Baidu has reasonable enough results for the Chinese web, but doesn't really search the English web at all. Google.cn does both very well.
I am guessing their problem is that a lot of the papers they wish to read and general scientific world is based on English. As there aren't any other major search engines in China, having only Baidu would be close to having no search engine at all for people who need to search English documents.
Android really hasn't fragmented. Windows apps tend to work everywhere because developers are very slow to adopt new APIs. The same is true of Android: the new APIs in 2.0/2.1 are neat, but most of them are either convenience or purely candy (like live wallpapers or social networking integration). Many apps don't need to use them and they'll work just fine on older OS versions. And then of course, you have the fact that Android updates itself automatically over the air. The 2.1 release is only a month old, it's hardly unexpected that so far other devices don't have it.
As somebody who has written an app for both J2ME and Android, that comparison is really far out.
There are some key differences between J2ME and Android:
Android is an implementation, J2ME is a spec. Developing for J2ME is a nightmare because practically every phone has a unique combination of serious bugs. This isn't true of Android because every phone is working off the same core code. If manufacturers want to modify that code, that's OK, there's a test suite they have to pass if they want certain things.
Androids "fragmentation" is actually simply that not everyone has upgraded yet - no different to any other platform (including the iPhone). But Android manufacturers have a history of upgrading versions, and I don't see much reason for them to change this: customers demand it and because Android is open source, doing the upgrade is effectively "free" by mobile development standards, even for things like Sense UI (the APIs are backwards compatible).
Android explicitly supports writing apps that run on multiple platform versions, where as J2ME never really did. For instance you can hide your app from earlier devices (in the market), or you can make your app gracefully degrade on older devices.
Android does actually standardise hardware, it just doesn't standardise on exactly one set. For instance, you can assume all Android devices have a decent touch screen, a display that uses one of a handful of resolutions and so on.
Google only lets manufacturers distribute certain closed source apps (like Market) if they meet certain conditions. Google has every incentive to keep as many people on the latest version of Android as possible. They have the leverage needed.
In short, it's a slow news day and InfoWorld know how to stir up a story.
Come on Bruce. Your argument for why Android low quality is that the kernel has patches and the "utility code is crap"?
Has it escaped your notice that unlike on the (presumably?) high quality desktop Linux distributions, the following things actually work in Android:
Sound mixing
Real-time software updates
Anti-malware sandboxing
Hardware acclerated video decoding
Android is selling by the truckload, it's doing far better than any other consumer open source OS ever has. Your customers want it because they know that the quality of an OS is not determined by how "hard coded" its "utility code" is - they can use it for themselves, they can see how well it's doing in the market place.
And as for the kernel arguments, wasn't that whole hooha debunked already? I mean, you know what the kernel "team" are like (I use the word team loosely). There's no clarity around who is responsible for approving or denying most changes. No decision can be made quickly. Everything you do will be flamed by someone and if you do get code merged, it'll quickly be rewritten or obsoleted by somebody else who may or may not have produced something usefully better. It's largely due to these high quality practices that desktop Linux remains a joke, to this day.
Android is what your customers demand because it works. If that's what low quality means, give me lower quality open source!
Is that unique to the NHS though? I find it unlikely that doctors would somehow be systematically more motivated or able under other systems.
If they are, I suspect it's a side effect of the system being free - whilst I'm sure your vision problem is quite real, I'd bet that the NHS gets a lot of hypochondriacs who have non-debilitating, difficult to diagnose illnesses which don't actually exist.
Also, no offence to your good self, but your blog says you've had quite a few tests including a full brain MRI which isn't at all cheap. I don't think that's really a great indictment of the NHS.
This might be a bit heavy for somebody as new to programming as you, but Java Puzzlers is a great book for Java devs of all skill levels to read. It's a series of small programs that screw up in ways you wouldn't expect, often related to bizarre gotchas in the design of Java itself. Newbies might not understand every puzzle, but generally it'll give you an appreciation for the fact that no tool is perfect, and insight into what sort of bugs you might find yourself writing in the real world.
Really? I don't think I've ever seen a 'scene' release that contained any kind of malware
That you know of. Remember that commercial AV packages have around an 80% miss rate, and some root kits go undetected on machines for years. If all you had to do was scan an EXE for viruses, we wouldn't be seeing botnets in the wild with millions of nodes - yet they exist.
