If you're entering a port where a) pirates could be lurking or b) you could bribe enough staff to look the other way, you have no business being in that port.
Are there physical interlocks in your engineered traffic light systems like the physical drums from days gone by? Those older analog drum systems would physically prevent an all-green situation. If your system has everything in firmware, and isn't protected by some physical relay system or interlock, the proper attacker could inflict an all-green situation.
Because if you make a bad batch of product, no matter what the product is, you can be liable for the value of the product and have to cough it up if your resellers/brick and mortars decide they want the money back they fronted for said defective product.
I'm saying there is limited additional cost involved. You need to know that X materials came from vendors, was put into Y products, and was shipped to Z customers or resellers. This is not about extra paperwork, it's about properly accounting for what you already need to keep track of your business.
That grease? Same thing. Tracability sometimes is quite important, and it costs a lot of money to maintain that paper trail.
Really? Competent businesses aren't keeping these records (electronically) as part of their normal supply chain process? That's somewhat sad to hear the government needs to pay extra for them to keep their own proper records that they'd use in day-to-day quality control.
Somewhat true. At my last gig, we were churning through collider data from the CMS detector at the LHC. We took the stream of data directly from CERN at 10-40Gbps, staged it to several TB of fast spinning disk while waiting to write it to a tape archive (17PB), and then ran specially written software on 5500 "workers" that would suck the data back in from a distributed filesystem and recreate collision events from the data. Using 10Gbps hardware, fast disks, and condor for work management, we were able to keep those workers pretty damn busy. Floored 100% of the time? No, but awfully close.
Note: I'm aware this is a special case, but don't discount how heavily computing resources are used in both industry and scientific/academic settings.
I think you'll see the move to photonics, with in silica light generation/detection used to push bits around. No voltage leakage, no heat issues (ergo, more reliable/longer lasting equipment). Intel is well on it's way with Light Peak.
You are correct. I do travel about 80K miles/year on Singapore Air though, so when I miss a flight, it is very painless getting rebooked on the next one even though I'm not ticketed all the way through.
Speaking of Star Alliance/OneWorld, if you ever get the chance, book an around-the-world ticket.
I paid $5K/ticket for both myself and my wife, and we traveled around the world *cheap* compared to buying individual flights (used OneWorld, got 16 segments across the globe).
That makes sense for Expedia, but AA pulled their listings from Orbitz as well, which is a joint ownership venture between several legacy airlines. I believe American has a share too, and with that share the ability to have input on the transaction costs associated with booking a flight through Orbitz. So why did they pull their listings from there?
It's more efficient to pipe wind, solar, and nuclear from rural areas to the city via HVDC transmission lines than it is for tens or hundreds of thousands of people to commute from the burbs to the city every day.
Also, rural areas will still need to produce food for the city, but will need to get it to the city in an efficient manner.
I'm saying that the current situation of burning 10-100 gallons of fuel a month (depending on your vehicle) to get to your city job from the suburbs will be unsustainable.
Localized population density is simply more efficient than the suburbs. I have no problem with people living in the suburbs, I ask they simply pay for the true cost of getting to their jobs 30-40 miles away and the cost to get goods to them when the price of petroleum goes up. Let the market work it out.
Yes, billions of people in India and China will be able to afford $150/barrel fuel, but people in first world nations won't. Nice logic there.
China already heavily subsidizes oil imports. Also, Americans need far more oil per capita vs third world citizens. If it costs them several dollars to fill their scooter, it's painful. If it costs you $150 to fill your car, it's crippling for the majority.
Which will work fine until oil is at $120-150/barrel, and you're spending a non-negligible amount on fuel for commuting and can't afford your mortgage and food.
There are billions of people between India and China who are going to be driving soon. And who will be using oil to do so. Don't kid yourself, the suburbs are unsustainable.
From what I've heard from internal Netflix IT folks, it's extremely dysfunctional in there, which may explain why it's easier for them to just cut a check and have someone else do it.
Now, if someone with IT competency were to purchase Netflix (Google? I mean, their job is to organize the world's information, and movie content *is* information), I'm sure you'd see a swift internal CDN roll out occur.
The trick would be to use HTTP redirects for large files where 1-2 seconds of initial latency isn't going to cause a problem, but still use the DNS/UDP/Anycast method for small files where you don't care if they're load balanced or not.
Slashdot Mirror
If you're entering a port where a) pirates could be lurking or b) you could bribe enough staff to look the other way, you have no business being in that port.
And how much does the insurance and ransoms cost, as well as the liability of losing tens-hundres of millions of dollars worth of cargo compare?
Airbags are expensive, but there is a reason they're in cars.
If the laser doesn't deter them, than the Phalanx CIWS will:
http://en.wikipedia.org/wiki/Phalanx_CIWS
Accuracy by volume.
Hence why they're locked up in the ship's gun safe when entering port.
Are there physical interlocks in your engineered traffic light systems like the physical drums from days gone by? Those older analog drum systems would physically prevent an all-green situation. If your system has everything in firmware, and isn't protected by some physical relay system or interlock, the proper attacker could inflict an all-green situation.
