> Why, in the name of the FSM, did my data have "stated income" as a yes-or-no field? Nobody sane would make a major loan to somebody who couldn't even verify his or her income, unless they had absolutely no concern about whether it would be paid
That is very close to the crux of the problem. So why *did* companies make those loans? Because they believed the fiction of their own flawed models that said they could predict the failure rates of such mortgages.
Simplistically, they just look historically at the percentage of such mortgages that fail and figured out what interest rate they needed to charge to ensure they still made a profit. Maybe even as many as 80% of them might fail, but you can still make profit if a) you charge enough interest and b) the value of the asset backing the loan doesn't fall. The fiction is that you can combine the failure probabilities together and lessen the risk.
Consider: if two mortgages with 80% chance of failure are combined together, what is the probability they will both fail? Simplistically, 64%, right? So keep on grinding, keep adding more... it keeps falling! Amazing! But it's completely wrong... these variables are not independent, it's high school statistics to know you cannot multiply the probabilities to compute the overall failure likelihood. But they essentially fell for a much more sophisticated version of this and keep computing lower and lower risk until someone like you or I would buy it and put it in our pension fund (or more correctly, we would choose a pension fund that would put it in there, but it comes to the same thing).
Mandate a standardized record format, and mandate that all patients can get their data in this format from anybody who provides them a medical service - full, complete, and free of charge. Done.
Once you have a standard format all else follows: there will immediately be a boom in services for maintaining patient controlled health records and existing services will move support the standard (Google Health, etc.). Then once the data is portable patients will suddenly be able to take their data from one place to another and the free market will begin to work. Small startup insurance companies will arrive and quote insurance based on your health record. Patients will no longer be locked into existing medical service providers and will seek the best, taking their data where they go. Medical service providers will no longer be locked into their IT vendors and will choose the best instead of whatever their existing vendor wants to shove down their throat. ISVs will need to compete based on quality and usability instead of how much lock-in they have.
It can all happen once the data is free, and it will never happen unless it is.
> What if the information you wanted withheld was critical, but you simply don't have the background or knowledge to know that it's critical?
If there is missing information that is critical then any responsible doctor should be capable of recognizing that - they will be in the same position as any doctor with no information and should ask questions and order fresh tests. You're right that this may cause delay and perhaps result in some slight increase in risk. However it also gives patients the chance to get a genuine second opinion which is vital.
It's interesting that they are attacking IE8 and not just IE6 & 7. They had good technical arguments for attacking IE6. IE7 somewhat, but almost none for IE8. This brings them 1 step closer to getting into monopolist problems with Chrome - they can hardly go to town claiming Microsoft is pushing IE unfairly on Windows when they themselves are cross-fertilizing their own browser from their search business.
No. Let's be clear: the legal issue is that they assist in copyright violation, not the name of their web site. The name of their web site is merely evidence that assists in proving their intent. Neither one by itself is enough but both combined together are sufficient.
You talk as if intent has nothing to do with the law / liability. In reality, where most of us live, intent is crucial to the law.
Killing someone intentionally is a massively different crime to killing them accidentally - in some cases it's not a crime at all (eg. self defense). It's ALL about intent. The problem for the pirate bay is exactly that their name expresses their intent to facilitate illegal activity.
I dare say a wide straight road named 'speed limit breakers highway', started by a company that charged tolls on the highway and obfuscated police efforts to control speeding there would also be on shaky ground.
> TPB is just linking to material. They don't host it. Yes, they 'make it easier to infringe', but the line between what TPB is doing and what e.g. the roads are doing (helping bank robbers get away, the horror!) is one of degree, and more importantly, it isn't clear where the line is
The problem with the pirate bay is that the name of their site directly expresses their intent to facilitate copyright infringement. If the did exactly what they do today but were called 'legalfilelinks.com' I would have no particular problem with it. To extend the bank robber / road analogy - what if a company built a road right to the rear of a bank where their vault was located, chopped away the wall so the vault was exposed and then named the road 'bank robber highway', and finally charged a 50c toll for each vehicle that passed through? I would say that the clearly expressed intent of the company's actions made them liable and would be completely different to a company that built a toll road that just happened to go past the bank.
