Re:Not a bad idea, but treat with caution.
on
Wii Check-Up Channel
·
· Score: 3, Interesting
I got a Wii Fit back in September and lost twenty pounds by New Year's Day. I was earning 30 Wii Fit points per day for most of that period, plus I was walking with my kids two miles every Saturday on a local nature trail.
Unfortunately, between the New Year's Eve party and a bunch of unhealthy left overs, I managed to put about a third of that back. Plus, bad weather has kept me off the nature trail, which certainly doesn't help. And all of that has hurt my motivation, so where before I'd done over 100 consecutive days of exercise, now I'm skipping more days than I'm doing. I've apparently fallen into the trap described in this article. Overall, though, I'm still doing better than most years, where I would put on ten pounds between Thanksgiving and New Year's, and never take it off. Damn you, homemade English Toffee!
BTW, Nintendo has created a pedometer that talks to your DS. The cartridge allows you to download your Mii for an experience similar to Wii Fit. What's missing is apparently any way to upload data into Wii Fit, but that may show up in a future channel.
If they'd made it possible to reflash, a zillion Linux weenies would have bought the devices just to put Rockbox on them.
I hadn't realized that you couldn't reflash a Zune. That's going to make it pretty hard to patch that leap-year bug, don't you think?
Speaking of which, I wonder what their sales figures looked like for the first three weeks of January. After this weekend, sales will definitely tank, but I'm sure somebody at the Harvard Business Review would like to know how many nails that bug put into the coffin.
I graduated from college in 1978 and moved to Silicon Valley. In a short time, I owned an Apple II, an Advent VideoBeam projection TV, and a copy of Bruce Artwick's original Flight Simulator. With a nominal 64" screen, it seemed as immersive as any commercial flight simulator of that time (even if it was low-res black-and-white wireframe). And when you bought it, you got a second floppy with the source code! I'm saddened to see the franchise end after 30+ years. Yes, I know that they'll probably farm it out, but it's lost its continuity.
No, a Rube Goldberg scheme would involve a catapult launching re-charged batteries from sunlit bases to those in the dark. The catapult would be triggered by a candle burning though a string, while a large catcher's mitt would collect the batteries. And a hamster in an exercise wheel would be involved somehow. Getting all of this to work in a vacuum is left as an exercise for the reader.
Maybe I'm just missing the point, but could anyone explain to me exactly why a moon-colony or whatever would have to be made out of "lunar rock-like" materials?
Because making it out of "Mars rock-like" materials would limit you to the wrong color scheme.
How about building a line of bases running east and west until you circle the moon? Lay down some aluminum power lines, and run the bases that are in darkness from the ones that are in sunlight. And you don't have to do it at the equator. Do it a few degrees from one of the poles, and you won't need that many bases or power conduit.
This thread began with a link to a site listing reasons.
Maybe people are condescending to you for a reason?
When was the last time you walked into a retail store and a salesman handed you a brochure and said "I'll be at the register if you need me." Shouting RTFM is the second most ineffective way to sell your technology next to calling potential converts imbeciles.
If I walked into the "Revision Control Systems 'R Us" store, I'd prefer live interaction as well. There are, however, lots of on-line stores (ever hear of Amazon?) where the sales process doesn't use salepeople, just lists of reasons why a produce is desirable.
Most data is static. Once you receive an email or IM, it never changes, while blogs mostly just append new posts and comments. As a result, pointer-based snapshots only need to track data that has changed since the snapshot was taken. If you have one TB of new/updated data per year, then 12 monthly snapshots will need 6.5 TB, 12 weekly ones will need 1.5 TB, and 12 daily 0.214 TB, totalling 8.214 TB. That may seem like a lot, but if you're backing up 20 TB then it's 42% overhead. And since the snapshots will probably be kept on 7+1 RAID, it could be less than 24% of your raw storage.
I thought that I'd do a not-so-random search of site:journalspace.com plus various common words, like "web". From looking at the results, I've got to go with the story's tag, "andnothingofvaluewaslost".
You store the disk drive in an off-site server, and keep it powered up all the time, slowly re-reading all the data so you can spt drive failures in a hurry. The big boys do off-site replication for disaster recovery, and take snapshots of that for off-site backups. The media never has to leave the DR site, yet it's safe from almost anything that would take out the primary data center.
(OK, I'll concede that an eruption of the Yellowstone supervolcano could take out data centers thousands of miles apart, but if that happens you've got other things to worry about.)
