I don't think any amount of education is going to help americans be healthier, but I'm a bit of a pessimist about my fellow citizens so maybe it's not as bad as I think. I was making a mathematical observation about concerns with your choice of statistics for correlation, I certainly wasn't disputing your main point. Where this a live discussion, I suspect we'd be in broad agreement but in a friendly discussion of detail; I'd have already bought you a beer because I did like Gapminder. Or something healthier, if you wanted to be unAmerican about it.
The fact that there are other correlations to be found here is sort of reinforcing the only thing I was highlighting to comment on: there are a lot more factors going into this mess than just health-care spending. The %GDP/capita metric is particularly suspect, because the GDP itself has its own trends to be considered. I have a pretty good picture of that in my head for that from years of studying the stock market, and you really need to overlay the raw GDP on all this to try to extract out gross economic trends. If the GDP grows due to general US business trends, but health care spending stays the same, that will shrink the %GDP/capita even with no policy change. Some of the funny business in the 80's was due to that, and 1980 and 1992 were both in the middle of significant turning points for the country's economy (both exits from mild recessions lined up with major administration changes). Five years from now, there's going to be another discontinuity right in the middle of 2008 too that looks just like it, except possibly reversed due to the country's productivity shrinking. There's a whole second level argument here around things like how bad economic times make people less healthy, which wanders all the way from stress to food consumption, and that adds a whole new dimension to trying to extract cause and effect from this sort of data.
Any way you measure it, health care in the US is more expensive than anywhere else relative to the quality of care it delivers, suggesting there is a massive improvement in efficiency possible. But as a realist who has spent a lot of time learning how US corporations and its government really work, actually doing something useful about it is one of the most difficult problems I've ever thought seriously about a solution to. That and the IRS are just mind-boggling problems to unwind, yet both are central to why the US is run so inefficiently relative to the rest of the modern countries.
The economics of the US health system are as screwed up as you say, but you may have picked the wrong cause for your correlation here. Anyone who visits both the UK and the US regularly will tell you their life expectancy is shorter because the US population is so fat and out of shape. As an English friend who was visiting me here recently put it, "I'm the fattest guy in my village, and I feel like a bloody model here".
Gentoo recently postponed using GNOME3 because it seemed like a "work in progress". Meanwhile, Fedora has shipped it, Ubuntu is now on the even less mature Ubiquity, and CentOS can't even get a modern release shipped out the door at all. Gentoo is looking like a stable Linux aimed at old geezers nowadays.
The upthread suggestion was that once you discover a misrouted message, you immediately reconfigure to dump all additional mail from that source into the trash. That's how you delete something without reading it. If I stop five lines into the message and say "this isn't meant for me", dump in the trash, and put all future copies into there automatically, I can certainly claim I never read even a single full e-mail from that source. Misrouted e-mails are toxic; unless you really do know the intended recipient, you have to minimize your exposure to them and eliminate them as quickly as possible from your life.
I'm not sure why you made that "if you'd bother reading.." comment as if I was disagreeing with you. I know, it is so rare now that someone replies to a message to add support to that person's opinion instead of trying to roast them.
It would take a much worse legal system than we currently have to reach a point where deleting an e-mail without reading it can get you into trouble. The opposite, acting on an e-mail you shouldn't have received, certainly can. There's a very good reason legal proceedings so often still require proof of physical delivery of something to a person in order to proceed. If you are physically served with a legal document and you are not its intended recipient, you damn well better make sure that's cleared up lest you become a guilty party too. E-mail intended for someone else? Get rid of that as soon as possible.
The only thing you didn't get quite right here is painting this picture as if it's specific to IT. You'll find the same cast of characters in every type of work: the lifer who will stab you to keep his deal going, the older guys getting kicked out the minute it's possible to replace them with a younger model, and the loss of any loyalty. This is business in the US in 2011, and the only unusual bit is that IT is new enough still that some workers can remember a time when it was different. I've spent a lot of my life working in manufacturing companies of various sorts. The brilliant mechanical guy who knows how to fix all the machines when they break, the one the whole factory would go down without, he gets the exact same fist.
