standard malloc implementation. Having some abstraction over malloc doesn't suddenly make your program managed.
At that point, a screen can be scraped and audio intercepted. It only ramps up the time to get a good movie rip, but if you want to show the content to users, there is a way to steal said content.
Sure, but you aren't going to get a quality rip that way. Pointing a camera at the screen is going to produce incredibly shitty results even compared to DVD. Do you really want to invite your friends round for a movie night and then show them a camcorded super compressed rip? It might be OK for personal use if you don't care about quality at all, but then a camcorded copy from a cinema will work just as well.
The updated BD+ is taking some time, but once they figure out how to emulate the virtual machine better, it will again fall. The downfall of Blu Ray is built in - accurately emulate an official virtual machine, and the disc will decrypt itself for playback.
In theory, you're right. In practice it's not that easy. Before you even get to the "accurately emulate" stage you need some player keys. Early BluRay players had woeful security, the keys were right there, unprotected and in one piece in memory. Player security has improved a lot and every three months, the AACS keys for software players rotate. So you need to be almost constantly reverse engineering existing players to get a steady supply of keys, with it getting harder every time.
After that, there are plenty of tricks BD+ can use to prove it's a real player. Movies can infinitely hash player memory and state just for one, so to even get past the first post you have to obtain, load and run the real player in parallel with your own decryptor. There are lots of techniques that can make this "non trivial". And remember time is not infinite... you have, at best, 3 months to do all this until you need to find a new set of hidden keys.
Bear in mind that if it's just a firmware upgrade, the BluRay disc can ship the upgrade on it and then require the player to update itself before it'll play. So no big deal.
Yeah but what's the incentive? You realize that implementing and keeping up with BD+ requires work measured in man years, right? At some point releasing movies for the scene recognition becomes insufficient. People who are skilled enough to do that work are also skilled enough to get well paying jobs that impress chicks, so why would they spend a crapload of money on buying disks then uploading them for nothing?
It's a required part of the BD+ spec, the VM is simple, I'd assume there's a crapload of unit tests for it and the complicated/hard BD+ codepaths are only taken for compromised players. So I think the answer will be "yes".
They don't care if the protection is broken eventually. If the movie is protected for a month or two after release then nearly all the sales will have been during the "no piracy" zone. Apparently they have figures that show pirates are just as impatient as regular Joes, perhaps even more so, and will eventually give up waiting for a pirate copy and go buy it if they really want that thing.
I don't have a BluRay player nor any involvement with BluRay at all. I just have an interest in DRM technologies. BTW the summary is correct - current BD+ programs have proved unbreakable to the open source community. That may well change in future of course. But for now the effort required appears to be prohibitive, as the previous open source implementors have said outright that it's too much work for them.
BD+ can hash arbitrary sections of player memory. This is a key problem for anybody wanting to build an "emulation" - you have no choice but to ship a complete firmware image with your alternative implementation. At that point you're committing good old fashioned copyright violation, not a DMCA violation. No problem for the pirates, but it is a problem for the company that actually develops and maintains the solution. It's also a problem for open source distributors.
Yeah, exactly. I'd be surprised if BD+ really reduced piracy. I suspect most pirates will just grab the lower quality but still highly watchable DVD rips. I guess if BluRay penetration increases studios might start releasing the DVD copies months after the BluRay copies, but there'll always be a large contigent of people who just don't care about the quality increase. I think piracy is mostly about convenience after all.
The BD+ VM is pretty simple actually and can be implemented in software. I imagine the bulk of the cost is in licensing, not the actual technical cost of implementation. And by the way, BD+ only really digs into your player if it's known to be compromised. If a new version of the player firmware is released that resecures it, BD+ programs won't bother to do any checks on it.
I assume it wasn't the processor, because it wasn't locked up -- so perhaps it was a modem issue.
Although Apple don't want you to know it, the iPhone OS is actually a multi-tasking OS at heart, so a spinning process won't lock up the processor. It'll just keep the CPU pegged at 100% and flatten the battery. What might cause such a thing? On Android, most of the software is written in Java so the only way to make it spin like that is to actually enter a real infinite loop. The iPhone is (rather questionably imho) written in Objective-C, which uses manual memory management and thus lets you do double frees, buffer overruns etc. A classic cause of infinite loops in C based apps is heap arena corruption - you smash the heap control structures in some way, and then malloc or free go into an infinite loop trying to find a free block. Most mallocs don't do much error checking for performance reasons.