Ummm, even Sun says you should not be launching threads inside of JEE containers. Or writing to the file system for that matter. In fact, most of Google's restrictions are identical to Sun's own restrictions for JEE, the only difference is, Sun didn't make it mandatory, and now they're complaining that Google are (because they have to actually host these things in a sane and manageable way).
The only thing that really irks me is no socket connections. Making HTTP the only avenue for network I/O, and even worse using Sun's crappy URL classes to do it, limits a whole lot of potentially useful applications one could write.
> Last time it was 19 updates. Came out somewhere around 32MB. I have a 24Mb connection, it took a long, long time to download them. It then sat and took the rest of the hour applying them, shutting down, applying some more during shutdown, booting and applying more during startup. And then it found more. It's slow and a bit of a shambles.
It was probably creating a system restore point, which on Vista may mean anything up to creating a shadow copy of every file on your system so that you can (at your option) revert to precisely the configuration you had before applying the updates, should there be any problem. Does your debian update do that as well?
Dynamic languages are gaining popularity extremely fast for a whole host of reasons. If you haven't used a dynamic language then you probably won't understand why they are attractive very easily, but if you have it will probably seem obvious. They do have a lot of downsides and I think the jury is still out to some extent whether the reams of code now being written in them is going to be maintainable down the track.
Groovy is Java made into a dynamic language and with innumerable convenience features added. The best way to describe it is to imagine if java was turned into Ruby.
I suspect one reason Google has focused on it is because of Grails, which is java's main player in the RAD web-app development arena (more or less the equivalent of Django, which they also support for the Python app engine). It doesn't work on App Engine yet, but I strongly suspect that is where they are headed.
Others have said it all, but I will throw in my lot too: groovy has a number of problems that weigh very heavily on productivity: the first and most important is that it's a dynamic language taken to an extreme. This means that the compiler never knows the type of any object and lets just about anything pass without complaining. For example, the following compiles fine but explodes at runtime:
int i = "cat"
You see, groovy is *soooo* groovy, that someone just might have modified things under the hood at runtime so that the = operator *will* work to assign a string to an int. So the compiler really can only verify that the most rudimentary syntax of the language is right and nothing else. If you start developing a large code base with this almost non-existent compiler checking of types, things start to fall apart quickly. You don't see this much at the start because most groovy is written on top of strongly typed java libraries which do have the benefit of strict typing. Once you start developing large groovy code bases in their own right it starts to hurt. The other problems are very poor error reporting (inscrutable 3000 line stack traces when exceptions occur) and poor IDE support (which I suspect partly flows from the loose nature of the language).
Scala solves all these problems. Thus I'm currently considering dropping groovy for Scala.
What makes you think the liberals won't support the filter?
Listen carefully to Minchin et al., they think the filter is a great idea and have pretty much said so outright on a number of occasions. The only thing they criticize is Labour's implementation of it.
> When you have someone offering you much more then your companies worth...you take it run and never look back. Especially with the bad economy.
Wow, so using that logic you must think that now is an excellent time to sell your house, right? Right after a massive collapse in prices and when nobody has any money or can get a loan - wait until there are at least 5 "for sale" signs in your street - the perfect time to sell!
I think you're 100% wrong - now is the worst time for any company to sell itself because there is literally no money out there except vultures looking for dead meat. Wait a few years and sell when the market is up. Sun has a lot of problems but they are not dead meat yet, and they shouldn't behave that way.
I agree with you in a sense, but I would put it rather differently: the problem with java is that it DID succeed in delivering homogenization and it turned out that nobody wanted it. The lowest common denominator of any group of things is just an all round sucky experience in general. Had java failed to deliver that, strangely enough, it just might have succeeded better on the desktop (because alternative libraries like SWT might have had a chance at becoming dominant and saved it).
One of the things that I find offensive about this whole debate is that the notion of censoring these kind of things is in fact in itself a proclamation that they are intrinsically dirty, sexual, and forbidden where in their original context they were not.
What this does is in fact sexualises children even more and makes them more "attractive" to those who have such leanings. For many who go down this path it is the very forbidden and unusual nature of it that excites them. I suspect that if the whole internet and every day life was filled with pictures of nude children there would probably be less child abuse than there is now because a whole segment of adults would completely lose interest.
I don't disagree with you about the irresponsibility of people letting their children get so far away.