Good luck Googling for the answer when your Cisco router is down. I have never used a vendor supplied documentation CD as heavily as the ones that came with my Cisco gear.
They've improved, but not eliminated, the problem I was commenting on. The data you want to operate on often starts in L1 cache of the CPU at the point where it might make sense to off-load to the GPU. The speed drop involved in the context switch to the GPU environment, which then has to fetch the data from main memory, is still large. Fusion drops the point at which the switch can make sense. But the interconnect to the GPU was already quite fast before, and it still got in the way. The speed advantage of code already working on a problem in the CPU, sitting in its instruction and data cache, has over transferring that same problem to the GPU is large. You really have to offload larger operations than you might think before it makes sense, or do as I suggested and move a lot of code onto the GPU.
As opposed to GTK+, where the project is healthy, the toolkit project is changing rapidly, and GNOME's future is uncertain because there's a giant user backlash over the changes.
Doing larger sorting operations usually involve breaking the problem into multiple data sets, then merging the sorted subsets back together again. If done right, you can get each of the GPU units working on sorting their own piece most of the time. The UNC results are typical, and note that data sizes to be sorted now are much, much larger than the right side of their graph--so the slower growing runtime is even more important.
Also, sorting time can take up a lot of the CPU resources on a busy database server, so being able to offload that portion to the GPU means more CPU time available for other tasks.
The odds of AMD inventing unicorns and saving the company with their sale are better than I'd give the idea that OpenCL will become popular.
I work with databases all day, and we regularly get people who say "why can't you accelerate sorting using a GPU?" The reason why you can't is that by the time you transmit the whole problem set over to the GPU, wait for it to compute, and then transfer the results back to the CPU again, you could have just sorted it on the CPU. This problem, that you have to load/process/return everything from the GPU, only goes away if they are capable of running much more general software. I'd have to move the entire database query executor onto the GPU, and it would need enough memory to do significant tasks, before it made sense here.
I see OpenCL continuing to be more popular in scientific computing applications, making far less nodes required in the computing clusters they tend to run. It's hard to imagine another area they have any real potential to be popular in.
You can buy a $5 Mini LED flashlight from the EFF that makes the yellow dots much more visible. I like to keep it around for when people accuse me of being too paranoid when I'm ranting about privacy and such. Grab the nearest color printout, shine the blue light on it, and show there are hidden yellow dots that can be used to track you; that's a good way to make people really nervous as they consider what else I'm saying is true. More fashionable than tin foil, too.
Also, your printer may work better if you plug the right kind of clock into the outlet next to it. Between that and the green rectangle trick, my printouts are so realistic they give me a giant Mpingo woodie every time I look at them.
Not just currently. Today's organizational turmoil within CentOS is nothing compared to when they lost access to much of the infrastructure a few years ago. I just wrote a blog entry on the rise of and fall of CentOS; the theme is why it's important to build an open community, not a tight clique, if you want an open-source project to scale.
ed is also two characters, and they are adjacent on the keyboard. If you are a touch-typist, it takes 50% of the hands and 50% of the fingers necessary to start that inefficient vi editor.
Blocks are now reallocated by hardware in the drive. You need to get some practice using whatever SMART utility is available for your OS. Look at the SMART attribute for "reallocated sectors" to figure out if this has been happening to you. I generally start migrating to another drive once any appear. You can't see them in software any other way now, the old "bad blocks" are now just reallocated when they occur.
As for testing every sector, kick off a SMART "extended self test". That does what you want. When it's finished, the SMART log will tell you what the test returned. If you see the reallocated sector count go upwards after the test, but it still passes, you should start getting suspicious of the drive.
Most drive manufacturers also have a "write all zeros" function available on the utility CDs they provide for testing drives out. Alternately you can boot from a Linux live CD and do something like "dd if=/dev/zero of=/dev/sda" to wipe them out.
