Yes a lot of tasks today are essentially serial. Way back in the 32K days, we had to partition our work so it would fit in a 24k. The OS took 8k the memory hogging pos.... Any rate, our tasks were broken down into little pieces of code that loaded serially one after the other. It was hard to imagine what one would do if the had a machine with 256K of memory. Nobody could ever use it! Reading your post reminded me of those days - some of us were hampered by a lack of imagination.
Some tasks are serial, others can be parallelized. You don't need fancy languages to do it either. To effectively partition tasks into threads using something as archaic as C, you can either fork or you can load different processes. Either way works. The trick is to shift one's thinking from "tasks are serial..." to "how could I speed my code up if I had multiple cpus available?"
Encoding music to put on some sort of music player that doesn't have replaceable batteries and is headed for landfill 18 months after it's purchased comes to mind. Partitioning the music can be done either by tunes or within a tune if the encoding scheme can be chunked.
AI is scalable via threading which mean a well laid out game architecture could scale with more hardware threads. A user with a single processor would only get a few smart enemies, a user with a cpu array could see lots of smart behavior such as some of the enemy deciding to flee while other warriors come charging into battle.
Folks who play with Photoshop or Gimp can easily soak their cpus. A blur operation is the kind of task that can be partioned to good effect. It's not programming languages that's keeping this from happening as much as not many people have the requisite hardware.
Personally, when I'm working, I'll be printing, scanning and reading the scanned input simultaneously. For some unknown reason, the printer driver soaks up all the cpu cycles which slows down my reader and scanner software. Being able to allocate the printer driver its own hardware would make the rest of my workflow smoother.
Some of us will be able to use the horsepower and some of us won't. Not much has changed in the past 40 years.
Yesterday's NY Times had an article reporting that IBM said Apple left because of pricing issues and Apple saying they left because of technology issues. Deeper in the article, there's a reference to IBM saying that Apple would have to kick in some cash if they wanted IBM to pour more resources into developing the PPC the way Apple needed it to go. It looks like both Apple and IBM are telling the truth - it was about both price and performance.
With IBM looking at the hundreds of millions of units going to the console market vs the few million Apple would sell, it's easy to see IBM's point of view on this.
Perhaps. Are you sure their model doesn't embed the data they're modeling? Did their model forecast the climate for 2004 in 1995 or did they have to wait until 2003 to do that?
I'm not trying to be argumentative - it's just I've seen so many models fall apart that when I read about Lorenz's work I came to understand why they'd fallen apart. What made me even more skeptical was my first job out of college I worked for an econometric model building firm. I saw things being done that we'd been specifically been taught not to do. The model builder was doing it because the legit model didn't cut the mustard and we had to have a deliverable. When a climatologist claims to have a model that works, I think of Lorenz's findings and my experience. I've seen first hand how models can be made to look much better than they actually are.
Failing that, how about starting the patent challenge clock when the patent holder issues a claim instead of starting the clock when the patent is issued. It'd deal with submarine patents as well as eliminate the need to look over the patent office's shoulder on each and every patent they issue.
Go take a look at the ERA-40 climate model, which does a hell of a lot better than just predicting cold winters and warm summers, and tell me that our models are utterly useless.
I may not have made myself clear. It wasn't that Lorenz's model failed to predict the climate accurately, it was that his model failed to predict itself accurately. It wasn't a matter of choosing the wrong equations or data - it's a matter of self-referential, non-linear equations are inherently unpredictable. Bigger computers and bigger models don't get away from that problem.
I quote from the ERA-40 site: After 10 days of coupled integrations, the model drift begins to be significant.
What they are saying is that the most reasonable explanation of what we are observing is that the warming is anthropogenic and that other attempts at explaining the current patterns are less plausible.
Hmmm. Here we have the 9 prior warmings, around 100,000 years apart which are obviously not anthropogenic. The last warming was about 100,000 years ago and to my eye, it seems more plausible that we're on a "seasonal" cycle than mankind has caused the current uptick in temperatures. Hell, this warming started over 17,000 years ago. Just exactly how is that mankind's fault?
I'm sorry, but you just don't understand what that means. What that means is that you can't predict whether it's going to rain on a specific day, say, 11291 days from now. It does not mean that you can't predict any kind of long term quantities.