I'd lose my progress once or twice a year if I did nothing all year but play Assassins Creed. Please think through your arguments first! And yes for people with shitty internet connections this will suck. But for people with children that eat DVDs, it will be better. Short of UbiSoft becoming a registered charity, some people will be inconvenienced by copy protection. The goal is to minimize that number.
The question is - more or less ridiculous than ever more aggressive DVD checks? On balance, I'd prefer being able to keep my DVDs somewhere safe and use an internet connection. Not saying UbiSoft will get it right, but I think internet based DRM will be an upgrade for most people if they do get it right.
Sure, DRM might be difficult for most pirates to overcome, but it only takes one pirate to crack it, and then the rest have access through torrents.
Yes, and UbiSoft know that. DRM on (PC) video games is all about the "time to crack". Look, this is how the modern piracy scene works and why UbiSoft are doing this.
A PC copy protection scheme on a major game will attract the attention of professional reverse engineers. These guys are likely paid to do it, because pirated games a the perfect way of getting people to install viruses and malware. The "pay per install" scene is largely based on torrents because it's so much easier than finding browser 0days.
Given this, it's not surprising that DVD-binding based DRM is weak. The techniques to defeat it are well known and the montary incentive is there.
So why move to internet based DRM? There are two reasons I can think of. Firstly, it's much stronger. Secondly, it solves the problem of people making backup DVDs, which is the traditional reason cited for why media binding is an unwanted technique. Internet connections these days are pretty damn reliable. Mine croaks maybe once or twice a year, and usually only for a few hours at worst. Trading a few hours of downtime a year for the ability to make backups seems like a pro-consumer move.
So what about strength? My gut feeling is that internet based DRM can be made significantly harder to break than media-binding based DRM. Even if it's still eventually done, if it reliably takes a month or two after release then it'll be considered a wild success. Consider the range of techniques available when an internet connection is active. The goal is to stop people sharing accounts, and to stop people removing the need for a connection. So, make every asset encrypted under a unique key that isn't stored on the DVD. As the player progresses through the game, it informs the server of where the player is up to. The server sends a small program to the game which then runs and gathers a hash of various bits of in-game state (like the values of certain memory locations) which "prove" the player has actually played that part of the game. The results of those hashes unlock the keys for the next areas. Of course all the usual anti-debugging tricks can be used, which are actually very effective (most cracks these days are about emulating the dvd drive rather than removing the checks, right).
In a non-linear game this approach will prove difficult to crack, because the cracker will have to play the game over and over to ensure he has actually reached every room, every level, every boss, every weapon. If he misses one, he produces an incomplete crack that will crash the game for some players. Of course the cracker might not care - pirated games are very often unstable and buggy compared to the retail version, as they only care about getting you to install their virus anyway. But it still increases the amount of work significantly.
AACS style broadcast encryption can be used to ensure every player who plays the game ends up with a uniquely watermarked/decrypted set of files, so the leaked version can be traced back to a credit card or buyer. So now the pirate also has to use a stolen CCN too. It's all about raising the bar.
How would that work? Even expert users can't easily know that an arbitrary CA follows a set of rules unless they are audited, and that's what the current process gives you (CNNIC passed the audits).
What makes Facebook so good is that it's all tied to people - even the fake accounts need to seem to be people.
Current prices for a facebook account on the spam markets are around $5 per 1000 friends. Creating fake accounts that seem to be people isn't that hard - just scrape pics of hot girls off MySpace, couple with a fake name generator and off you go (assuming you can get past Facebooks defences of course, but then GMail has defences too).
Imagine somebody turned up at your front door and say, hey, why don't you give me some money? I'll make it easier for you to donate to charity! And you say, wait, didn't I read about you being found guilty of something, and he says "yes but don't worry, I'm appealing!" - is that going to reassure you? It sure as hell wouldn't make me feel better.
But it gets worse for Mr "brokep". The Pirate Bay deliberately antagonized the people he is now trying to help. He ignored DMCA takedowns and told anyone who would listen why he was right and the rest of the world was wrong. During the trial where he was found guilty, he believed it was going "quite well". And his business partners have a history of harassment, like the "ddo$" scheme against the opposition law firm.