The Unicomp keyboards are not as good as the original Model M designs. Lexmark changed the design in 1995 to make it cheaper to build, and every Model M since then has been made from cheaper parts than the earlier ones. Unicomp's changes have continued that trend downward. The best years of Model M are 1993 and 1994: by then they had switched to the PS/2 connector (earlier ones may have the old AT connector), but it's before the quality started dropping.
uids on Debian derived systems start with 1000. If you build a system based on either RedHat or Debian standards right now, the uid numbers are going to be in a completely different range. This makes it even more annoying than usual to do things like create shared volumes where the file permissions are right, when mounted by two different servers running different distributions.
It's also a hassle on dual-boot systems, or ones where you converted from one distribution to the other, too. For example, last week I changed a RHEL server to run Debian. I had to go through 2TB of files and switch all of the UIDs over to the Debian standard for them, mapping user 500 -> 1000. Even though I could have just made the main account use the default RedHat range UID, that 500 number is technically reserved in Debian.
Small things, but definitely moving in the right direction as far as making Linux distributions less different from one another.
Well, sure, if you only have 512GB of RAM on your laptop, you're not going to be able to use all the GNOME Shell capabilities. Systems with less than a terabyte of RAM will have to adopt lightweight alternatives like Compiz or Plasma. Be sure to upgrade your hardware if you want to take advantages of all the latest wizardry from the GNOME project!
You missed where he mentioned "sales and marketing, closing new business". Every other part of a startup has trivial problems to solve compared to those. Programmers, project managers, administration, all irrelevant unless you figure out how to turn that work into money.
Yes, that's what Linus will say. I was commenting on how businesses will perceive things, regardless of the technical message that comes along with it. "2.8.1" or even worse "3.0.1" will be considered toxic no matter how it's presented to business people. And with so many distributions lined up with the same kernel version right now, it's a decent time to do it; I think the sort of places that think like that will be happy with the currently available options until enough Linux version bumps that the number doesn't sound as scary.
Slashdot Mirror
I don't think any amount of education is going to help americans be healthier, but I'm a bit of a pessimist about my fellow citizens so maybe it's not as bad as I think. I was making a mathematical observation about concerns with your choice of statistics for correlation, I certainly wasn't disputing your main point. Where this a live discussion, I suspect we'd be in broad agreement but in a friendly discussion of detail; I'd have already bought you a beer because I did like Gapminder. Or something healthier, if you wanted to be unAmerican about it.
The fact that there are other correlations to be found here is sort of reinforcing the only thing I was highlighting to comment on: there are a lot more factors going into this mess than just health-care spending. The %GDP/capita metric is particularly suspect, because the GDP itself has its own trends to be considered. I have a pretty good picture of that in my head for that from years of studying the stock market, and you really need to overlay the raw GDP on all this to try to extract out gross economic trends. If the GDP grows due to general US business trends, but health care spending stays the same, that will shrink the %GDP/capita even with no policy change. Some of the funny business in the 80's was due to that, and 1980 and 1992 were both in the middle of significant turning points for the country's economy (both exits from mild recessions lined up with major administration changes). Five years from now, there's going to be another discontinuity right in the middle of 2008 too that looks just like it, except possibly reversed due to the country's productivity shrinking. There's a whole second level argument here around things like how bad economic times make people less healthy, which wanders all the way from stress to food consumption, and that adds a whole new dimension to trying to extract cause and effect from this sort of data.
Any way you measure it, health care in the US is more expensive than anywhere else relative to the quality of care it delivers, suggesting there is a massive improvement in efficiency possible. But as a realist who has spent a lot of time learning how US corporations and its government really work, actually doing something useful about it is one of the most difficult problems I've ever thought seriously about a solution to. That and the IRS are just mind-boggling problems to unwind, yet both are central to why the US is run so inefficiently relative to the rest of the modern countries.
The economics of the US health system are as screwed up as you say, but you may have picked the wrong cause for your correlation here. Anyone who visits both the UK and the US regularly will tell you their life expectancy is shorter because the US population is so fat and out of shape. As an English friend who was visiting me here recently put it, "I'm the fattest guy in my village, and I feel like a bloody model here".
Gentoo recently postponed using GNOME3 because it seemed like a "work in progress". Meanwhile, Fedora has shipped it, Ubuntu is now on the even less mature Ubiquity, and CentOS can't even get a modern release shipped out the door at all. Gentoo is looking like a stable Linux aimed at old geezers nowadays.
