And of course, it's an excellent opportunity to shaft the patients a bit more.
The latest revision to the International Classification of Diseases has had an *explosion* of complexity. Ostensibly this is to make it more accurate. What I suspect it's really for is to make it easier to make an error. Because if you make an error in medical records, your HMO can deny you payment.
Version 2 : The number one complaint I hear about version 2 is that the extensibility features are abused. Yes, it's a standard that defines a way to make it's messages non-standard, and most people exploit that feature to excess.
It's a fairly simple character-delimited-text protocol. Even so, I've seen implementations screw it up so badly that their HL7 patient admin interface didn't even emit valid HL7 messages.
That said, I'm sure there are bits of HL7 V2 that are successful, most notably the parts regarding billing patients.
Version 3 : Version three is incredibly complex and difficult to implement systems on.
And the kicker - no-one in the corporate market wants standards. They want their own system to rule everywhere. I've specified message formats, and the company involved have just come right back and told me flat out that they won't be implementing the standard for unknown field values, they'll be keeping their stupid magic numbers, thank you very much. Then I've had the management agree with them because the contract says that if we change that part of their system they can demand to be paid millions for a full system test.
The dirty secret of healthcare is that you don't need the government for safety. You need consumers to be more vigilant and involved in their own care.
In other words, consumers have to become medical experts, and polymaths at that, when all the best paid doctors are highly specialized.
Bullshit. There's a reason the phrase for people selling something that doesn't work is "snake oil". Healthcare is a complex subject. If it's too complex for a single professional to grasp the entirety of it, what hope do consumers have?
The dirty secret of healthcare is that HMOs exist to deny you treatment, because that's how they make more money, and that they co-opt doctors to help them. The only bits of healthcare record interoperability that really work and have been around for a while? The bits governing billing.
In both single payer, and HMO models, there is cost cutting going on. But at least with single-payer, the motive is that they want to cut costs so they can do the most net good to all their patients. With the HMO, they want to cut costs so they can do the most net good to their shareholders pockets.
Who would you rather give your hard earned dollar to, someone who has your back? Or someone who wants to take the shirt off it?
I worked for the NHS, then for suppliers serving the NHS, now I work for the NHS again (in the department that used to be the National Programme for IT). The problem is Cathedral mentality. Rather than do something simple that you can expand, everyone wants an all-singing, all-dancing, solves every problem out of the box EHR system.
But everyone wants a system that will conform to the little local quirks - in effect, they want their current system, But With A Computer (tm).
One of my hobbies when I was developing EHR systems was collecting hospital stationery, so I could compare their forms and charts. You'd think that in a single-payer system like the NHS, there would be a single set of forms and papers for everything, but they all have their own.
So in order to enter the market, you have to make a system that can basically be configured to do anything - a platform for making EHR systems, if you like. Or you have to force the players to use the same standard system (not just the software - but how it's used, how it's configured, what the standard data is, etc), which would be much better.
Our National Programme for IT (in the NHS) was a much-publicised £12B failure.
Can you imagine what could have been achieved if that had been spent properly? We could have instituted a programme producing standard, Free (as in speech) software for solving healthcare IT problems. Even if they'd just shoved £12B into a savings account and made software with the interest, we'd probably have some really kick ass software (and have thrown a lot of dross away in the process of getting it).
Instead, we were tasked with producing systems that try to make the mess of corporate systems interoperate properly. As far as I can tell, this will continue to be the goal of central NHS IT planning, because this is the approach that favours spending public money on corporate services the most, and this seems to be the most prominent common theme amongst all UK politicians at the moment.
The easiest way to make systems interoperate, as Slashdotters already know, is to make them all the same system. It works for Office - whatever you want to think about Microsoft, you can exchange files between computers running Office and expect compatibility.
The downside is that this doesn't work for the NHS - every site has it's own peculiar quirks and requirements, and typically require extended pre-sales and after-sales support to get their systems configured. The staff required to do this do not scale, so even the largest of companies does not have the capacity to service the whole market simultaneously or even expeditiously.
If the software were Free, of course, you could rapidly scale an industry of consultants who made it their profession to configure and support the software. But that would put a kink in the rails of the gravy train for the large incumbent players, so it's unlikely to be permitted.
If the whole *system* were Free - instead of just the software, the system of managing EHR data for the hospital - the task might be less herculean. But you'd have to convince every player to give up their local quirks and start marching to the same tune. Which isn't going to happen either - people defend their personal fiefdoms.