You know, I've thought once or twice about building something that reads robots.txt files and archives everything that's denied to archive.org. Until today, I couldn't think of a good business case.
but the thumbnail of the article's picture, when scaled to width="50" height="37", looks an awful lot like an airliner flying towards an very skinny skyscraper.
I also want to know why the thumbnail wasn't scaled to the standard slashdot size to begin with.
Mass transit is neither cost-effective nor green. Per passenger-mile, it costs more to operate and generates more green-house gases than private automobiles. But don't believe me. Check out what the Federal Reserve Bank of St. Louis, The Atlantic, and others have to say about it.
Based solely on dollar cost, the annual light-rail subsidies could instead be used to buy an environmentally friendly hybrid Toyota Prius every five years for each poor rider and even to pay annual maintenance costs of $6,000. Increases in pollution would be minimal with the hybrid vehicle, and 7,700 new vehicles on the roadway would result in only a 0.5 percent increase in traffic congestion.3 And there would still be funds left overâ"about $49 million per year. These funds could be given to all other MetroLink riders (amounting to roughly $1,045 per person per year) and be used for cab fare, bus fare, etc.
The received wisdom on this topic is easily stated: 1. It is self-evident that public transportation is vastly more energy-efficient than automobiles; 2. It is self-evident that investing money to improve transit facilities will attract many more passengers. Therefore, the national energy policy ought to give major attention to building new transit systems and revitalizing old ones. Unfortunately, both of these "self-evident" premises turn out to be false.
Particularly disturbing were the numbers for some of the worst transit systems, including the light rail in San Jose, which I sometimes ride. That system takes twice as much energy per passenger than private cars do. It's not even the worst -- that's Cleveland, which also is part of a grid more dependent on fossil fuels than San Jose.
From a purely utilitarian point of view, it would be cheaper to simply buy compact cars for the poorest of the poor, or even subsidize some kind of taxi cab service for poor people. But that idea is too "way out there" - much stranger than ripping up our cities for years and years while the planners implement their expensive dreams.
Requiring that OS X only be used with Apple hardware is exactly like Ford requiring that the customer use only Ford-made replacement parts. Ford most emphatically isn't allowed to do that, so neither can Apple!
No, requiring that OS X only be used with Apple hardware is like Ford requiring that Ford-made replacement parts be used in a Ford. I don't think that anyone has ruled one way or the other on that question.
Iâ(TM)m intrigued that the article's illustration (presumably from the patent application) states that there are 31,351 capacitors, while the text of the article states that there are 31,353. Which is it? The first number is prime, which means that there isnâ(TM)t a simple rectangular grid containing all of the capacitors (although a staggered arrangement, such as the stars in the current US flag might be possible). 31,353, on the other hand, is 3*7*1493, however it also doesnâ(TM)t easily form a rectangular array. I have to wonder how either of these numbers came to be involved in the design of this a device.
I know someone who was in a car accident at age ten. He damaged one of the growth plates in his left leg, as a result he now has one leg a couple of inches shorter than the other. To date, there's been no good way to lengthen the short leg to the full length of the uninjured leg, and he isn't fond of the idea of shortening the longer one. Will this be able to help him?
The main depreciation of a used car isn't because newer and cooler cars come out (it only plays a small part), but actual mechanical wear.
Googling the phrase "drive off the lot" returned the factoid that "With a hybrid car, buyers do not always immediately lose that 30 percent of the value the minute they drive off the lot." People who aren't extolling the virtues of hybrids quote slightly lower numbers, some 10-15%, some 15-20%. Still, do you really think that there is that much mechanical wear associated with driving 500 feet?
Every car I've ever owned was purchased from the used-car department at a local dealer's. My current car was a year old and came with the remainder of the manufacturer's warranty. I saved a lot more than the percentage of its lifetime that had been "used up". My wife is a bit more paranoid, so her cars are purchased at year-end clearance sales. The savings aren't as large, but she gets a car with less than a hundred miles on the odometer.
A couple of days ago Slashdot had Twenty Years of Dijkstra's Cruelty. In it, he proposed using a non-executable language to force the students to use formal methods to prove the correctness, rather than trial-and-error testing. [...] But would Dijkstra's course be a better starting point?
Another way to show that LISP was neater than Turing machines was to write a universal LISP function [...] Writing eval required inventing a notation representing LISP functions as LISP data, [...] S.R. Russell noticed that eval could serve as an interpreter for LISP, promptly hand coded it, and we now had a programming language with an interpreter.