Like all the of the guys who ran the pirate bay, they have spent the last few years proving the world that they're stupid and untrustworthy. I can't think of a worse person to run something like Flattr.
The TPM is an open solution. The chips behavior is determined by open standards and there are multiple competing vendors of these chips.
The fact that you can mount sophisticated silicon attacks on a TPM is not a "flaw" because nobody knows how to make completely impenetrable chips. The TPM does what it was designed to do - provide a good level of security for very low cost. If you lose your laptop and it uses a TPM based product, chances are really great that the thieves won't get data out of it. That is not the same thing as "completely invulnerable to SEMs" and nobody ever claimed it was.
Um, why should opening your TV and replacing its components void the warranty? Because it's no longer the thing you bought, so expecting the manufactuers to fix it seems wildly unreasonable.
there are (or soon will be) numerous alternatives that are not as tightly locked.
Sure. Like the Nexus One. Not only is the SDK free, easy to program (java), flexible (you can replace most of the built in apps) but the phone itself isn't locked. Watch this video if you don't believe me... the shipping phone doesn't need a "jailbreak" because you can simply run an officially provided command and after informing you that you void the warranty, the phone will let you reflash to any OS (it changes the bootup logo to make it harder to resell trojaned/warranty voided phones but that seems reasonable).
This sort of hair splitting is merely playing with words; it has no intellectual merit.
The point of copyright is to blur the line between information and physical property, economically, to provide a market in which the fruits of intellectual labor can be traded. Pointing that blurryness out isn't making any interesting or new point, especially not here on Slashdot. You're just distracting from the essence of the debate by discussing what people said and not what they obviously meant.
Yeah but it's not like the iphone piracy scene has found a way to make a radically better app store. Besides, there's piracy on Android too, and you can get a refund for any app after 24 hours there. Basically no matter how great the deal is, some people will find an excuse to pirate.
The sad thing about this is that if you have a good quality product that meets the consumers needs and is at an affordable price, then people will buy them.
Actually, the sad thing is that this theory has been pretty much disproven in recent years by the iPhone phenomenon, in particular the way apps which cost $1 end up with 90% piracy rates (ie, rates comparable to desktop apps).
Pirates are, by definition, people who take something without paying for it. Whether an app costs $1 or $99 probably won't make much difference as long as piracy is equally convenient.
For example, I have moved over to Linux, but still play games through Wine. I try out (and regularly buy) several casual games and some of the bigger ones as well (like StarCraft). DRM on this software will make it harder to run on this platform, and will drive me away from those companies
Oddly enough, I used to work on Wine (have several hundred patches in there). So I've "examined" more than my fair share of copy protection schemes. There are two things you should know.
The first is that you're in a tiny minority and always will be. In my years of using Linux, its market share has never increased and shows no sign of doing so anytime soon.
The second thing is that fortunately, that probably won't matter in a few years. It seems likely that PC gaming DRM will move to internet binding rather than media binding in future, which is likely to not only make it far more robust but also make it a lot more compatible with emulators like Wine, because the game won't have any interest in poking around in kernel mode trying to distinguish fake DVD drives from real. So I wouldn't try and overgeneralize from the sort of DRM we have today to all DRM.
There are precious few people who can make their Windows machine as secure as *nix or Mac.
I think that's pretty inaccurate. Out of the box Windows is just as secure as Linux or a Mac is. It's something that Slashdotters don't like to discuss (easier to blame mothers and sisters I guess) but a lot of malware gets onto systems via warez and the like. For example, here is a point and click tutorial on uploading infected warez, even including how to avoid bans from torrent sites. The appeal of this option is obvious - there are lots of people out there downloading and running binaries from entirely untrustworthy sources. Unless you think using Mac or Windows makes somebody inherently more virtuous overnight, the only reason these platforms don't have the same problem is lack of market share.
Seriously now. Any multi-tasking solution in which apps randomly crash and lose data due to memory pressure cannot be said to have "smooth fluid handling".
Not that it makes a difference. For the price of the rocket you need to launch one panel, you can buy hundreds of panels. That will generate hundreds of times the power. It's an utterly stupid concept.
That doesn't make sense. The whole point of putting them in space is that they work better there. So if you had 1 panel in space and 100 on the ground, I don't know what the real ratio would be but it'd clearly not be 1:100.
