Blocking at the web browser level, where the blocking program has an idea of what's going on, works best. Blocking at the IP level will stall out some sites. It's technically possible to block in the browser in such a way that the site can't figure out that it's being blocked. Few sites detect ad blockers yet, but more could. It may be worthwhile to delay loads of ad sites and see if this stalls the loading of the real content. For mobile, it would be amusing to have an ad-blocking proxy site which reads the ads into the proxy machine but never sends them over the air link.
We need a new level of popup-blocking technology, one that understands HTML layers and decides which ones get to appear. Anybody working on this? Also, most of the existing ad blockers run off of big lists of regular expressions, which are manually updated. That's rather retro technology. They should be using classifiers.
Blocking tracking sites is usually a win. For this page, I'm blocking Google Analytics and Comscore Beacon, using Abine's DoNotTrackMe Firefox add-on. This blocking has the amusing side effect that CBS shows will run without showing any ads.
Of course, with "apps", it's much tougher to block. It may be necessary to run apps under a virtual machine that prevents the app from doing certain things. An ad-hostile version of Flash might be worth constructing.
Should some ads get through? We offer Ad Limiter, which declutters Google search result pages by removing all but one ad. We pick the one ad based on our ratings of site legitimacy. Interestingly, most users of that add-on seem to be business sites - usage is high on weekdays and drops off on weekends. There may be a market for business-based ad blocking products.
It's hard because the space guys want exclusive use of a chunk of spectrum over most of the planet. Then they don't use it that often.
If there was enough space travel to justify it, there would be something like Aeronautical Radio, Inc. to handle it, with ground stations used by many parties. Or Iridium would offer links that were aimed at other directions than the ground. Commercial space isn't big enough yet to justify the investment.
"Home control" has been around since the 1990s. It was once promoted with some really annoying blinking pop-up ads for the X10 wireless control system. Around 2001, X10 was the fourth most popular property on the web. You can still buy X10 gear. It works fine. Nobody cares.
Then there was Echelon LonWorks. This was a technically better system than X10 (which was mostly one-way), and it's widely used in commercial buildings. It has really good noise immunity, which has resulted in it being used to control auxiliary systems (lights, HVAC, destination signs, etc.) in subway trains. As a home control system, which was the original plan, it went nowhere.
There's no problem doing this, and plenty of products are available.
Remote off/on control of home lights and appliances just isn't that useful.
Hello? Why is a web browser launching other applications without explicit user consent? Ever?
This was the classic Microsoft security hole - executing anything that came in which could possibly be executed - Word documents, spreadsheets, autoplay files, Universal Plug and Play. Microsoft has now turned most of that off. Apple is replicating a classic Microsoft mistake here.
Those aren't the phishers you're really worried about. There seem to be about ten "usual suspects" we keep seeing on our phishing reports. The low-end ones are trolling for Habbo Hotel accounts. A few notches up are phony logins for bank accounts (PayPal and HSBC are popular targets. New this week: Swedish tax refunds. And, for some reason, several new phish sites for AOL 9.0 accounts.)
We track these, but they're more of a nuisance than a real threat.
The ones to worry about are better targeted and are of better quality. Those are aimed at corporate login info. Those won't be seen by broad-based phishing detection services because they're only sent to people who might have those logins. So they tend not to be blacklisted.
You'd think I'd be mega unhappy like everyone else at the constant waiting and lack of actually being able to play a game I purchased.
Well, you'd be wrong.
The hours upon hours since launch that I haven't been able to log in, whether it be sitting in queues, or server busy messages, or just plain old not working screens, I've managed to do a heap of things that I never do when I'm locked in my man cave playing video games.
I've washed the dishes, the laundry, changed the oil in the car, mopped the floors, dusted, did a spot of gardening, greeted my children who I hadn't really seen since Christmas, walked the dog, asked how my wife's day has been and listened to the entire response, restocked the groceries and many more things! My family has never been happier that they've got a father and husband again.
In fact, I feel like Simcity has given me a new lease on life. This wouldn't have been possible without the seemingly crazy decision to have constant online connections and server side save points even for single player.