Lorenz had a simple climate model based on 3 non-linear equations. He had perfect data and perfect knowledge of his little universe. Nonetheless, his model was predicting hot winters and cold summers within a few years. Using your Las Vegas analogy, it was as if he ended up owning the casino. His point wasn't that his model was wrong. His point was that non-linear models are very sensitive to their input values and that they behave in unpredictable fashions very quickly.
None of todays climate models come even close to having the perfect knowledge or perfect data Lorenz had and yet they're making claims as to what the climate will be 50 years out. It's bullshit and you know it.
There's a strong correlation between atmospheric CO2 and warming and well understood atmospheric interactions of CO2, but some try to point the finger elsewhere or back to natural patterns.
Before you assert categorically that global warming is anthropogenic, you have to explain why the data that shows it's a natural phenomenon do not apply. There are ample oxygen isotope data that indicate the interglacials have had a 100,000 year cycle for at least the past million years. We weren't around in any significant numbers for the last 9 interglacials that have preceeded this one. Why is this interglacial anthropogenic when it's on the same cycle as the previous 9 interglacials?
What's truly astounding is the massively increasing level of outright propaganda on the subject. The scientists appear to be being left behind and the propagandists (sponsored by private industry) are taking over the show.
Propaganda isn't the sole province of partisan politicians. Arthur Eddington was convinced Chandra's theory that black holes could exist was wrong and he browbeat anyone who disagreed with him. It wasn't until Eddington died in 1944 that any progress on black holes was possible. There's a story about Shapely erasing data that disproved his hypothesis that the Milky Way was the whole universe. At the time, Shapely was the director of Harvard's observatory. The point is, just because some scientist believe something to be true doesn't necessarily mean it is - no matter how reputable the scientist is.
You might counter "isn't it better to act than to wait until we're sure?" The answer is "it depends on the cost of acting and being right vs. the cost of acting and being wrong." Moreover, you have to know what to do if you choose to act. Don't look to climate models for guidance - they're not worth much. The salient quote:
Lorenz showed that with a set of simple differential equations seemingly very complex turbulent behaviour could be created that would previously have been considered as random. He further showed that accurate longer range forecasts in any chaotic system were impossible, thereby overturning the previous orthodoxy. It had been believed that the more equations you add to describe a system, the more accurate will be the eventual forecast.
Plop a windows box and a Mac in front of a newbie and see which one is "0wned" first.
Your experiment reminds me of my friend who is a Christian Scientist. She doesn't believe in germs so refuses to get her children innoculated against the various diseases that afflict kids. Not so surprisingly, her kids are fine. She views the fact that they're healthy as proof that she's right. She doesn't understand that her kids survive because the overwhelming majority of kids surrounding hers are disease free. If there were more Christian Scientist out there, her strategy would fail.
Macs are relatively virus free because the majority of virii out there aren't aimed at Macs. If the Mac ever regained a significant market share, virus writers would start aiming at the platform and your experiment would show different results.
Hawaii went whole hog for OTEC back in the late 70's. They sunk a pipe and hauled cold water up to use it as a heatsink for a steam engine driven by alcohol. The warm surface water would heat the alcohol to boiling, they'd run the vapor through a turbine and then use the cold water to condense the alcohol vapor.
Long story short, it didn't work very well. My physics prof pointed out that the theoretical limit on their technology was
(303-273)/303
or about 10% where 303 is the boiling temperature in Kelvin and 273 the cold water temperature in Kelvin. After subtracting the various inefficiencies, there wasn't enough power left over to do anything with.
All was not lost however, the Hawaiians ended up using the cold, nutrient rich water to feed aqua culture enterprises that would use it to grow lobster, abalone, kelp and nori (the seaweed you wrap sushi in.) Aqua culture was so successful that the farmers started sinking their own pipes because the state couldn't meet the demand for cold water.
I couldn't find any details on the prize, i.e., what the design parameters are.
The lunar environment is so radically different there that it changes a lot of the design parameters. Structures weigh 1/6 as much there as they do here. Build a structure that is strong enough for Lunar gravity and it'll collapse here. You've got both a 250 degree F heat source and -250 degrees heatsink readily available on the Moon which makes for a nice heat engine but again, it only works on the Moon.