Astrium isn't exactly a fly by night outfit. If they think they can get the numbers to where a panel in space is significantly more efficient than on the ground, it may not matter that it costs a lot to launch as the launch costs can be amortized over the lifetime of the satellite, the expected future cost of energy and so on.
You realize that Hadoop is a reimplementation of the MapReduce technology widely in use inside Google for a long time. Google invented it, filed a patent on it, published a paper on it, and Hadoop reimplemented it... then finally the US PTO granted the patent. Clear?
Slashdot Mirror
Baidu has reasonable enough results for the Chinese web, but doesn't really search the English web at all. Google.cn does both very well.
I am guessing their problem is that a lot of the papers they wish to read and general scientific world is based on English. As there aren't any other major search engines in China, having only Baidu would be close to having no search engine at all for people who need to search English documents.
Android really hasn't fragmented. Windows apps tend to work everywhere because developers are very slow to adopt new APIs. The same is true of Android: the new APIs in 2.0/2.1 are neat, but most of them are either convenience or purely candy (like live wallpapers or social networking integration). Many apps don't need to use them and they'll work just fine on older OS versions. And then of course, you have the fact that Android updates itself automatically over the air. The 2.1 release is only a month old, it's hardly unexpected that so far other devices don't have it.
As somebody who has written an app for both J2ME and Android, that comparison is really far out.
There are some key differences between J2ME and Android:
Android is an implementation, J2ME is a spec. Developing for J2ME is a nightmare because practically every phone has a unique combination of serious bugs. This isn't true of Android because every phone is working off the same core code. If manufacturers want to modify that code, that's OK, there's a test suite they have to pass if they want certain things.
Androids "fragmentation" is actually simply that not everyone has upgraded yet - no different to any other platform (including the iPhone). But Android manufacturers have a history of upgrading versions, and I don't see much reason for them to change this: customers demand it and because Android is open source, doing the upgrade is effectively "free" by mobile development standards, even for things like Sense UI (the APIs are backwards compatible).
You can see what versions are in use. So there's no guessing about what you can or cannot use, API wise, it's easy to see.
Android explicitly supports writing apps that run on multiple platform versions, where as J2ME never really did. For instance you can hide your app from earlier devices (in the market), or you can make your app gracefully degrade on older devices.
Android does actually standardise hardware, it just doesn't standardise on exactly one set. For instance, you can assume all Android devices have a decent touch screen, a display that uses one of a handful of resolutions and so on.
Google only lets manufacturers distribute certain closed source apps (like Market) if they meet certain conditions. Google has every incentive to keep as many people on the latest version of Android as possible. They have the leverage needed.
In short, it's a slow news day and InfoWorld know how to stir up a story.
Come on Bruce. Your argument for why Android low quality is that the kernel has patches and the "utility code is crap"?
Has it escaped your notice that unlike on the (presumably?) high quality desktop Linux distributions, the following things actually work in Android:
Android is selling by the truckload, it's doing far better than any other consumer open source OS ever has. Your customers want it because they know that the quality of an OS is not determined by how "hard coded" its "utility code" is - they can use it for themselves, they can see how well it's doing in the market place.
And as for the kernel arguments, wasn't that whole hooha debunked already? I mean, you know what the kernel "team" are like (I use the word team loosely). There's no clarity around who is responsible for approving or denying most changes. No decision can be made quickly. Everything you do will be flamed by someone and if you do get code merged, it'll quickly be rewritten or obsoleted by somebody else who may or may not have produced something usefully better. It's largely due to these high quality practices that desktop Linux remains a joke, to this day.
Android is what your customers demand because it works. If that's what low quality means, give me lower quality open source!
Is that unique to the NHS though? I find it unlikely that doctors would somehow be systematically more motivated or able under other systems.
If they are, I suspect it's a side effect of the system being free - whilst I'm sure your vision problem is quite real, I'd bet that the NHS gets a lot of hypochondriacs who have non-debilitating, difficult to diagnose illnesses which don't actually exist.
Also, no offence to your good self, but your blog says you've had quite a few tests including a full brain MRI which isn't at all cheap. I don't think that's really a great indictment of the NHS.