So I can only thank EA and Maxis. Your failures have been my rewards. 5 stars!
Fastenal uses the exact same vending machine (minus all the photos of course!).
Right. It's an Edge 5000 Industrial Vending Machine from Apex Supply Chain Technologies. Fastenall has about 30,000 installed. Facebook, not so many.
These industrial vending machines look like candy machines, but they're more versatile. They can be configured for a wide range of item sizes, they have an IR beam system to make sure the item actually is dispensed, and there are options available for soft handling of fragile items. The machines have an Internet connection, report who got what, and tell the supplier what needs refilling.
Facebook isn't being original here.
Fastenall, which sells cutting tools, bolts, and other useful things used by people who make Real Stuff, has special vending machines for industrial plants. Employees use their employee badge or a PIN to get tools and supplies. The machines report back to Fastenall, and they restock the machines. The customer only pays for items when they're vended.
Here's the Youtube video. Fastenall vends electric drills and WD-40, rather than keyboards and cables. They have little machines for things like drill bits, and locker-sized bins for big items. So they're already doing what Facebook is only talking about.
Destructoid.com - stuck trying to read assets from "craveonline.com", "bulk2.destructoid.com", etc. When it finally comes up, we get a giant picture from Teenage Pokemon, followed by clips from stories. Plus lots of ads.
Their RSS feed is more readable and loads quickly. Now we get to see the content.
It's just some gamer's blog. "This is my favorite episode so far." "There's not a whole lot of information disclosed on how and when the game will released". "Ten golden rules of online gaming." (the usual excuse for hanging ads on every paragraph.) "We had a delightful little Saturday Morning Hangover this morning, playing the recently released Phantom Breaker: Battle Grounds." No insights. No inside information. Not even good trip reports.
Why should this guy expect to make money for writing a personal blog about his hobby?
The image shown as an example has most of the screen real estate tied up with a useless background of car images. Then there's a tiny map. The screen contains no useful information about bypassing the delay. The actual info is less useful than what 511.org or calling 511 provides.
As for dialog boxes, Apple had a spec for those in the original Macintosh user interface guidelines. Trouble dialogs should be two sentences. The first sentence describes the problem. The second suggests corrective action. And you should never have to tell the machine something it already knows.
What they actually say about their design sounds like the design spec for Metro, except without the emphasis on square flat-shaded icons. Scrollable grids of icons presented in more or less random order do not scale well.
The "Computing Research Association" is a lobbying group. It's not on K Street NW in DC like most lobbyists. It's on L street, one block over. It's a lobby for federal funding for college CS departments.
Here's the actual report. Two charts are upside down. The focus is on race and gender. There's little discussion of CS vs IT vs EE vs CE degrees, although there are some separate table columns. Employment statistics are provided only for PhD graduates.
The data seems to be self-reported by the institutions involved.
The district court ruled that the evidence obtained by the search should be suppressed and not admissible as evidence, which means the defendant could not be convicted. The Court of Appeals for the 9th circuit reversed that decision and held that the search was permissible:
For the above reasons, we conclude
that the examination of Cotterman's electronic devices was supported by reasonable suspicion and that the scope and manner of the
search were reasonable under the Fourth
Amendment. Cotterman's motion to
suppress therefore was erroneously
granted.
REVERSED.
Cotterman (the defendant) goes to jail.
Yes, there's some dicta there indicating that there are limits on search authority at the border, but the threshold of suspicion is so low as to be meaningless.
Honestly, sometimes it just seems like Google doesn't now how to run themselves in spite of billions in profit. The are succeeding in spite of themselves.
Google's business is pay-per-click ads. They're really good at pay-per-click ads. Most of their other activities lose money. Google has tried a wide range of products and services, looking for the Next Big Moneymaker. So far, fail.
This is their second try at phones. In a year, we'll know if it generates profits.
New Mexico already was screwed by Virgin Galactic on this. The state built Spaceport America for Virgin. Then Virgin demanded a better deal, and got it. Once in a while Armadillo Aerospace launches some test rocket from there. The terminal building is used for bus tours.