There are other significant differences so I'm curious whether NASA plans on testing the machines using Earth design rules or Moon design rules.
If that were true, the Soviet Union would not have been able to fly over U.S. territory and vice-versa. It was a deliberate choice Eisenhower made in 1955 when he proposed his "Open Skies" initiative. When Sputnik flew a few years later, he didn't complain about its flying over US territory because he wanted the right to do the same thing. In 1960 when Corona flew, it made a hash of the fear that the Soviets had an advantage over us and enabled Eisenhower to focus on domestic issues instead of meeting a non-existant military threat.
I started up with Walmart when they got going so it's possible the West Coast crew did their job better or they were closer to a major mail hub than the Arkansas site.
As to going back to NF, I don't think so - leastwise not today. Perhaps I'll check out greencine. My son swears by it.
It's ironic that Walmart pointed me back at Netflix. Netflix pulled a fast one back when Walmart was just getting going and switched their terms of service while claiming that the terms hadn't changed. They capped the two-dvd service at 4 dvds/month and introduced an uncapped 2 dvd service simultanously. Instead of pitching the capped service as a way to avoid a price rise, they chose to claim the old cap had always been there. Since it hadn't been, I felt Netflix was being dishonest and left Netflix for Walmart. Now Walmart points me back to NF.
You're the Netflix manager in charge of the transition from Netflix's side. Do you
A - Let Walmart assign the coding task to one of their programmers who is about to get laid off or
B - Do you claim that task as one of yours? Remember that Netflix is allegedly getting 100K customers out of this transaction. Lose 5% of them because of a coding screwup and you've cost Netflix $75,000 per month.
If you chose A, then you shouldn't be working for Netflix.
If you chose B and you still screwed it up,...did you know Wendy's is hiring?
WalMart shipped from Arkansas, so you can imagine how long it took for disks to travel back and forth to the rest of the country. It was usually a 2-3 day trip each way here to Dallas, Texas, for example. And the selection was even worse than Blockbuster's online service, if you can believe it.
Perhaps for Texas clients they shipped out of Arkansas but here in Northern California, they were shipping out of the SF Bay Area. My turnaround was frequently 2 days after I mailed in a DVD.
As to availability, my choices were 50-50 labeled "Available Now" and "Very Long Wait." I'm not sure what distinguished "Very Long Wait" because those selections came as quickly as the "Available Now" selections. In one case, Deadwood, the labeling transitioned to "Available Now" as I went through the first season. I'm guessing that they were building their inventory depending on which "Very Long Wait" titles were getting ordered.
For me Walmart was fine. It had a decent enough selection that I could keep my queue active. Clearly, we had different experiences.
I got an email from Walmart this morning telling me about the change over. The email had a link that took me to a page on Walmart's site explaining they were shutting down. On the page was a button that said "Click here to get your account transfered to Netflix" along with some text that explained my rental queue and terms of service would be transferred intact by clicking on the button.
Didn't work that way. Got transferred to Netflix's new account page with the wrong service option selected and everything else blank. Had the software worked properly, I would now be a Netflix customer but it didn't so I checked out Blockbuster's offer. So if anyone at Netflix is reading this, especially if you're the developer, your bug cost you a customer. Blockbuster looks to have a better deal than Netflix - something I probably wouldn't have noticed had your code worked properly.
It's a shame Walmart quit. When they entered the market, Netflix raised their monthly service fee, realized that wasn't the brightest thing to way to compete with the world's largest retailer and subsequently lowered their fees to match Walmart's. Both companies then proceeded to add a whole lot more titles to their mix - or at least claim they did. The overall effect of Walmart's entry into this field was to increase competition and improve the offerings. I frequently saw a 2 day turnaround from mailing a DVD to receiving its replacement. With Walmart's exit, I wouldn't be surprised to see a concurrent dimunition of service from the remaining players.
The motivation behind a transition would be to go to a significantly faster cpu. X-86 has been, and always will be, crippled by backward compatibility requirements. Toss that requirement and you can do a lot more with today's fabrication technologies. The market however, won't let you toss its investment in software so you resort to emulation.