This might be a bit heavy for somebody as new to programming as you, but Java Puzzlers is a great book for Java devs of all skill levels to read. It's a series of small programs that screw up in ways you wouldn't expect, often related to bizarre gotchas in the design of Java itself. Newbies might not understand every puzzle, but generally it'll give you an appreciation for the fact that no tool is perfect, and insight into what sort of bugs you might find yourself writing in the real world.
That you know of. Remember that commercial AV packages have around an 80% miss rate, and some root kits go undetected on machines for years. If all you had to do was scan an EXE for viruses, we wouldn't be seeing botnets in the wild with millions of nodes - yet they exist.
I'd lose my progress once or twice a year if I did nothing all year but play Assassins Creed. Please think through your arguments first! And yes for people with shitty internet connections this will suck. But for people with children that eat DVDs, it will be better. Short of UbiSoft becoming a registered charity, some people will be inconvenienced by copy protection. The goal is to minimize that number.
Those 40% probably don't overlap much with the people who play Assassins Creed.
The question is - more or less ridiculous than ever more aggressive DVD checks? On balance, I'd prefer being able to keep my DVDs somewhere safe and use an internet connection. Not saying UbiSoft will get it right, but I think internet based DRM will be an upgrade for most people if they do get it right.
Yes, and UbiSoft know that. DRM on (PC) video games is all about the "time to crack". Look, this is how the modern piracy scene works and why UbiSoft are doing this.
A PC copy protection scheme on a major game will attract the attention of professional reverse engineers. These guys are likely paid to do it, because pirated games a the perfect way of getting people to install viruses and malware. The "pay per install" scene is largely based on torrents because it's so much easier than finding browser 0days.
Given this, it's not surprising that DVD-binding based DRM is weak. The techniques to defeat it are well known and the montary incentive is there.
So why move to internet based DRM? There are two reasons I can think of. Firstly, it's much stronger. Secondly, it solves the problem of people making backup DVDs, which is the traditional reason cited for why media binding is an unwanted technique. Internet connections these days are pretty damn reliable. Mine croaks maybe once or twice a year, and usually only for a few hours at worst. Trading a few hours of downtime a year for the ability to make backups seems like a pro-consumer move.
So what about strength? My gut feeling is that internet based DRM can be made significantly harder to break than media-binding based DRM. Even if it's still eventually done, if it reliably takes a month or two after release then it'll be considered a wild success. Consider the range of techniques available when an internet connection is active. The goal is to stop people sharing accounts, and to stop people removing the need for a connection. So, make every asset encrypted under a unique key that isn't stored on the DVD. As the player progresses through the game, it informs the server of where the player is up to. The server sends a small program to the game which then runs and gathers a hash of various bits of in-game state (like the values of certain memory locations) which "prove" the player has actually played that part of the game. The results of those hashes unlock the keys for the next areas. Of course all the usual anti-debugging tricks can be used, which are actually very effective (most cracks these days are about emulating the dvd drive rather than removing the checks, right).
In a non-linear game this approach will prove difficult to crack, because the cracker will have to play the game over and over to ensure he has actually reached every room, every level, every boss, every weapon. If he misses one, he produces an incomplete crack that will crash the game for some players. Of course the cracker might not care - pirated games are very often unstable and buggy compared to the retail version, as they only care about getting you to install their virus anyway. But it still increases the amount of work significantly.
AACS style broadcast encryption can be used to ensure every player who plays the game ends up with a uniquely watermarked/decrypted set of files, so the leaked version can be traced back to a credit card or buyer. So now the pirate also has to use a stolen CCN too. It's all about raising the bar.
Key changes are a part of life though. Your proposed solution can't distinguish between key rotation and attack, which is a non starter.
How would that work? Even expert users can't easily know that an arbitrary CA follows a set of rules unless they are audited, and that's what the current process gives you (CNNIC passed the audits).
Current prices for a facebook account on the spam markets are around $5 per 1000 friends. Creating fake accounts that seem to be people isn't that hard - just scrape pics of hot girls off MySpace, couple with a fake name generator and off you go (assuming you can get past Facebooks defences of course, but then GMail has defences too).
Yes, but under "common sense law" it does.