What I suspect happens is that there's a subtle memory error in a part of iPhone 3.0 which causes some background process to start spinning inside malloc. The iPhone 3GS has a more powerful processor than its predecessor and it's possible that Apple decided (riskily imho) that it was OK if they couldn't dissipate the heat from a pegged CPU because they would try and ensure the CPU never pegged in software. For instance they might be throttling user applications (I have no evidence of that, it's just a theory). However that overlooks the possibility that their own software would accidentally peg the CPU for some time.... leading to the result we have here.
It does go away if you uninstall Google software. It'll take a few hours for it to wake up and notice that all the software it's meant to update has gone, but it will leave eventually.
Naver isn't exactly a search engine per se. It's really a questions and answer service. It solved the problem the Korean market had at the time, which was that there wasn't much on the web written in Korean. Unfortunately they robots.txt out all their content so they are the only ones that can search it, thus Naver is the de-facto search engine in Korea now because they "own" peoples questions and answers.
Re:Android just won't catch up with iPhone
on
Unlocking Android
·
· Score: 1
I've written an Android app but not an iPhone app, so, my opinion is less informed than yours. That said, you can write native code now. It's not really meant to write entire apps in because the APIs are still Java and that is unlikely to change, but if you want to do something compute intensive in C/C++ that's certainly now an option. Personally I think Objective-C is a pretty stupid language compared to Java, so, having at least the option to mix and match approaches is nice.
There is a visual UI designer, so you don't have to write XML. That said, the UI designer sucks donkey balls, so it's not a superior alternative. It's balanced by the Android layout managers which are superb. RelativeLayout in particular is very intuitive. You just say "I want this control to be below that one, and that control to be to the right side of this one" etc. It's probably the best GUI toolkit I've used so far.
You don't have to design in DIPs. You can use regular pixel measurements if you want. Of course then you get to do extra work if you want to support other screen resolutions, but that's the inevitable result of Androids "many hardware designs" philosophy. Is the iPhone hardware the last word in handset design? If you believe it is, then DIPs are a waste of time. If (like me) it seems unlikely, then DIPs are probably a good idea.
As to Core Animation, I'll have to take your word for it. There is animation support in the Android UI toolkit and I've used it to make nice swipes/fades etc. I don't know how much more sophisticated Core Animation is, but the Android API demos show how to do things like 3D screen flips etc.
HTC is a Taiwanese electronics firm. The CEO in question is (according to the previous summary) running HCL Technologies. Of course if the previous summary is inaccurate as well, this is also wrong. But that's slashdot - they should replace the/. logo with a box of chocolates...
Yeah, I think it's pretty unlikely they'll do online activation just yet for the 360 at least. There are too many people who aren't connected. What they might start doing is releasing the game online a month or two before the disc based version is out. There isn't any way to pirate online distributed games on modern consoles and it seems unlikely that this will change, so they can exploit the desire pirates have to play the newest stuff to get a few extra sales without reducing their eventual overall market.
Actually it's not possible to copy games on the PS3, period. It's also not possible to copy internet distributed games on PS3 or Xbox 360. It is possible to copy game discs on the 360, but it requires some convoluted modding process and if Microsoft detect you the console will be banned from Xbox Live. At least for me, that's a major part of the Xbox 360 value, so, getting the box banned would be a big risk.
Eh? The GFW has been known to run out of capacity at peak times and stop reliably blocking things. They don't even have the resources to do stream re-assembly, splitting the stopwords across packet boundaries is enough to defeat it. I don't think they'll be decrypting traffic in realtime anytime soon.
I think a spokesman from Nokia claimed that installation of such systems is legally required to build a cellphone network in the western world, so it's not like they'd have had a strong moral standing to deny the sale.
I wonder if it's actually impossible for him to have used 9G of data. If he had a wireless USB modem, is it possible he had it in all the time and assumed it wouldn't be in use if he wasn't browsing the web? If his machine has a bot on it, I bet it could chew through 9G of data no trouble and he'd never even know.
How many "anti copyright professional associations" are there? Even if you consider the pirate party to be such an organization, that's pretty new. I don't know of any others. Probably because "professionals" tend to want to get paid for their work, so if they produce creative works they tend to support copyright.
Or.... Sweden might not be the banana republic some posters seem to think it is. Perhaps the judge was not in fact biased. I mean, saying the judge is biased because he belongs to some groups related to copyright law is like saying I'm obviously pro file-sharing because I'm a member of slashdot. But I'm not.
Technicalities like that are exactly the sort of thing the court decided it did not care about. Sort of like how an arms fair doesn't actually "sell arms", that's what the individual dealers do, and hey how were the organizers supposed to know that if they created the "Ukranian Arms Fair" and reserved lots of space that a bunch of weapon dealers would move in? Totally unreasonable to expect them to deal with that!