It's not as it has never been done before. Apple did it when they transitioned from the 68xxx cpu to the PowerPc. There's no reason Microsoft couldn't pull the same stunt. The payoff would be
More powerful hardware enables you to add more features to the OS and hence engender another sales cycle.
Microsoft ends up owning the underlying hardware design as well.
In case you missed it, Microsoft owns the cpu design in the new Xbox. You can bet Microsoft would love to not only put their desktop on every cpu, they'd love to put their cpu on every desktop as well.
The motivation could run deeper - Microsoft may jetison x-86 altogether and bring Windows onto a proprietary platform. They'd need x-86 backward compatibility to get the market to follow. Xbox may be just the first step.
Consider the Trends in International Mathematics and Science Study eighth-grade science test, for instance, and the scores achieved by Colorado, Connecticut, Iowa, Maine, Massachusetts, Minnesota, Montana, Nebraska North Dakota, Oregon, Utah, Vermont, Wisconsin and Wyoming. Had these states--none of which has a substantial underclass--been treated as separate nations, each of them would have been outscored only by Singapore. (China, the nation that produced the ACM contest winner this year, has refused to participate in TIMMS.)
I couldn't find the TMSS data broken out by state that he refers to.
Minnesota's presence is notable because the original studies in 86 and 92 that compared Minnesotan children against Taiwanese and Japanese students found huge educational gaps. Moreover, the Minnesotan children chosen for the 1986 and 1992 followup were chosen from an upper-income Minnesota neighborhood. The schools were chosen because they were considered among the best in the United States.
I'd be very interested in seeing the data the author is referring to.
The article the grandparent references states that the plane was dismantled in Hawaii. Looking at the plane on the Mojave tarmac from about 1/2 a mile away, the plane looked intact - wings, fuselage etc. Perhaps someone who has seen the plane closer than that could comment.
What's left of the Aloha Airlines plane is stored at Mojave Airport where Scaled Composites is sited. It's a boneyard for planes and to pass the time while we waited for the second SpaceShipOne flight, we were scoping out the various planes that were parked across the runway. It was tucked away next to a 747 so you either had to be very sharp eyed or had to look for it to spot it. The entire plane was there, it wasn't chopped into chunks.
Now the question is, who had the cajones to fly the plane from Hawaii to Mojave? Safely landing the plane after the roof peeled off is one thing. It's quite another to get into an airplane whose fuselage has partially peeled away and fly it several thousand miles. Especially if most of the flight is over water. That took big brass ones in my book.
I thought it was a pretty good idea and installed the tool the day it became available. I didn't mind it accessing the web while I was browsing but when I turned off my browser, the tool kept going to the net for lord knows what. I wouldn't have noticed except ZoneAlarm kept flashing despite the fact I wasn't explicitly running anything. Turned off the tool and the flashing stopped.
I figured that saving a few seconds browsing wasn't worth relinquishing one of the few tools I have to detect unusual net traffic. The tool lasted perhaps 2 hours on my machine before I removed it.
You might want to watch this talk Bob Colwell gave at Stanford. Colwell was one of the Pentium architects. It's clear from the talk - that Itanium was having problems early on and - it wasn't an internal secret and - there were groups within Intel that were pushing for other architectures and - legacy support can bog alternative projects down and - Blue Crystals (marketing) can drive a company into making the wrong technological choices.
It's an interesting video, well worth watching for the insight Colwell displayed.
But I still wonder what this has that water does not?:-) Cooling wise, I mean.
Galinstan conducts heat far better than water. Galinstan's conductivity is 16.5 W/(MK) vs. water's value of between.4 to.7.
Several posters have suggested gallium which melts at 40 C. Using pure gallium would be a disaster because when it freezes, it expands like water does. It'd rupture the plumbing inside your computer. Galinstan stays liquid down to -19 C. A spec sheet is available.
Galinstan has a couple of drawbacks. A, it's corrosive and B, it sticks to most surfaces unless the surface is properly prepared. That means your radiator, water block and all the tubing has to be internally coated before you pour Galinstan into your cooling rig.