Whenever I hear the name "VistA" I shudder inwardly. Why? Because it's written in MUMPS. I mean, FFS, this is a language that has had two articles all to itself on DailyWTF.
For my sins, I had to do some work on a system written in MUMPS. I guess it's a rite of passage that you just have to endure in the healthcare IT world if you want to graduate to the more wizardly ranks. I had to deal with it for a mere two days. I never want to see another line of MUMPS code again.
This includes the right to access a copy of the information comprised in their personal data, so you are effectively a joint owner of all data concerning you, except where it runs into one of the exceptions like purposes of national security, crime, taxation, and data held merely domestically.
The data holder is permitted to charge you a "reasonable fee" to process the data if you request a copy.
MIT were more dragged into it by the Feds. MIT had no desire to pursue legal action against Swartz. The Feds pushed this matter, one might infer, because they had an axe to grind with him.
You'd think so, but some of the e-books I've bought were *terrible* in terms of typesetting, almost like they OCRed the print copy to make the ebook. Anathem being one of them.
Indeed - they could mandate that the cash registers are also TiVoized to prevent them running anything but the approved build, but then it isn't Free software.
SSD are good for random IO of lots of tiny files because of the low latency, but spinning rust is about equivalent for bulk IO. If the bulk transfer speed of the media approaches the speed of the SATA bus, there isn't any real difference.
If you're a half decent game developer, you write your levels as large files that can be read sequentially and not randomly. SSDs should only improve loading times for badly written games.
Spinning rust is cheap, and game textures are large.
That's true ; but mostly because the tactics of the tax accountants have evolved faster than legislation, which has been deliberately retarded and loosened by conservatives to produce these loopholes.
For example, the loophole exploited by Starbucks UK, where they pay a subsidiary in another country with a lower tax rate for their beans, etc. They pay a rate that is so far removed from what coffee beans actually cost that their entire UK operation hasn't booked a profit in years, despite taking billions over the counter. Now they have "voluntarily" offered to pay a mere £20m in tax over the next three years, a figure which is both insulting and ludicrous. We know they are worming out of their responsibilities, but it's all perfectly legal.
Coffee houses that are entirely based in the UK don't have the facilities to get away with this bullshit, so they are at a disadvantage. That's the kind of "why would they try?" that you want to be worried about - the complete antithesis of what capitalists think they believe in, a healthy fair market where innovation and agility matter more than the amount of capital you hold.
The original game is no longer compelling and doesn't really stand up well against games from systems with more capacity.
It was groundbreaking and amazing at the time, from a 9 year old child's eyes. But even then the combat became laughably simple after you upgraded your ship.
For Nokia, not so much. If they were ready to acknowledge that they couldn't go on with their own separate phone OS any more, they should have bitten the bullet, embraced Android, and done what Nokia do best - good hardware with good antennae.
Instead they were sold the pipe dream of remaining special and recapturing the market, which they were all too eager to buy into. A shame it was a tar baby. Microsoft will pick over the corpse of their corporation for the juicy bits in a few years.
It means they've done usability testing, and have numbers to back up their assertion that Unity works well.
I quite like it, TBH.
Many of the features are just Compiz features, things like windows snapping to screen positions, etc.
The window button placements make total sense with regard to the placement of launcher buttons (in both Unity and the older GNOME 2 desktop) - the buttons are in the closest corner to the button you are going to push next - the one that starts a new app, switches to a new app, etc. It does this better than both OSX (buttons top left, launcher bottom middle) and Windows (buttons top right, launcher bottom left).
The HUD menu is great, especially when you are using an IDE like Eclipse (after hacking the blacklist so it works again..) with about a hundred menu items. I just wish it worked with more of the widget toolkits used for GUI apps on Linux.
Yes, it's an adjustment to get used to it, but the frustration you initially feel is no different to me walking up to a Windows machine and finding that ctrl-alt-T doesn't open a terminal, the default CLI interface is horribly crippled, and the "full function" shell is distinctly different in terms of usage and convention.
The Amazon thing was ill-judged though. By all means, monetize peoples searches (everyone else does..), but only the ones you intended to be directing to an online shopping site.
I hated that mealy-mouthed bullshit so much the first time around that on subsequent playthroughs I always just menaced the shopkeepers into giving me a discount instead.
When you're hashing your password, you only have to do it once. Your multiplier is 1.