But I also think you need to temper your own feelings in cases like the child in the road. Obviously I don't know the details, but (and I don't know if you are a parent), it is seriously hard to control little children. They go through phases of really wanting to run away and get extremely good at it. I'm not at all condoning the situation, but I know with great familiarity the challenge of trying to manage kids in car parks and the difficulty posed in controlling them. In the end, we are humans and raising children is a fundamental part of our lives and roles on this planet. They need to go out in the world to face and see dangers in order to learn to survive them. There's no choice of complete safety that works. We need to take a communal responsibility towards looking after their safety while they are doing this.
You did a great job spotting and stopping for the child and you should be relieved and proud that you are a good driver and not feel guilty at all. The parent has learned a lesson and should be supported to get better at controlling their kids. We all need to be working together to do this and not get angry at each other.
Yes and no. Just because the experiment had a negative result doesn't mean we learned anything from it. Maybe we could have predicted this outcome all along in which case we learned nothing. Maybe there are other ways to do the experiment that might avoid this outcome, in which case we have only proved it doesn't work by precisely this method and have learned *almost* nothing.
It *may* be that this is a useful negative result, but you can't jump to the conclusion that the experiment itself was a success unless the result truly helps us generalize and understand some fundamental aspects of the problem better.
That's why they dog food the OSes internally for years during development as has been documented previously.
It is actually harder to get the interface right for new users exactly because all the internal staff are expert users. Windows has plenty of shortcuts for experts. There's nothing wrong with MS trying to make it right for both groups by testing on non-experts - they are a legitimate and large portion of the users of windows.
You wrote that whole rant about a version of Windows that is now almost 2 major releases old and which has been superseded by a version whose primary focus was improving the kind of security problems you cite.
"At the last election the labour party promised to implement optional internet filtering in Australia. After the election you broke that promise by changing the scheme to mandatory filtering for all internet connections, something for which you had no mandate based on your election platform.
The question is firstly, why did you break that promise? and secondly, given that you broke such an important promise, how can Australians now trust you to implement this scheme without further violating their precious democratic right to free speech?"
Australia and New Zealand are small nations where the people very much look to their governments to care for all their needs in a very paternalistic kind of way. I attribute this to the fact that they were born from peaceful colonies with no violent history or war to gain their independence or freedom.
Example - we just had a bunch of guys walk into Sydney airport (supposedly one of the most secure areas in the country), pick up the posts of the security railing and beat a man for 15 minutes until he died in front of dozens of on-lookers. No police or any kind of security intervened during this time. The conclusion of the politicians and Australian public? They want more laws to increase police powers! The fact that no police were actually on the scene, and that beating someone to death might just already possibly be illegal enough to give police power to intervene doesn't seem to have occurred to anybody. The solution, it seems, is to give police the power to arrest anybody they want based on mere suspicion or association.
> That's why that right is protected in the US Bill of Rights.
Absolutely! Just make sure you confine your protest to the mandated "free speech zones" and stick to the pre-ordained "free speech times" while you are protesting. Let me see, I think today's zone is at the disused factory 25 miles out of town and a full 15 minutes has been alotted for protesting. Let freedom ring!
> At first, I sort of liked Windows 7 and ran it for a couple of weeks. Then I booted back into XP (not a fresh installation of XP, mind you). I was depressed by how much snappier XP feels. I was hoping to have a good reason to ditch XP.
You may be correct, but consider also that it may just be the "feel" of the OS that you are noticing, not the speed per se. All those fancy animated window effects (fading in and out, sliding etc.) that Vista and 7 do actually do reduce the feel of "snappiness" of the OS and make things seem slower. Try turning them all off and see how it feels. Many ordinary users like the effects but for power users they just get in the way.
Yes - massively so, than before, though the gains are not as much as other browsers. This link is old but you can see even in beta 2 IE8 was at least 4 times as fast as IE7 and that was before a lot of optimisation was done. In the final cut I would expect an order of magnitude improvement - which of course, is *still* less than other browsers, but nonetheless it approaches being game changing in terms of what you can do... if only we didn't have to support IE6.... argh!
Slashdot Mirror
> Why, in the name of the FSM, did my data have "stated income" as a yes-or-no field? Nobody sane would make a major loan to somebody who couldn't even verify his or her income, unless they had absolutely no concern about whether it would be paid
That is very close to the crux of the problem. So why *did* companies make those loans? Because they believed the fiction of their own flawed models that said they could predict the failure rates of such mortgages.