Slashdot Mirror
Some tasks are serial, others can be parallelized. You don't need fancy languages to do it either. To effectively partition tasks into threads using something as archaic as C, you can either fork or you can load different processes. Either way works. The trick is to shift one's thinking from "tasks are serial..." to "how could I speed my code up if I had multiple cpus available?"
Encoding music to put on some sort of music player that doesn't have replaceable batteries and is headed for landfill 18 months after it's purchased comes to mind. Partitioning the music can be done either by tunes or within a tune if the encoding scheme can be chunked.
AI is scalable via threading which mean a well laid out game architecture could scale with more hardware threads. A user with a single processor would only get a few smart enemies, a user with a cpu array could see lots of smart behavior such as some of the enemy deciding to flee while other warriors come charging into battle.
Folks who play with Photoshop or Gimp can easily soak their cpus. A blur operation is the kind of task that can be partioned to good effect. It's not programming languages that's keeping this from happening as much as not many people have the requisite hardware.
Personally, when I'm working, I'll be printing, scanning and reading the scanned input simultaneously. For some unknown reason, the printer driver soaks up all the cpu cycles which slows down my reader and scanner software. Being able to allocate the printer driver its own hardware would make the rest of my workflow smoother.
Some of us will be able to use the horsepower and some of us won't. Not much has changed in the past 40 years.
With IBM looking at the hundreds of millions of units going to the console market vs the few million Apple would sell, it's easy to see IBM's point of view on this.
I'm not trying to be argumentative - it's just I've seen so many models fall apart that when I read about Lorenz's work I came to understand why they'd fallen apart. What made me even more skeptical was my first job out of college I worked for an econometric model building firm. I saw things being done that we'd been specifically been taught not to do. The model builder was doing it because the legit model didn't cut the mustard and we had to have a deliverable. When a climatologist claims to have a model that works, I think of Lorenz's findings and my experience. I've seen first hand how models can be made to look much better than they actually are.
Failing that, how about starting the patent challenge clock when the patent holder issues a claim instead of starting the clock when the patent is issued. It'd deal with submarine patents as well as eliminate the need to look over the patent office's shoulder on each and every patent they issue.
I may not have made myself clear. It wasn't that Lorenz's model failed to predict the climate accurately, it was that his model failed to predict itself accurately. It wasn't a matter of choosing the wrong equations or data - it's a matter of self-referential, non-linear equations are inherently unpredictable. Bigger computers and bigger models don't get away from that problem.
I quote from the ERA-40 site: After 10 days of coupled integrations, the model drift begins to be significant.
It's exactly the same problem Lorenz had.
Hmmm. Here we have the 9 prior warmings, around 100,000 years apart which are obviously not anthropogenic. The last warming was about 100,000 years ago and to my eye, it seems more plausible that we're on a "seasonal" cycle than mankind has caused the current uptick in temperatures. Hell, this warming started over 17,000 years ago. Just exactly how is that mankind's fault?
I'm sorry, but you just don't understand what that means. What that means is that you can't predict whether it's going to rain on a specific day, say, 11291 days from now. It does not mean that you can't predict any kind of long term quantities.
Lorenz had a simple climate model based on 3 non-linear equations. He had perfect data and perfect knowledge of his little universe. Nonetheless, his model was predicting hot winters and cold summers within a few years. Using your Las Vegas analogy, it was as if he ended up owning the casino. His point wasn't that his model was wrong. His point was that non-linear models are very sensitive to their input values and that they behave in unpredictable fashions very quickly.
None of todays climate models come even close to having the perfect knowledge or perfect data Lorenz had and yet they're making claims as to what the climate will be 50 years out. It's bullshit and you know it.
Before you assert categorically that global warming is anthropogenic, you have to explain why the data that shows it's a natural phenomenon do not apply. There are ample oxygen isotope data that indicate the interglacials have had a 100,000 year cycle for at least the past million years. We weren't around in any significant numbers for the last 9 interglacials that have preceeded this one. Why is this interglacial anthropogenic when it's on the same cycle as the previous 9 interglacials?
What's truly astounding is the massively increasing level of outright propaganda on the subject. The scientists appear to be being left behind and the propagandists (sponsored by private industry) are taking over the show.