And that's the problem with Dijkstra's idea. As soon as the non-executable language is specified, some smart-alec's going to code an interpreter for it, and *poof* you're right back where you started.
When WotLK came out, I followed a link to a PA strip featuring the Lich King talking to a sycophant. The King was bored, and the sycophant was discussing ways to cheer him up, without success. Each idea had its own punchline, and one of the was killing high level players; the King replied that anyone who had plateaued already had no life and death would be a mercy.
Since then, I've been unable to find the strip. I thought it referred to the WotLK expansion, but may have been talking about the previous one. I've singled-stepped through several weeks worth of strips around the date of both releases, but haven't found anything. If anyone can help, I'll mod you up the next time i have points.
What language would you recommend for an OO programmer to start with?
Any language will work to start with, just commit to not using assignment statements. Twenty years ago, I was a FORTRAN programmer who got introduced to LISP. I decided to try rewriting some of my programs without assignments. Since FORTRAN didn't allow recursion, there were lot of places where I got stuck, but it was surprising how I was able to get done. And I seem to recall that the programs were easier to debug, since there were fewer places where data could get clobbered. It should be easier these days since modern languages support functional programing, you just need the self-discipline to stick to that style.
Hint: any place that you want to make an assignment, turn the rest of the block into a function and pass the value as a parameter. If you have a loop, try coding it as a reduce-type algorithm; it's mentally easier than turning it into a recursive algorithm and it maps well to massively parallel architectures. Python somehow got a lot of functional stuff added in when Guido wasn't paying close attention (http://www.artima.com/forums/flat.jsp?forum=106&thread=98196), so it's a great language to play with.
After that, follow the suggestions in the article. Learn Scala, F#, Erlang or Haskell. Get a copy of Mathematica. Have fun, and you'll learn a lot.
in cases where internet access is prohibitively expensive or even impossible, it makes perfect sense to have everything in easily-syncable git repositories.
Slashdot Mirror
I got a Wii Fit back in September and lost twenty pounds by New Year's Day. I was earning 30 Wii Fit points per day for most of that period, plus I was walking with my kids two miles every Saturday on a local nature trail.
Unfortunately, between the New Year's Eve party and a bunch of unhealthy left overs, I managed to put about a third of that back. Plus, bad weather has kept me off the nature trail, which certainly doesn't help. And all of that has hurt my motivation, so where before I'd done over 100 consecutive days of exercise, now I'm skipping more days than I'm doing. I've apparently fallen into the trap described in this article. Overall, though, I'm still doing better than most years, where I would put on ten pounds between Thanksgiving and New Year's, and never take it off. Damn you, homemade English Toffee!
BTW, Nintendo has created a pedometer that talks to your DS. The cartridge allows you to download your Mii for an experience similar to Wii Fit. What's missing is apparently any way to upload data into Wii Fit, but that may show up in a future channel.
If they'd made it possible to reflash, a zillion Linux weenies would have bought the devices just to put Rockbox on them.
I hadn't realized that you couldn't reflash a Zune. That's going to make it pretty hard to patch that leap-year bug, don't you think?
Speaking of which, I wonder what their sales figures looked like for the first three weeks of January. After this weekend, sales will definitely tank, but I'm sure somebody at the Harvard Business Review would like to know how many nails that bug put into the coffin.
I graduated from college in 1978 and moved to Silicon Valley. In a short time, I owned an Apple II, an Advent VideoBeam projection TV, and a copy of Bruce Artwick's original Flight Simulator. With a nominal 64" screen, it seemed as immersive as any commercial flight simulator of that time (even if it was low-res black-and-white wireframe). And when you bought it, you got a second floppy with the source code! I'm saddened to see the franchise end after 30+ years. Yes, I know that they'll probably farm it out, but it's lost its continuity.
Hey, you kids, get off my tarmac!
No, a Rube Goldberg scheme would involve a catapult launching re-charged batteries from sunlit bases to those in the dark. The catapult would be triggered by a candle burning though a string, while a large catcher's mitt would collect the batteries. And a hamster in an exercise wheel would be involved somehow. Getting all of this to work in a vacuum is left as an exercise for the reader.
Maybe I'm just missing the point, but could anyone explain to me exactly why a moon-colony or whatever would have to be made out of "lunar rock-like" materials?
Because making it out of "Mars rock-like" materials would limit you to the wrong color scheme.