Simplistically, they just look historically at the percentage of such mortgages that fail and figured out what interest rate they needed to charge to ensure they still made a profit. Maybe even as many as 80% of them might fail, but you can still make profit if a) you charge enough interest and b) the value of the asset backing the loan doesn't fall. The fiction is that you can combine the failure probabilities together and lessen the risk.
Consider: if two mortgages with 80% chance of failure are combined together, what is the probability they will both fail? Simplistically, 64%, right? So keep on grinding, keep adding more ... it keeps falling! Amazing! But it's completely wrong ... these variables are not independent, it's high school statistics to know you cannot multiply the probabilities to compute the overall failure likelihood. But they essentially fell for a much more sophisticated version of this and keep computing lower and lower risk until someone like you or I would buy it and put it in our pension fund (or more correctly, we would choose a pension fund that would put it in there, but it comes to the same thing).
I actually think it's completely the reverse.
Mandate a standardized record format, and mandate that all patients can get their data in this format from anybody who provides them a medical service - full, complete, and free of charge. Done.
Once you have a standard format all else follows: there will immediately be a boom in services for maintaining patient controlled health records and existing services will move support the standard (Google Health, etc.). Then once the data is portable patients will suddenly be able to take their data from one place to another and the free market will begin to work. Small startup insurance companies will arrive and quote insurance based on your health record. Patients will no longer be locked into existing medical service providers and will seek the best, taking their data where they go. Medical service providers will no longer be locked into their IT vendors and will choose the best instead of whatever their existing vendor wants to shove down their throat. ISVs will need to compete based on quality and usability instead of how much lock-in they have.
It can all happen once the data is free, and it will never happen unless it is.
> What if the information you wanted withheld was critical, but you simply don't have the background or knowledge to know that it's critical?
If there is missing information that is critical then any responsible doctor should be capable of recognizing that - they will be in the same position as any doctor with no information and should ask questions and order fresh tests. You're right that this may cause delay and perhaps result in some slight increase in risk. However it also gives patients the chance to get a genuine second opinion which is vital.
It's interesting that they are attacking IE8 and not just IE6 & 7. They had good technical arguments for attacking IE6. IE7 somewhat, but almost none for IE8. This brings them 1 step closer to getting into monopolist problems with Chrome - they can hardly go to town claiming Microsoft is pushing IE unfairly on Windows when they themselves are cross-fertilizing their own browser from their search business.
No. Let's be clear: the legal issue is that they assist in copyright violation, not the name of their web site. The name of their web site is merely evidence that assists in proving their intent. Neither one by itself is enough but both combined together are sufficient.
You talk as if intent has nothing to do with the law / liability. In reality, where most of us live, intent is crucial to the law.
Killing someone intentionally is a massively different crime to killing them accidentally - in some cases it's not a crime at all (eg. self defense). It's ALL about intent. The problem for the pirate bay is exactly that their name expresses their intent to facilitate illegal activity.
I dare say a wide straight road named 'speed limit breakers highway', started by a company that charged tolls on the highway and obfuscated police efforts to control speeding there would also be on shaky ground.
> TPB is just linking to material. They don't host it. Yes, they 'make it easier to infringe', but the line between what TPB is doing and what e.g. the roads are doing (helping bank robbers get away, the horror!) is one of degree, and more importantly, it isn't clear where the line is
The problem with the pirate bay is that the name of their site directly expresses their intent to facilitate copyright infringement. If the did exactly what they do today but were called 'legalfilelinks.com' I would have no particular problem with it. To extend the bank robber / road analogy - what if a company built a road right to the rear of a bank where their vault was located, chopped away the wall so the vault was exposed and then named the road 'bank robber highway', and finally charged a 50c toll for each vehicle that passed through? I would say that the clearly expressed intent of the company's actions made them liable and would be completely different to a company that built a toll road that just happened to go past the bank.
Ummm, even Sun says you should not be launching threads inside of JEE containers. Or writing to the file system for that matter. In fact, most of Google's restrictions are identical to Sun's own restrictions for JEE, the only difference is, Sun didn't make it mandatory, and now they're complaining that Google are (because they have to actually host these things in a sane and manageable way).