Propaganda isn't the sole province of partisan politicians. Arthur Eddington was convinced Chandra's theory that black holes could exist was wrong and he browbeat anyone who disagreed with him. It wasn't until Eddington died in 1944 that any progress on black holes was possible. There's a story about Shapely erasing data that disproved his hypothesis that the Milky Way was the whole universe. At the time, Shapely was the director of Harvard's observatory. The point is, just because some scientist believe something to be true doesn't necessarily mean it is - no matter how reputable the scientist is.
You might counter "isn't it better to act than to wait until we're sure?" The answer is "it depends on the cost of acting and being right vs. the cost of acting and being wrong." Moreover, you have to know what to do if you choose to act. Don't look to climate models for guidance - they're not worth much. The salient quote:
- Have a video that documents the iPod was designed as a disposable item.
- Subsequently reward Apple with record sales and then
- Sue Apple for selling disposable iPods.
Nonetheless, Apple brought this suit upon themselves by refusing to design the iPod so the end user could easily replace the battery.Your experiment reminds me of my friend who is a Christian Scientist. She doesn't believe in germs so refuses to get her children innoculated against the various diseases that afflict kids. Not so surprisingly, her kids are fine. She views the fact that they're healthy as proof that she's right. She doesn't understand that her kids survive because the overwhelming majority of kids surrounding hers are disease free. If there were more Christian Scientist out there, her strategy would fail.
Macs are relatively virus free because the majority of virii out there aren't aimed at Macs. If the Mac ever regained a significant market share, virus writers would start aiming at the platform and your experiment would show different results.
Long story short, it didn't work very well. My physics prof pointed out that the theoretical limit on their technology was
or about 10% where 303 is the boiling temperature in Kelvin and 273 the cold water temperature in Kelvin. After subtracting the various inefficiencies, there wasn't enough power left over to do anything with.All was not lost however, the Hawaiians ended up using the cold, nutrient rich water to feed aqua culture enterprises that would use it to grow lobster, abalone, kelp and nori (the seaweed you wrap sushi in.) Aqua culture was so successful that the farmers started sinking their own pipes because the state couldn't meet the demand for cold water.
The lunar environment is so radically different there that it changes a lot of the design parameters. Structures weigh 1/6 as much there as they do here. Build a structure that is strong enough for Lunar gravity and it'll collapse here. You've got both a 250 degree F heat source and -250 degrees heatsink readily available on the Moon which makes for a nice heat engine but again, it only works on the Moon.
There are other significant differences so I'm curious whether NASA plans on testing the machines using Earth design rules or Moon design rules.
If that were true, the Soviet Union would not have been able to fly over U.S. territory and vice-versa. It was a deliberate choice Eisenhower made in 1955 when he proposed his "Open Skies" initiative. When Sputnik flew a few years later, he didn't complain about its flying over US territory because he wanted the right to do the same thing. In 1960 when Corona flew, it made a hash of the fear that the Soviets had an advantage over us and enabled Eisenhower to focus on domestic issues instead of meeting a non-existant military threat.
Outer space is open to whomever can get there.
As to going back to NF, I don't think so - leastwise not today. Perhaps I'll check out greencine. My son swears by it.
It's ironic that Walmart pointed me back at Netflix. Netflix pulled a fast one back when Walmart was just getting going and switched their terms of service while claiming that the terms hadn't changed. They capped the two-dvd service at 4 dvds/month and introduced an uncapped 2 dvd service simultanously. Instead of pitching the capped service as a way to avoid a price rise, they chose to claim the old cap had always been there. Since it hadn't been, I felt Netflix was being dishonest and left Netflix for Walmart. Now Walmart points me back to NF.
- A - Let Walmart assign the coding task to one of their programmers who is about to get laid off or
- B - Do you claim that task as one of yours? Remember that Netflix is allegedly getting 100K customers out of this transaction. Lose 5% of them because of a coding screwup and you've cost Netflix $75,000 per month.
If you chose A, then you shouldn't be working for Netflix.If you chose B and you still screwed it up,...did you know Wendy's is hiring?
Perhaps for Texas clients they shipped out of Arkansas but here in Northern California, they were shipping out of the SF Bay Area. My turnaround was frequently 2 days after I mailed in a DVD.