The upthread suggestion was that once you discover a misrouted message, you immediately reconfigure to dump all additional mail from that source into the trash. That's how you delete something without reading it. If I stop five lines into the message and say "this isn't meant for me", dump in the trash, and put all future copies into there automatically, I can certainly claim I never read even a single full e-mail from that source. Misrouted e-mails are toxic; unless you really do know the intended recipient, you have to minimize your exposure to them and eliminate them as quickly as possible from your life.
I'm not sure why you made that "if you'd bother reading.." comment as if I was disagreeing with you. I know, it is so rare now that someone replies to a message to add support to that person's opinion instead of trying to roast them.
It would take a much worse legal system than we currently have to reach a point where deleting an e-mail without reading it can get you into trouble. The opposite, acting on an e-mail you shouldn't have received, certainly can. There's a very good reason legal proceedings so often still require proof of physical delivery of something to a person in order to proceed. If you are physically served with a legal document and you are not its intended recipient, you damn well better make sure that's cleared up lest you become a guilty party too. E-mail intended for someone else? Get rid of that as soon as possible.
I think the author could use a bit more intellect in the area of making sure their blog will survive before submitting a self-promotional link.
The only thing you didn't get quite right here is painting this picture as if it's specific to IT. You'll find the same cast of characters in every type of work: the lifer who will stab you to keep his deal going, the older guys getting kicked out the minute it's possible to replace them with a younger model, and the loss of any loyalty. This is business in the US in 2011, and the only unusual bit is that IT is new enough still that some workers can remember a time when it was different. I've spent a lot of my life working in manufacturing companies of various sorts. The brilliant mechanical guy who knows how to fix all the machines when they break, the one the whole factory would go down without, he gets the exact same fist.
Good luck Googling for the answer when your Cisco router is down. I have never used a vendor supplied documentation CD as heavily as the ones that came with my Cisco gear.
That's not the whole story. Don't forget that Ubuntu has shat out a turd called Ubiquity, too.
With GNOME 3.0, Ubiquity, and the uncertainty around QT/KDE, I haven't been this scared for my Linux desktop in years.
They've improved, but not eliminated, the problem I was commenting on. The data you want to operate on often starts in L1 cache of the CPU at the point where it might make sense to off-load to the GPU. The speed drop involved in the context switch to the GPU environment, which then has to fetch the data from main memory, is still large. Fusion drops the point at which the switch can make sense. But the interconnect to the GPU was already quite fast before, and it still got in the way. The speed advantage of code already working on a problem in the CPU, sitting in its instruction and data cache, has over transferring that same problem to the GPU is large. You really have to offload larger operations than you might think before it makes sense, or do as I suggested and move a lot of code onto the GPU.
As opposed to GTK+, where the project is healthy, the toolkit project is changing rapidly, and GNOME's future is uncertain because there's a giant user backlash over the changes.
Doing larger sorting operations usually involve breaking the problem into multiple data sets, then merging the sorted subsets back together again. If done right, you can get each of the GPU units working on sorting their own piece most of the time. The UNC results are typical, and note that data sizes to be sorted now are much, much larger than the right side of their graph--so the slower growing runtime is even more important.
Also, sorting time can take up a lot of the CPU resources on a busy database server, so being able to offload that portion to the GPU means more CPU time available for other tasks.
The odds of AMD inventing unicorns and saving the company with their sale are better than I'd give the idea that OpenCL will become popular.
I work with databases all day, and we regularly get people who say "why can't you accelerate sorting using a GPU?" The reason why you can't is that by the time you transmit the whole problem set over to the GPU, wait for it to compute, and then transfer the results back to the CPU again, you could have just sorted it on the CPU. This problem, that you have to load/process/return everything from the GPU, only goes away if they are capable of running much more general software. I'd have to move the entire database query executor onto the GPU, and it would need enough memory to do significant tasks, before it made sense here.