The only thing that really irks me is no socket connections. Making HTTP the only avenue for network I/O, and even worse using Sun's crappy URL classes to do it, limits a whole lot of potentially useful applications one could write.
> Last time it was 19 updates. Came out somewhere around 32MB. I have a 24Mb connection, it took a long, long time to download them. It then sat and took the rest of the hour applying them, shutting down, applying some more during shutdown, booting and applying more during startup. And then it found more. It's slow and a bit of a shambles.
It was probably creating a system restore point, which on Vista may mean anything up to creating a shadow copy of every file on your system so that you can (at your option) revert to precisely the configuration you had before applying the updates, should there be any problem. Does your debian update do that as well?
Dynamic languages are gaining popularity extremely fast for a whole host of reasons. If you haven't used a dynamic language then you probably won't understand why they are attractive very easily, but if you have it will probably seem obvious. They do have a lot of downsides and I think the jury is still out to some extent whether the reams of code now being written in them is going to be maintainable down the track.
Groovy is Java made into a dynamic language and with innumerable convenience features added. The best way to describe it is to imagine if java was turned into Ruby.
I suspect one reason Google has focused on it is because of Grails, which is java's main player in the RAD web-app development arena (more or less the equivalent of Django, which they also support for the Python app engine). It doesn't work on App Engine yet, but I strongly suspect that is where they are headed.
Others have said it all, but I will throw in my lot too: groovy has a number of problems that weigh very heavily on productivity: the first and most important is that it's a dynamic language taken to an extreme. This means that the compiler never knows the type of any object and lets just about anything pass without complaining. For example, the following compiles fine but explodes at runtime:
int i = "cat"
You see, groovy is *soooo* groovy, that someone just might have modified things under the hood at runtime so that the = operator *will* work to assign a string to an int. So the compiler really can only verify that the most rudimentary syntax of the language is right and nothing else. If you start developing a large code base with this almost non-existent compiler checking of types, things start to fall apart quickly. You don't see this much at the start because most groovy is written on top of strongly typed java libraries which do have the benefit of strict typing. Once you start developing large groovy code bases in their own right it starts to hurt. The other problems are very poor error reporting (inscrutable 3000 line stack traces when exceptions occur) and poor IDE support (which I suspect partly flows from the loose nature of the language).
Scala solves all these problems. Thus I'm currently considering dropping groovy for Scala.
What makes you think the liberals won't support the filter?
Listen carefully to Minchin et al., they think the filter is a great idea and have pretty much said so outright on a number of occasions. The only thing they criticize is Labour's implementation of it.
> When you have someone offering you much more then your companies worth...you take it run and never look back. Especially with the bad economy.
Wow, so using that logic you must think that now is an excellent time to sell your house, right? Right after a massive collapse in prices and when nobody has any money or can get a loan - wait until there are at least 5 "for sale" signs in your street - the perfect time to sell!
I think you're 100% wrong - now is the worst time for any company to sell itself because there is literally no money out there except vultures looking for dead meat. Wait a few years and sell when the market is up. Sun has a lot of problems but they are not dead meat yet, and they shouldn't behave that way.
I agree with you in a sense, but I would put it rather differently: the problem with java is that it DID succeed in delivering homogenization and it turned out that nobody wanted it. The lowest common denominator of any group of things is just an all round sucky experience in general. Had java failed to deliver that, strangely enough, it just might have succeeded better on the desktop (because alternative libraries like SWT might have had a chance at becoming dominant and saved it).
One of the things that I find offensive about this whole debate is that the notion of censoring these kind of things is in fact in itself a proclamation that they are intrinsically dirty, sexual, and forbidden where in their original context they were not.
What this does is in fact sexualises children even more and makes them more "attractive" to those who have such leanings. For many who go down this path it is the very forbidden and unusual nature of it that excites them. I suspect that if the whole internet and every day life was filled with pictures of nude children there would probably be less child abuse than there is now because a whole segment of adults would completely lose interest.
I don't disagree with you about the irresponsibility of people letting their children get so far away.