As to availability, my choices were 50-50 labeled "Available Now" and "Very Long Wait." I'm not sure what distinguished "Very Long Wait" because those selections came as quickly as the "Available Now" selections. In one case, Deadwood, the labeling transitioned to "Available Now" as I went through the first season. I'm guessing that they were building their inventory depending on which "Very Long Wait" titles were getting ordered.
For me Walmart was fine. It had a decent enough selection that I could keep my queue active. Clearly, we had different experiences.
Didn't work that way. Got transferred to Netflix's new account page with the wrong service option selected and everything else blank. Had the software worked properly, I would now be a Netflix customer but it didn't so I checked out Blockbuster's offer. So if anyone at Netflix is reading this, especially if you're the developer, your bug cost you a customer. Blockbuster looks to have a better deal than Netflix - something I probably wouldn't have noticed had your code worked properly.
It's a shame Walmart quit. When they entered the market, Netflix raised their monthly service fee, realized that wasn't the brightest thing to way to compete with the world's largest retailer and subsequently lowered their fees to match Walmart's. Both companies then proceeded to add a whole lot more titles to their mix - or at least claim they did. The overall effect of Walmart's entry into this field was to increase competition and improve the offerings. I frequently saw a 2 day turnaround from mailing a DVD to receiving its replacement. With Walmart's exit, I wouldn't be surprised to see a concurrent dimunition of service from the remaining players.
It's not as it has never been done before. Apple did it when they transitioned from the 68xxx cpu to the PowerPc. There's no reason Microsoft couldn't pull the same stunt. The payoff would be
- More powerful hardware enables you to add more features to the OS and hence engender another sales cycle.
- Microsoft ends up owning the underlying hardware design as well.
In case you missed it, Microsoft owns the cpu design in the new Xbox. You can bet Microsoft would love to not only put their desktop on every cpu, they'd love to put their cpu on every desktop as well.The motivation could run deeper - Microsoft may jetison x-86 altogether and bring Windows onto a proprietary platform. They'd need x-86 backward compatibility to get the market to follow. Xbox may be just the first step.
I couldn't find the TMSS data broken out by state that he refers to.
Minnesota's presence is notable because the original studies in 86 and 92 that compared Minnesotan children against Taiwanese and Japanese students found huge educational gaps. Moreover, the Minnesotan children chosen for the 1986 and 1992 followup were chosen from an upper-income Minnesota neighborhood. The schools were chosen because they were considered among the best in the United States.
I'd be very interested in seeing the data the author is referring to.
The article the grandparent references states that the plane was dismantled in Hawaii. Looking at the plane on the Mojave tarmac from about 1/2 a mile away, the plane looked intact - wings, fuselage etc. Perhaps someone who has seen the plane closer than that could comment.
Now the question is, who had the cajones to fly the plane from Hawaii to Mojave? Safely landing the plane after the roof peeled off is one thing. It's quite another to get into an airplane whose fuselage has partially peeled away and fly it several thousand miles. Especially if most of the flight is over water. That took big brass ones in my book.
I figured that saving a few seconds browsing wasn't worth relinquishing one of the few tools I have to detect unusual net traffic. The tool lasted perhaps 2 hours on my machine before I removed it.
- that Itanium was having problems early on and
- it wasn't an internal secret and
- there were groups within Intel that were pushing for other architectures and
- legacy support can bog alternative projects down and
- Blue Crystals (marketing) can drive a company into making the wrong technological choices.
It's an interesting video, well worth watching for the insight Colwell displayed.
Galinstan conducts heat far better than water. Galinstan's conductivity is 16.5 W/(MK) vs. water's value of between .4 to .7.
Several posters have suggested gallium which melts at 40 C. Using pure gallium would be a disaster because when it freezes, it expands like water does. It'd rupture the plumbing inside your computer. Galinstan stays liquid down to -19 C. A spec sheet is available.
Galinstan has a couple of drawbacks. A, it's corrosive and B, it sticks to most surfaces unless the surface is properly prepared. That means your radiator, water block and all the tubing has to be internally coated before you pour Galinstan into your cooling rig.
Why do you think it'll take that long for a dupe?