When you has to defeat a password, you have to do it as many times as you need to, to match the previously hashed password. Let's say you have a really crap password (a 6 digit number) and your attacker knows this. Your multiplier is between 1 and 1,000,000
The hash is the difficulty factor. If your hash is optimized for throughput, it's cheap. The advantage is that it's really good for establishing data integrity - large volumes of data pass through it quickly. This is the opposite of what you want in a password hash, where the data volume is small.
If we take hypothetical hashes A and B, where the cost of A is 1 and the cost of B is 1,000 , we can see that the cost of hashing data with B is too much for bulk data, but the cost of doing it for passwords makes it harder for an attacker, who has to hash a vast amount of data compared to someone logging into their account.
A : 1 x 1 - 1,000,000 = 1 - 1,000,000 B : 1,000 x 1 - 1,000,000 = 1,000 - 1,000,000,000
You might also hypothesise that the barbed sting delivers a larger dose to the thicker skinned larger animal which needs a bigger deterrent, because it stays stuck in their hide, which is advantageous to the colony because they learn not to attack the hive.
Bees die when they sting humans at least - their sting is barbed and they can't get it out, so the injuries they receive when they rip part of their butt off are fatal. It only applies to animals with a thick skin/
If this membrane is tough enough, they won't be able to penetrate it, and their stinger will stay on. It would also work if it's weak enough to pull the stinger loose.
If you block it, how do you catch paedos by tracing it?
Child porno already has relevant laws covering it. You don't need to add a means of censorship which can and will be arbitrarily extended to cover anything else the government doesn't like, like people who don't like them.
The UK is much the same ; EU law forces everything to be labelled in metric but people generally still think* in pounds and ounces and pints and feet and inches, partly because of history, partly because they are more "human" units.
No-one ever said "I'm going down the pub for a 0.57 of a litre". Except in beer commercials taking the piss out of metric.
* We think in imperial. We calculate in metric. Converting chains (22 yards, the length of a cricket pitch) to metres is a bitch.
Slashdot Mirror
And of course, it's an excellent opportunity to shaft the patients a bit more.
The latest revision to the International Classification of Diseases has had an *explosion* of complexity. Ostensibly this is to make it more accurate. What I suspect it's really for is to make it easier to make an error. Because if you make an error in medical records, your HMO can deny you payment.
There are two versions of HL7
Version 2 : The number one complaint I hear about version 2 is that the extensibility features are abused. Yes, it's a standard that defines a way to make it's messages non-standard, and most people exploit that feature to excess.
It's a fairly simple character-delimited-text protocol. Even so, I've seen implementations screw it up so badly that their HL7 patient admin interface didn't even emit valid HL7 messages.
That said, I'm sure there are bits of HL7 V2 that are successful, most notably the parts regarding billing patients.
Version 3 : Version three is incredibly complex and difficult to implement systems on.
And the kicker - no-one in the corporate market wants standards. They want their own system to rule everywhere. I've specified message formats, and the company involved have just come right back and told me flat out that they won't be implementing the standard for unknown field values, they'll be keeping their stupid magic numbers, thank you very much. Then I've had the management agree with them because the contract says that if we change that part of their system they can demand to be paid millions for a full system test.
The dirty secret of healthcare is that you don't need the government for safety. You need consumers to be more vigilant and involved in their own care.
In other words, consumers have to become medical experts, and polymaths at that, when all the best paid doctors are highly specialized.
Bullshit. There's a reason the phrase for people selling something that doesn't work is "snake oil". Healthcare is a complex subject. If it's too complex for a single professional to grasp the entirety of it, what hope do consumers have?
The dirty secret of healthcare is that HMOs exist to deny you treatment, because that's how they make more money, and that they co-opt doctors to help them. The only bits of healthcare record interoperability that really work and have been around for a while? The bits governing billing.
In both single payer, and HMO models, there is cost cutting going on. But at least with single-payer, the motive is that they want to cut costs so they can do the most net good to all their patients. With the HMO, they want to cut costs so they can do the most net good to their shareholders pockets.
Who would you rather give your hard earned dollar to, someone who has your back? Or someone who wants to take the shirt off it?
I worked for the NHS, then for suppliers serving the NHS, now I work for the NHS again (in the department that used to be the National Programme for IT). The problem is Cathedral mentality. Rather than do something simple that you can expand, everyone wants an all-singing, all-dancing, solves every problem out of the box EHR system.
But everyone wants a system that will conform to the little local quirks - in effect, they want their current system, But With A Computer (tm).