But I also think you need to temper your own feelings in cases like the child in the road. Obviously I don't know the details, but (and I don't know if you are a parent), it is seriously hard to control little children. They go through phases of really wanting to run away and get extremely good at it. I'm not at all condoning the situation, but I know with great familiarity the challenge of trying to manage kids in car parks and the difficulty posed in controlling them. In the end, we are humans and raising children is a fundamental part of our lives and roles on this planet. They need to go out in the world to face and see dangers in order to learn to survive them. There's no choice of complete safety that works. We need to take a communal responsibility towards looking after their safety while they are doing this.
You did a great job spotting and stopping for the child and you should be relieved and proud that you are a good driver and not feel guilty at all. The parent has learned a lesson and should be supported to get better at controlling their kids. We all need to be working together to do this and not get angry at each other.
Yes and no. Just because the experiment had a negative result doesn't mean we learned anything from it. Maybe we could have predicted this outcome all along in which case we learned nothing. Maybe there are other ways to do the experiment that might avoid this outcome, in which case we have only proved it doesn't work by precisely this method and have learned *almost* nothing.
It *may* be that this is a useful negative result, but you can't jump to the conclusion that the experiment itself was a success unless the result truly helps us generalize and understand some fundamental aspects of the problem better.
That's why they dog food the OSes internally for years during development as has been documented previously.
It is actually harder to get the interface right for new users exactly because all the internal staff are expert users. Windows has plenty of shortcuts for experts. There's nothing wrong with MS trying to make it right for both groups by testing on non-experts - they are a legitimate and large portion of the users of windows.
You wrote that whole rant about a version of Windows that is now almost 2 major releases old and which has been superseded by a version whose primary focus was improving the kind of security problems you cite.
I would like the following question asked:
"At the last election the labour party promised to implement optional internet filtering in Australia. After the election you broke that promise by changing the scheme to mandatory filtering for all internet connections, something for which you had no mandate based on your election platform.
The question is firstly, why did you break that promise? and secondly, given that you broke such an important promise, how can Australians now trust you to implement this scheme without further violating their precious democratic right to free speech?"
Australia and New Zealand are small nations where the people very much look to their governments to care for all their needs in a very paternalistic kind of way. I attribute this to the fact that they were born from peaceful colonies with no violent history or war to gain their independence or freedom.
Example - we just had a bunch of guys walk into Sydney airport (supposedly one of the most secure areas in the country), pick up the posts of the security railing and beat a man for 15 minutes until he died in front of dozens of on-lookers. No police or any kind of security intervened during this time. The conclusion of the politicians and Australian public? They want more laws to increase police powers! The fact that no police were actually on the scene, and that beating someone to death might just already possibly be illegal enough to give police power to intervene doesn't seem to have occurred to anybody. The solution, it seems, is to give police the power to arrest anybody they want based on mere suspicion or association.
> That's why that right is protected in the US Bill of Rights.
Absolutely! Just make sure you confine your protest to the mandated "free speech zones" and stick to the pre-ordained "free speech times" while you are protesting. Let me see, I think today's zone is at the disused factory 25 miles out of town and a full 15 minutes has been alotted for protesting. Let freedom ring!
> At first, I sort of liked Windows 7 and ran it for a couple of weeks. Then I booted back into XP (not a fresh installation of XP, mind you). I was depressed by how much snappier XP feels. I was hoping to have a good reason to ditch XP.
You may be correct, but consider also that it may just be the "feel" of the OS that you are noticing, not the speed per se. All those fancy animated window effects (fading in and out, sliding etc.) that Vista and 7 do actually do reduce the feel of "snappiness" of the OS and make things seem slower. Try turning them all off and see how it feels. Many ordinary users like the effects but for power users they just get in the way.
One subtlety often missed is that raw javascript performance itself is not enough.
What matters is Javascript + DOM - how fast the browser can react to changes to the DOM by javascript.
Some browsers blitz the pure JS benchmarks but then do DOM manipulation slowly which is not so useful.
> Does IE have better javascript performance?
Yes - massively so, than before, though the gains are not as much as other browsers. This link is old but you can see even in beta 2 IE8 was at least 4 times as fast as IE7 and that was before a lot of optimisation was done. In the final cut I would expect an order of magnitude improvement - which of course, is *still* less than other browsers, but nonetheless it approaches being game changing in terms of what you can do ... if only we didn't have to support IE6 .... argh!
http://ejohn.org/blog/javascript-performance-rundown/