I see OpenCL continuing to be more popular in scientific computing applications, making far less nodes required in the computing clusters they tend to run. It's hard to imagine another area they have any real potential to be popular in.
I can't think of any single hobby that gives 20/30/40-something guys more of an excuse to burn through cash like there's no tomorrow.
I see you've never met someone who owns a boat.
The goat actually felt more violated by how Zuckerberg treated his personal privacy than he did about the whole throat cutting thing.
You can buy a $5 Mini LED flashlight from the EFF that makes the yellow dots much more visible. I like to keep it around for when people accuse me of being too paranoid when I'm ranting about privacy and such. Grab the nearest color printout, shine the blue light on it, and show there are hidden yellow dots that can be used to track you; that's a good way to make people really nervous as they consider what else I'm saying is true. More fashionable than tin foil, too.
Also, your printer may work better if you plug the right kind of clock into the outlet next to it. Between that and the green rectangle trick, my printouts are so realistic they give me a giant Mpingo woodie every time I look at them.
Not just currently. Today's organizational turmoil within CentOS is nothing compared to when they lost access to much of the infrastructure a few years ago. I just wrote a blog entry on the rise of and fall of CentOS; the theme is why it's important to build an open community, not a tight clique, if you want an open-source project to scale.
ed is also two characters, and they are adjacent on the keyboard. If you are a touch-typist, it takes 50% of the hands and 50% of the fingers necessary to start that inefficient vi editor.
Blocks are now reallocated by hardware in the drive. You need to get some practice using whatever SMART utility is available for your OS. Look at the SMART attribute for "reallocated sectors" to figure out if this has been happening to you. I generally start migrating to another drive once any appear. You can't see them in software any other way now, the old "bad blocks" are now just reallocated when they occur.
As for testing every sector, kick off a SMART "extended self test". That does what you want. When it's finished, the SMART log will tell you what the test returned. If you see the reallocated sector count go upwards after the test, but it still passes, you should start getting suspicious of the drive.
Most drive manufacturers also have a "write all zeros" function available on the utility CDs they provide for testing drives out. Alternately you can boot from a Linux live CD and do something like "dd if=/dev/zero of=/dev/sda" to wipe them out.
The Unicomp keyboards are not as good as the original Model M designs. Lexmark changed the design in 1995 to make it cheaper to build, and every Model M since then has been made from cheaper parts than the earlier ones. Unicomp's changes have continued that trend downward. The best years of Model M are 1993 and 1994: by then they had switched to the PS/2 connector (earlier ones may have the old AT connector), but it's before the quality started dropping.
uids on Debian derived systems start with 1000. If you build a system based on either RedHat or Debian standards right now, the uid numbers are going to be in a completely different range. This makes it even more annoying than usual to do things like create shared volumes where the file permissions are right, when mounted by two different servers running different distributions.
It's also a hassle on dual-boot systems, or ones where you converted from one distribution to the other, too. For example, last week I changed a RHEL server to run Debian. I had to go through 2TB of files and switch all of the UIDs over to the Debian standard for them, mapping user 500 -> 1000. Even though I could have just made the main account use the default RedHat range UID, that 500 number is technically reserved in Debian.
Small things, but definitely moving in the right direction as far as making Linux distributions less different from one another.
Well, sure, if you only have 512GB of RAM on your laptop, you're not going to be able to use all the GNOME Shell capabilities. Systems with less than a terabyte of RAM will have to adopt lightweight alternatives like Compiz or Plasma. Be sure to upgrade your hardware if you want to take advantages of all the latest wizardry from the GNOME project!
You missed where he mentioned "sales and marketing, closing new business". Every other part of a startup has trivial problems to solve compared to those. Programmers, project managers, administration, all irrelevant unless you figure out how to turn that work into money.
Yes, that's what Linus will say. I was commenting on how businesses will perceive things, regardless of the technical message that comes along with it. "2.8.1" or even worse "3.0.1" will be considered toxic no matter how it's presented to business people. And with so many distributions lined up with the same kernel version right now, it's a decent time to do it; I think the sort of places that think like that will be happy with the currently available options until enough Linux version bumps that the number doesn't sound as scary.