One of my hobbies when I was developing EHR systems was collecting hospital stationery, so I could compare their forms and charts. You'd think that in a single-payer system like the NHS, there would be a single set of forms and papers for everything, but they all have their own.
So in order to enter the market, you have to make a system that can basically be configured to do anything - a platform for making EHR systems, if you like. Or you have to force the players to use the same standard system (not just the software - but how it's used, how it's configured, what the standard data is, etc), which would be much better.
Yes, it's because of greed.
Our National Programme for IT (in the NHS) was a much-publicised £12B failure.
Can you imagine what could have been achieved if that had been spent properly? We could have instituted a programme producing standard, Free (as in speech) software for solving healthcare IT problems. Even if they'd just shoved £12B into a savings account and made software with the interest, we'd probably have some really kick ass software (and have thrown a lot of dross away in the process of getting it).
Instead, we were tasked with producing systems that try to make the mess of corporate systems interoperate properly. As far as I can tell, this will continue to be the goal of central NHS IT planning, because this is the approach that favours spending public money on corporate services the most, and this seems to be the most prominent common theme amongst all UK politicians at the moment.
The easiest way to make systems interoperate, as Slashdotters already know, is to make them all the same system. It works for Office - whatever you want to think about Microsoft, you can exchange files between computers running Office and expect compatibility.
The downside is that this doesn't work for the NHS - every site has it's own peculiar quirks and requirements, and typically require extended pre-sales and after-sales support to get their systems configured. The staff required to do this do not scale, so even the largest of companies does not have the capacity to service the whole market simultaneously or even expeditiously.
If the software were Free, of course, you could rapidly scale an industry of consultants who made it their profession to configure and support the software. But that would put a kink in the rails of the gravy train for the large incumbent players, so it's unlikely to be permitted.
If the whole *system* were Free - instead of just the software, the system of managing EHR data for the hospital - the task might be less herculean. But you'd have to convince every player to give up their local quirks and start marching to the same tune. Which isn't going to happen either - people defend their personal fiefdoms.
Whenever I hear the name "VistA" I shudder inwardly. Why? Because it's written in MUMPS. I mean, FFS, this is a language that has had two articles all to itself on DailyWTF.
For my sins, I had to do some work on a system written in MUMPS. I guess it's a rite of passage that you just have to endure in the healthcare IT world if you want to graduate to the more wizardly ranks. I had to deal with it for a mere two days. I never want to see another line of MUMPS code again.
Depends on the jurisdiction.
Under the terms of the Data Protection Act 1998 (which itself aligns the UK with EU legislation), the individual has rights with respect to data held about them.
This includes the right to access a copy of the information comprised in their personal data, so you are effectively a joint owner of all data concerning you, except where it runs into one of the exceptions like purposes of national security, crime, taxation, and data held merely domestically.
The data holder is permitted to charge you a "reasonable fee" to process the data if you request a copy.
MIT were more dragged into it by the Feds. MIT had no desire to pursue legal action against Swartz. The Feds pushed this matter, one might infer, because they had an axe to grind with him.
ebooks are also subject to tax in some jurisdictions, where paper books are not, the UK being one.
You'd think so, but some of the e-books I've bought were *terrible* in terms of typesetting, almost like they OCRed the print copy to make the ebook. Anathem being one of them.
Indeed - they could mandate that the cash registers are also TiVoized to prevent them running anything but the approved build, but then it isn't Free software.
CCTV does not... without gait recognition software. Or other kinds of image processing.
SSD are good for random IO of lots of tiny files because of the low latency, but spinning rust is about equivalent for bulk IO. If the bulk transfer speed of the media approaches the speed of the SATA bus, there isn't any real difference.
If you're a half decent game developer, you write your levels as large files that can be read sequentially and not randomly. SSDs should only improve loading times for badly written games.
Spinning rust is cheap, and game textures are large.
That's true ; but mostly because the tactics of the tax accountants have evolved faster than legislation, which has been deliberately retarded and loosened by conservatives to produce these loopholes.
For example, the loophole exploited by Starbucks UK, where they pay a subsidiary in another country with a lower tax rate for their beans, etc. They pay a rate that is so far removed from what coffee beans actually cost that their entire UK operation hasn't booked a profit in years, despite taking billions over the counter. Now they have "voluntarily" offered to pay a mere £20m in tax over the next three years, a figure which is both insulting and ludicrous. We know they are worming out of their responsibilities, but it's all perfectly legal.