How about building a line of bases running east and west until you circle the moon? Lay down some aluminum power lines, and run the bases that are in darkness from the ones that are in sunlight. And you don't have to do it at the equator. Do it a few degrees from one of the poles, and you won't need that many bases or power conduit.
This thread began with a link to a site listing reasons.
Maybe people are condescending to you for a reason?
When was the last time you walked into a retail store and a salesman handed you a brochure and said "I'll be at the register if you need me." Shouting RTFM is the second most ineffective way to sell your technology next to calling potential converts imbeciles.
If I walked into the "Revision Control Systems 'R Us" store, I'd prefer live interaction as well. There are, however, lots of on-line stores (ever hear of Amazon?) where the sales process doesn't use salepeople, just lists of reasons why a produce is desirable.
Most data is static. Once you receive an email or IM, it never changes, while blogs mostly just append new posts and comments. As a result, pointer-based snapshots only need to track data that has changed since the snapshot was taken. If you have one TB of new/updated data per year, then 12 monthly snapshots will need 6.5 TB, 12 weekly ones will need 1.5 TB, and 12 daily 0.214 TB, totalling 8.214 TB. That may seem like a lot, but if you're backing up 20 TB then it's 42% overhead. And since the snapshots will probably be kept on 7+1 RAID, it could be less than 24% of your raw storage.
I thought that I'd do a not-so-random search of site:journalspace.com plus various common words, like "web". From looking at the results, I've got to go with the story's tag, "andnothingofvaluewaslost".
You store the disk drive in an off-site server, and keep it powered up all the time, slowly re-reading all the data so you can spt drive failures in a hurry. The big boys do off-site replication for disaster recovery, and take snapshots of that for off-site backups. The media never has to leave the DR site, yet it's safe from almost anything that would take out the primary data center.
(OK, I'll concede that an eruption of the Yellowstone supervolcano could take out data centers thousands of miles apart, but if that happens you've got other things to worry about.)
Sort of reminds me of this.
You know, I've thought once or twice about building something that reads robots.txt files and archives everything that's denied to archive.org. Until today, I couldn't think of a good business case.
but the thumbnail of the article's picture, when scaled to width="50" height="37", looks an awful lot like an airliner flying towards an very skinny skyscraper.
I also want to know why the thumbnail wasn't scaled to the standard slashdot size to begin with.
Mass transit is neither cost-effective nor green. Per passenger-mile, it costs more to operate and generates more green-house gases than private automobiles. But don't believe me. Check out what the Federal Reserve Bank of St. Louis, The Atlantic, and others have to say about it.
http://www.stlouisfed.org/publications/re/2004/c/pages/light_rail.html
Based solely on dollar cost, the annual light-rail subsidies could instead be used to buy an environmentally friendly hybrid Toyota Prius every five years for each poor rider and even to pay annual maintenance costs of $6,000. Increases in pollution would be minimal with the hybrid vehicle, and 7,700 new vehicles on the roadway would result in only a 0.5 percent increase in traffic congestion.3 And there would still be funds left overâ"about $49 million per year. These funds could be given to all other MetroLink riders (amounting to roughly $1,045 per person per year) and be used for cab fare, bus fare, etc.
http://www.theatlantic.com/doc/197910/197910
The received wisdom on this topic is easily stated: 1. It is self-evident that public transportation is vastly more energy-efficient than automobiles; 2. It is self-evident that investing money to improve transit facilities will attract many more passengers. Therefore, the national energy policy ought to give major attention to building new transit systems and revitalizing old ones. Unfortunately, both of these "self-evident" premises turn out to be false.
http://www.templetons.com/brad/transit-myth.html
Particularly disturbing were the numbers for some of the worst transit systems, including the light rail in San Jose, which I sometimes ride. That system takes twice as much energy per passenger than private cars do. It's not even the worst -- that's Cleveland, which also is part of a grid more dependent on fossil fuels than San Jose.
http://www.gregburch.net/cars/plans.html
From a purely utilitarian point of view, it would be cheaper to simply buy compact cars for the poorest of the poor, or even subsidize some kind of taxi cab service for poor people. But that idea is too "way out there" - much stranger than ripping up our cities for years and years while the planners implement their expensive dreams.
Requiring that OS X only be used with Apple hardware is exactly like Ford requiring that the customer use only Ford-made replacement parts. Ford most emphatically isn't allowed to do that, so neither can Apple!
No, requiring that OS X only be used with Apple hardware is like Ford requiring that Ford-made replacement parts be used in a Ford. I don't think that anyone has ruled one way or the other on that question.