Imagine somebody turned up at your front door and say, hey, why don't you give me some money? I'll make it easier for you to donate to charity! And you say, wait, didn't I read about you being found guilty of something, and he says "yes but don't worry, I'm appealing!" - is that going to reassure you? It sure as hell wouldn't make me feel better.
But it gets worse for Mr "brokep". The Pirate Bay deliberately antagonized the people he is now trying to help. He ignored DMCA takedowns and told anyone who would listen why he was right and the rest of the world was wrong. During the trial where he was found guilty, he believed it was going "quite well". And his business partners have a history of harassment, like the "ddo$" scheme against the opposition law firm.
Like all the of the guys who ran the pirate bay, they have spent the last few years proving the world that they're stupid and untrustworthy. I can't think of a worse person to run something like Flattr.
Gah. This whole conversation is retarded.
Um, why should opening your TV and replacing its components void the warranty? Because it's no longer the thing you bought, so expecting the manufactuers to fix it seems wildly unreasonable.
Sure. Like the Nexus One. Not only is the SDK free, easy to program (java), flexible (you can replace most of the built in apps) but the phone itself isn't locked. Watch this video if you don't believe me ... the shipping phone doesn't need a "jailbreak" because you can simply run an officially provided command and after informing you that you void the warranty, the phone will let you reflash to any OS (it changes the bootup logo to make it harder to resell trojaned/warranty voided phones but that seems reasonable).
This sort of hair splitting is merely playing with words; it has no intellectual merit.
The point of copyright is to blur the line between information and physical property, economically, to provide a market in which the fruits of intellectual labor can be traded. Pointing that blurryness out isn't making any interesting or new point, especially not here on Slashdot. You're just distracting from the essence of the debate by discussing what people said and not what they obviously meant.
Yeah but it's not like the iphone piracy scene has found a way to make a radically better app store. Besides, there's piracy on Android too, and you can get a refund for any app after 24 hours there. Basically no matter how great the deal is, some people will find an excuse to pirate.
Actually, the sad thing is that this theory has been pretty much disproven in recent years by the iPhone phenomenon, in particular the way apps which cost $1 end up with 90% piracy rates (ie, rates comparable to desktop apps).
Pirates are, by definition, people who take something without paying for it. Whether an app costs $1 or $99 probably won't make much difference as long as piracy is equally convenient.
Oddly enough, I used to work on Wine (have several hundred patches in there). So I've "examined" more than my fair share of copy protection schemes. There are two things you should know.
The first is that you're in a tiny minority and always will be. In my years of using Linux, its market share has never increased and shows no sign of doing so anytime soon.
The second thing is that fortunately, that probably won't matter in a few years. It seems likely that PC gaming DRM will move to internet binding rather than media binding in future, which is likely to not only make it far more robust but also make it a lot more compatible with emulators like Wine, because the game won't have any interest in poking around in kernel mode trying to distinguish fake DVD drives from real. So I wouldn't try and overgeneralize from the sort of DRM we have today to all DRM.
I think that's pretty inaccurate. Out of the box Windows is just as secure as Linux or a Mac is. It's something that Slashdotters don't like to discuss (easier to blame mothers and sisters I guess) but a lot of malware gets onto systems via warez and the like. For example, here is a point and click tutorial on uploading infected warez, even including how to avoid bans from torrent sites. The appeal of this option is obvious - there are lots of people out there downloading and running binaries from entirely untrustworthy sources. Unless you think using Mac or Windows makes somebody inherently more virtuous overnight, the only reason these platforms don't have the same problem is lack of market share.
LOL. But did you lose any data?
Seriously now. Any multi-tasking solution in which apps randomly crash and lose data due to memory pressure cannot be said to have "smooth fluid handling".
That doesn't make sense. The whole point of putting them in space is that they work better there. So if you had 1 panel in space and 100 on the ground, I don't know what the real ratio would be but it'd clearly not be 1:100.
Astrium isn't exactly a fly by night outfit. If they think they can get the numbers to where a panel in space is significantly more efficient than on the ground, it may not matter that it costs a lot to launch as the launch costs can be amortized over the lifetime of the satellite, the expected future cost of energy and so on.
You realize that Hadoop is a reimplementation of the MapReduce technology widely in use inside Google for a long time. Google invented it, filed a patent on it, published a paper on it, and Hadoop reimplemented it ... then finally the US PTO granted the patent. Clear?