Slashdot Mirror
You've never seen an infinite loop inside malloc? I guess I have been unlucky enough to debug more memory corruptions than you then. They're not as common as crashes but they do happen on every platform I've worked with.
Nothing I said is false. Objective C requires you to manually manage memory. Yes, the Cocoa libraries provide some (weak) abstractions to help you with that, not that auto-release pools are very helpful for managing non-event scoped state. Yet it is based on C and does not bounds check, it does not prevent memory errors and it uses a standard malloc implementation. Having some abstraction over malloc doesn't suddenly make your program managed.
Sure, but you aren't going to get a quality rip that way. Pointing a camera at the screen is going to produce incredibly shitty results even compared to DVD. Do you really want to invite your friends round for a movie night and then show them a camcorded super compressed rip? It might be OK for personal use if you don't care about quality at all, but then a camcorded copy from a cinema will work just as well.
In theory, you're right. In practice it's not that easy. Before you even get to the "accurately emulate" stage you need some player keys. Early BluRay players had woeful security, the keys were right there, unprotected and in one piece in memory. Player security has improved a lot and every three months, the AACS keys for software players rotate. So you need to be almost constantly reverse engineering existing players to get a steady supply of keys, with it getting harder every time.
After that, there are plenty of tricks BD+ can use to prove it's a real player. Movies can infinitely hash player memory and state just for one, so to even get past the first post you have to obtain, load and run the real player in parallel with your own decryptor. There are lots of techniques that can make this "non trivial". And remember time is not infinite ... you have, at best, 3 months to do all this until you need to find a new set of hidden keys.
Bear in mind that if it's just a firmware upgrade, the BluRay disc can ship the upgrade on it and then require the player to update itself before it'll play. So no big deal.
Yeah but what's the incentive? You realize that implementing and keeping up with BD+ requires work measured in man years, right? At some point releasing movies for the scene recognition becomes insufficient. People who are skilled enough to do that work are also skilled enough to get well paying jobs that impress chicks, so why would they spend a crapload of money on buying disks then uploading them for nothing?
It's a required part of the BD+ spec, the VM is simple, I'd assume there's a crapload of unit tests for it and the complicated/hard BD+ codepaths are only taken for compromised players. So I think the answer will be "yes".
They don't care if the protection is broken eventually. If the movie is protected for a month or two after release then nearly all the sales will have been during the "no piracy" zone. Apparently they have figures that show pirates are just as impatient as regular Joes, perhaps even more so, and will eventually give up waiting for a pirate copy and go buy it if they really want that thing.
I don't have a BluRay player nor any involvement with BluRay at all. I just have an interest in DRM technologies. BTW the summary is correct - current BD+ programs have proved unbreakable to the open source community. That may well change in future of course. But for now the effort required appears to be prohibitive, as the previous open source implementors have said outright that it's too much work for them.
BD+ can hash arbitrary sections of player memory. This is a key problem for anybody wanting to build an "emulation" - you have no choice but to ship a complete firmware image with your alternative implementation. At that point you're committing good old fashioned copyright violation, not a DMCA violation. No problem for the pirates, but it is a problem for the company that actually develops and maintains the solution. It's also a problem for open source distributors.
Yeah, exactly. I'd be surprised if BD+ really reduced piracy. I suspect most pirates will just grab the lower quality but still highly watchable DVD rips. I guess if BluRay penetration increases studios might start releasing the DVD copies months after the BluRay copies, but there'll always be a large contigent of people who just don't care about the quality increase. I think piracy is mostly about convenience after all.
The BD+ VM is pretty simple actually and can be implemented in software. I imagine the bulk of the cost is in licensing, not the actual technical cost of implementation. And by the way, BD+ only really digs into your player if it's known to be compromised. If a new version of the player firmware is released that resecures it, BD+ programs won't bother to do any checks on it.
Although Apple don't want you to know it, the iPhone OS is actually a multi-tasking OS at heart, so a spinning process won't lock up the processor. It'll just keep the CPU pegged at 100% and flatten the battery. What might cause such a thing? On Android, most of the software is written in Java so the only way to make it spin like that is to actually enter a real infinite loop. The iPhone is (rather questionably imho) written in Objective-C, which uses manual memory management and thus lets you do double frees, buffer overruns etc. A classic cause of infinite loops in C based apps is heap arena corruption - you smash the heap control structures in some way, and then malloc or free go into an infinite loop trying to find a free block. Most mallocs don't do much error checking for performance reasons.