Iâ(TM)m intrigued that the article's illustration (presumably from the patent application) states that there are 31,351 capacitors, while the text of the article states that there are 31,353. Which is it? The first number is prime, which means that there isnâ(TM)t a simple rectangular grid containing all of the capacitors (although a staggered arrangement, such as the stars in the current US flag might be possible). 31,353, on the other hand, is 3*7*1493, however it also doesnâ(TM)t easily form a rectangular array. I have to wonder how either of these numbers came to be involved in the design of this a device.
Hmmm, five category icons for one story. Is that a record?
I know someone who was in a car accident at age ten. He damaged one of the growth plates in his left leg, as a result he now has one leg a couple of inches shorter than the other. To date, there's been no good way to lengthen the short leg to the full length of the uninjured leg, and he isn't fond of the idea of shortening the longer one. Will this be able to help him?
So *that's* where little electrodes come from!
The main depreciation of a used car isn't because newer and cooler cars come out (it only plays a small part), but actual mechanical wear.
Googling the phrase "drive off the lot" returned the factoid that "With a hybrid car, buyers do not always immediately lose that 30 percent of the value the minute they drive off the lot." People who aren't extolling the virtues of hybrids quote slightly lower numbers, some 10-15%, some 15-20%. Still, do you really think that there is that much mechanical wear associated with driving 500 feet?
Every car I've ever owned was purchased from the used-car department at a local dealer's. My current car was a year old and came with the remainder of the manufacturer's warranty. I saved a lot more than the percentage of its lifetime that had been "used up". My wife is a bit more paranoid, so her cars are purchased at year-end clearance sales. The savings aren't as large, but she gets a car with less than a hundred miles on the odometer.
A couple of days ago Slashdot had Twenty Years of Dijkstra's Cruelty. In it, he proposed using a non-executable language to force the students to use formal methods to prove the correctness, rather than trial-and-error testing. [...] But would Dijkstra's course be a better starting point?
Do you know how LISP was created? http://www-formal.stanford.edu/jmc/history/lisp/node3.html#SECTION00030000000000000000
Another way to show that LISP was neater than Turing machines was to write a universal LISP function [...] Writing eval required inventing a notation representing LISP functions as LISP data, [...] S.R. Russell noticed that eval could serve as an interpreter for LISP, promptly hand coded it, and we now had a programming language with an interpreter.
And that's the problem with Dijkstra's idea. As soon as the non-executable language is specified, some smart-alec's going to code an interpreter for it, and *poof* you're right back where you started.
Excellent! Someone needs to tell Gabe to quit plagiarizing Tim's style. :-)
When WotLK came out, I followed a link to a PA strip featuring the Lich King talking to a sycophant. The King was bored, and the sycophant was discussing ways to cheer him up, without success. Each idea had its own punchline, and one of the was killing high level players; the King replied that anyone who had plateaued already had no life and death would be a mercy.
Since then, I've been unable to find the strip. I thought it referred to the WotLK expansion, but may have been talking about the previous one. I've singled-stepped through several weeks worth of strips around the date of both releases, but haven't found anything. If anyone can help, I'll mod you up the next time i have points.
Thanks!
What language would you recommend for an OO programmer to start with?
Any language will work to start with, just commit to not using assignment statements. Twenty years ago, I was a FORTRAN programmer who got introduced to LISP. I decided to try rewriting some of my programs without assignments. Since FORTRAN didn't allow recursion, there were lot of places where I got stuck, but it was surprising how I was able to get done. And I seem to recall that the programs were easier to debug, since there were fewer places where data could get clobbered. It should be easier these days since modern languages support functional programing, you just need the self-discipline to stick to that style.
Hint: any place that you want to make an assignment, turn the rest of the block into a function and pass the value as a parameter. If you have a loop, try coding it as a reduce-type algorithm; it's mentally easier than turning it into a recursive algorithm and it maps well to massively parallel architectures. Python somehow got a lot of functional stuff added in when Guido wasn't paying close attention (http://www.artima.com/forums/flat.jsp?forum=106&thread=98196), so it's a great language to play with.
After that, follow the suggestions in the article. Learn Scala, F#, Erlang or Haskell. Get a copy of Mathematica. Have fun, and you'll learn a lot.
You're used to permanent online Internet access.
in cases where internet access is prohibitively expensive or even impossible, it makes perfect sense to have everything in easily-syncable git repositories.
You missed a few scenarios.