Coffee houses that are entirely based in the UK don't have the facilities to get away with this bullshit, so they are at a disadvantage. That's the kind of "why would they try?" that you want to be worried about - the complete antithesis of what capitalists think they believe in, a healthy fair market where innovation and agility matter more than the amount of capital you hold.
They're doing him a favour - if he stops drink driving, he's much less likely to end up in a body bag, or worse, maimed or in jail.
Or if you're a real man, you leave your number under his windshield wiper, fess up (to the hit, not the drinking) and pay for the damage.
Then you stop being a murderous punk-ass little bitch who'd drink and then handle a giant steel lethal weapon.
The original game is no longer compelling and doesn't really stand up well against games from systems with more capacity.
It was groundbreaking and amazing at the time, from a 9 year old child's eyes. But even then the combat became laughably simple after you upgraded your ship.
Elop did the right thing for Microsoft.
For Nokia, not so much. If they were ready to acknowledge that they couldn't go on with their own separate phone OS any more, they should have bitten the bullet, embraced Android, and done what Nokia do best - good hardware with good antennae.
Instead they were sold the pipe dream of remaining special and recapturing the market, which they were all too eager to buy into. A shame it was a tar baby. Microsoft will pick over the corpse of their corporation for the juicy bits in a few years.
It means they've done usability testing, and have numbers to back up their assertion that Unity works well.
I quite like it, TBH.
Many of the features are just Compiz features, things like windows snapping to screen positions, etc.
The window button placements make total sense with regard to the placement of launcher buttons (in both Unity and the older GNOME 2 desktop) - the buttons are in the closest corner to the button you are going to push next - the one that starts a new app, switches to a new app, etc. It does this better than both OSX (buttons top left, launcher bottom middle) and Windows (buttons top right, launcher bottom left).
The HUD menu is great, especially when you are using an IDE like Eclipse (after hacking the blacklist so it works again..) with about a hundred menu items. I just wish it worked with more of the widget toolkits used for GUI apps on Linux.
Yes, it's an adjustment to get used to it, but the frustration you initially feel is no different to me walking up to a Windows machine and finding that ctrl-alt-T doesn't open a terminal, the default CLI interface is horribly crippled, and the "full function" shell is distinctly different in terms of usage and convention.
The Amazon thing was ill-judged though. By all means, monetize peoples searches (everyone else does..), but only the ones you intended to be directing to an online shopping site.
I hated that mealy-mouthed bullshit so much the first time around that on subsequent playthroughs I always just menaced the shopkeepers into giving me a discount instead.
When you're hashing your password, you only have to do it once. Your multiplier is 1.
When you has to defeat a password, you have to do it as many times as you need to, to match the previously hashed password. Let's say you have a really crap password (a 6 digit number) and your attacker knows this. Your multiplier is between 1 and 1,000,000
The hash is the difficulty factor. If your hash is optimized for throughput, it's cheap. The advantage is that it's really good for establishing data integrity - large volumes of data pass through it quickly. This is the opposite of what you want in a password hash, where the data volume is small.
If we take hypothetical hashes A and B, where the cost of A is 1 and the cost of B is 1,000 , we can see that the cost of hashing data with B is too much for bulk data, but the cost of doing it for passwords makes it harder for an attacker, who has to hash a vast amount of data compared to someone logging into their account.
A : 1 x 1 - 1,000,000 = 1 - 1,000,000
B : 1,000 x 1 - 1,000,000 = 1,000 - 1,000,000,000
You might also hypothesise that the barbed sting delivers a larger dose to the thicker skinned larger animal which needs a bigger deterrent, because it stays stuck in their hide, which is advantageous to the colony because they learn not to attack the hive.
Bees die when they sting humans at least - their sting is barbed and they can't get it out, so the injuries they receive when they rip part of their butt off are fatal. It only applies to animals with a thick skin/
If this membrane is tough enough, they won't be able to penetrate it, and their stinger will stay on. It would also work if it's weak enough to pull the stinger loose.
If you block it, how do you catch paedos by tracing it?
Child porno already has relevant laws covering it. You don't need to add a means of censorship which can and will be arbitrarily extended to cover anything else the government doesn't like, like people who don't like them.
The UK is much the same ; EU law forces everything to be labelled in metric but people generally still think* in pounds and ounces and pints and feet and inches, partly because of history, partly because they are more "human" units.
No-one ever said "I'm going down the pub for a 0.57 of a litre". Except in beer commercials taking the piss out of metric.
* We think in imperial. We calculate in metric. Converting chains (22 yards, the length of a cricket pitch) to metres is a bitch.