Because if you make a bad batch of product, no matter what the product is, you can be liable for the value of the product and have to cough it up if your resellers/brick and mortars decide they want the money back they fronted for said defective product.
I'm saying there is limited additional cost involved. You need to know that X materials came from vendors, was put into Y products, and was shipped to Z customers or resellers. This is not about extra paperwork, it's about properly accounting for what you already need to keep track of your business.
That grease? Same thing. Tracability sometimes is quite important, and it costs a lot of money to maintain that paper trail.
Really? Competent businesses aren't keeping these records (electronically) as part of their normal supply chain process? That's somewhat sad to hear the government needs to pay extra for them to keep their own proper records that they'd use in day-to-day quality control.
Somewhat true. At my last gig, we were churning through collider data from the CMS detector at the LHC. We took the stream of data directly from CERN at 10-40Gbps, staged it to several TB of fast spinning disk while waiting to write it to a tape archive (17PB), and then ran specially written software on 5500 "workers" that would suck the data back in from a distributed filesystem and recreate collision events from the data. Using 10Gbps hardware, fast disks, and condor for work management, we were able to keep those workers pretty damn busy. Floored 100% of the time? No, but awfully close.
Note: I'm aware this is a special case, but don't discount how heavily computing resources are used in both industry and scientific/academic settings.
I think you'll see the move to photonics, with in silica light generation/detection used to push bits around. No voltage leakage, no heat issues (ergo, more reliable/longer lasting equipment). Intel is well on it's way with Light Peak.
I read those in between cross country or transatlantic flights. Still have too much to do.
Reading books in digital is great because it is a linear process. But how many people read magazines in a start to finish fashion?
*Raises hand* Scientific American, Wired, The Economist, and MAKE.
I've used an iPad to manage a server from 35K feet on a Virgin Atlantic flight. Not everyone uses it for games.
You are correct. I do travel about 80K miles/year on Singapore Air though, so when I miss a flight, it is very painless getting rebooked on the next one even though I'm not ticketed all the way through.
Speaking of Star Alliance/OneWorld, if you ever get the chance, book an around-the-world ticket.
http://www.fourhourworkweek.com/blog/2010/10/08/round-the-world-plane-ticket/
I paid $5K/ticket for both myself and my wife, and we traveled around the world *cheap* compared to buying individual flights (used OneWorld, got 16 segments across the globe).
Why would someone ever *choose* to live in such an environment? *shivers*
I flew from LAX to Tokyo last year for a tour of Japan, and Singapore Air was glorious. I now try to fly them international whenever possible.
That makes sense for Expedia, but AA pulled their listings from Orbitz as well, which is a joint ownership venture between several legacy airlines. I believe American has a share too, and with that share the ability to have input on the transaction costs associated with booking a flight through Orbitz. So why did they pull their listings from there?
I've been begging Virgin America to fly into either ORD or MDW. Until then, they aren't much use to someone in Chicago =(
In that case, I usually use JetBlue or Southwest to get to New York or SFO and then fly Singapore Air (since rarely do they come into Chicago).
It's more efficient to pipe wind, solar, and nuclear from rural areas to the city via HVDC transmission lines than it is for tens or hundreds of thousands of people to commute from the burbs to the city every day.
Also, rural areas will still need to produce food for the city, but will need to get it to the city in an efficient manner.
I'm saying that the current situation of burning 10-100 gallons of fuel a month (depending on your vehicle) to get to your city job from the suburbs will be unsustainable.
Localized population density is simply more efficient than the suburbs. I have no problem with people living in the suburbs, I ask they simply pay for the true cost of getting to their jobs 30-40 miles away and the cost to get goods to them when the price of petroleum goes up. Let the market work it out.
Yes, billions of people in India and China will be able to afford $150/barrel fuel, but people in first world nations won't. Nice logic there.
China already heavily subsidizes oil imports. Also, Americans need far more oil per capita vs third world citizens. If it costs them several dollars to fill their scooter, it's painful. If it costs you $150 to fill your car, it's crippling for the majority.
And god forbid your system halts and you lose any data you haven't already committed to persistent storage.
Which will work fine until oil is at $120-150/barrel, and you're spending a non-negligible amount on fuel for commuting and can't afford your mortgage and food.
There are billions of people between India and China who are going to be driving soon. And who will be using oil to do so. Don't kid yourself, the suburbs are unsustainable.
From what I've heard from internal Netflix IT folks, it's extremely dysfunctional in there, which may explain why it's easier for them to just cut a check and have someone else do it.
Now, if someone with IT competency were to purchase Netflix (Google? I mean, their job is to organize the world's information, and movie content *is* information), I'm sure you'd see a swift internal CDN roll out occur.
The trick would be to use HTTP redirects for large files where 1-2 seconds of initial latency isn't going to cause a problem, but still use the DNS/UDP/Anycast method for small files where you don't care if they're load balanced or not.