What I suspect happens is that there's a subtle memory error in a part of iPhone 3.0 which causes some background process to start spinning inside malloc. The iPhone 3GS has a more powerful processor than its predecessor and it's possible that Apple decided (riskily imho) that it was OK if they couldn't dissipate the heat from a pegged CPU because they would try and ensure the CPU never pegged in software. For instance they might be throttling user applications (I have no evidence of that, it's just a theory). However that overlooks the possibility that their own software would accidentally peg the CPU for some time .... leading to the result we have here.
It does go away if you uninstall Google software. It'll take a few hours for it to wake up and notice that all the software it's meant to update has gone, but it will leave eventually.
Naver isn't exactly a search engine per se. It's really a questions and answer service. It solved the problem the Korean market had at the time, which was that there wasn't much on the web written in Korean. Unfortunately they robots.txt out all their content so they are the only ones that can search it, thus Naver is the de-facto search engine in Korea now because they "own" peoples questions and answers.
I've written an Android app but not an iPhone app, so, my opinion is less informed than yours. That said, you can write native code now. It's not really meant to write entire apps in because the APIs are still Java and that is unlikely to change, but if you want to do something compute intensive in C/C++ that's certainly now an option. Personally I think Objective-C is a pretty stupid language compared to Java, so, having at least the option to mix and match approaches is nice.
There is a visual UI designer, so you don't have to write XML. That said, the UI designer sucks donkey balls, so it's not a superior alternative. It's balanced by the Android layout managers which are superb. RelativeLayout in particular is very intuitive. You just say "I want this control to be below that one, and that control to be to the right side of this one" etc. It's probably the best GUI toolkit I've used so far.
You don't have to design in DIPs. You can use regular pixel measurements if you want. Of course then you get to do extra work if you want to support other screen resolutions, but that's the inevitable result of Androids "many hardware designs" philosophy. Is the iPhone hardware the last word in handset design? If you believe it is, then DIPs are a waste of time. If (like me) it seems unlikely, then DIPs are probably a good idea.
As to Core Animation, I'll have to take your word for it. There is animation support in the Android UI toolkit and I've used it to make nice swipes/fades etc. I don't know how much more sophisticated Core Animation is, but the Android API demos show how to do things like 3D screen flips etc.
HTC is a Taiwanese electronics firm. The CEO in question is (according to the previous summary) running HCL Technologies. Of course if the previous summary is inaccurate as well, this is also wrong. But that's slashdot - they should replace the /. logo with a box of chocolates ...
Yeah, I think it's pretty unlikely they'll do online activation just yet for the 360 at least. There are too many people who aren't connected. What they might start doing is releasing the game online a month or two before the disc based version is out. There isn't any way to pirate online distributed games on modern consoles and it seems unlikely that this will change, so they can exploit the desire pirates have to play the newest stuff to get a few extra sales without reducing their eventual overall market.
Actually it's not possible to copy games on the PS3, period. It's also not possible to copy internet distributed games on PS3 or Xbox 360. It is possible to copy game discs on the 360, but it requires some convoluted modding process and if Microsoft detect you the console will be banned from Xbox Live. At least for me, that's a major part of the Xbox 360 value, so, getting the box banned would be a big risk.
Eh? The GFW has been known to run out of capacity at peak times and stop reliably blocking things. They don't even have the resources to do stream re-assembly, splitting the stopwords across packet boundaries is enough to defeat it. I don't think they'll be decrypting traffic in realtime anytime soon.
I think a spokesman from Nokia claimed that installation of such systems is legally required to build a cellphone network in the western world, so it's not like they'd have had a strong moral standing to deny the sale.
I wonder if it's actually impossible for him to have used 9G of data. If he had a wireless USB modem, is it possible he had it in all the time and assumed it wouldn't be in use if he wasn't browsing the web? If his machine has a bot on it, I bet it could chew through 9G of data no trouble and he'd never even know.
How many "anti copyright professional associations" are there? Even if you consider the pirate party to be such an organization, that's pretty new. I don't know of any others. Probably because "professionals" tend to want to get paid for their work, so if they produce creative works they tend to support copyright.
Or .... Sweden might not be the banana republic some posters seem to think it is. Perhaps the judge was not in fact biased. I mean, saying the judge is biased because he belongs to some groups related to copyright law is like saying I'm obviously pro file-sharing because I'm a member of slashdot. But I'm not.
Technicalities like that are exactly the sort of thing the court decided it did not care about. Sort of like how an arms fair doesn't actually "sell arms", that's what the individual dealers do, and hey how were the organizers supposed to know that if they created the "Ukranian Arms Fair" and reserved lots of space that a bunch of weapon dealers would move in? Totally unreasonable to expect them to deal with that!
Wow, bravo. That is an excellent post. I had no idea GTA4 made over a billion dollars. That's amazing!
Huh? There are quite a few platformers on Xbox Live.