This is the first time in decades that any country with nuclear weapons has seriously threatened to start a nuclear war. Iraq never had nuclear weapons. Iran probably doesn't. North Korea has demonstrated both nuclear weapons and an ICBM. Maybe they can't cross the Pacific Ocean with it, but they can certainly hit South Korea, China, or Japan.
This is the real worry in the world today. Afghanistan is irrelevant. Iran/Israel is a sideshow. North Korea has the military power to do real damage.
If you're in South Korea, I would suggest getting a long way from Seoul for a while.
The problem with building trustworthy robots is that the computer industry can't do it. The computer industry has a culture of irresponsibility. Software companies are not routinely held liable for their mistakes.
Automotive companies are held liable for their mistakes. Structural engineering companies are. Aircraft companies are. Engineers who do civil or structural engineering carry liability insurance, take exams, and put seals of approval on their work.
Trustworthy robots are going to require the kinds of measures take in avionics - multiple redundant systems, systems that constantly check other systems, and backup systems which are completely different from the primary system. None of the advanced robot projects of which I am aware do any of that. The Segway is one of the few consumer robotic-like products with any real redundancy and checking.
The software industry is used to sleazing by on these issues. Much medical equipment runs Windows. We're not ready for trustworthy robots.
What killed the RISC alternatives to x86 was the Pentium Pro. Before the Pentium Pro, the industry consensus was that the way to faster machines was RISC. Then Intel developed a superscalar x86 machine and beat out RISC hardware.
It was an incredible technical achievement to make an instruction set designed for zero parallelism go superscalar. All previous superscalar machines, from the IBM 7030 and CDC 6600 of the 1960s, had imposed restrictions on what programs could do to accommodate the problems of concurrency.
The Pentium Pro didn't do that. All the awful cases were handled. Exceptions were exact. Storing into code just ahead of execution was allowed. It took Intel 3,000 engineers to make that work. Nobody had ever put that level of effort into a CPU design before. The design team for a MIPS processor was about 15 people.
The Pentium Pro was designed for 32-bit code, but still ran 16-bit code. Intel thought that by the time the thing shipped in 1995, the desktop world would be 32-bit. After all, it had been 10 years since the 386 introuced 32-bit mode. The desktop world still wasn't ready. Many users ran Windows 3.1/DOS on the Pentium Pro and complained of slow performance. It ran Windows NT quite well, but NT hadn't achieved much market share yet, much to Microsoft's annoyance. So the Pentium II had more transistors devoted to 16-bit support, fixing that problem. The Pentium II and III use modified Pentium Pro architecture. The Pentium 4 (late 2000) was the next new design.
That was the beginning of the end for RISC. RISC could get a simple CPU to one instruction per clock. Superscalar machines could beat one instruction per clock. Superscalar RISC machines had all the complexity of superscalar CISC machines, combined with a lower code density and thus higher demands on memory bandwidth.
As it turned out, x86 wasn't a bad instruction set to make go fast. RISC thinking was that having lots of registers would help. It doesn't. On a superscalar machine, commits to memory are deferred, and most stack accesses are really coming from registers within the execution units. So there's no win in having lots of user-visible registers. Also, if you have a huge number of registers like a SPARC does, time is wasted saving and restoring them. On the stack, you just move the stack pointer.
Also, RISC code is about twice as large as x86 code. Making all the instructions the same length bloats all the small ones.
The Itanium was an attempt to introduce a proprietary architecture that couldn't be cloned. The Itanium has lots of original, patented technology. It was very different from other CPUs. However, it wasn't better. Just different. Compiling fast code for it was really hard. It was a "build it and they will come" architecture, like the Cell. Except they didn't come.
very little of Coursera's network requirement is "scripts and redirects and google metrics."
Yes, it is. Look at the source of their home page. They load stuff from "ogp.me", lots of stuff from "cloudfront.net", jQuery, then they "boot up Coursera", as a comment in the code says. The home page even has a GIF animation of a loading icon for when this is taking some time. The loader creates a new document which is forced in with "document.write()". The new document loads still more Javascript, chosen by the loader.
Amusingly, there are special cases not only for Opera but for the Playstation 3 browser.
What's the point of a manned ballistic fly-by? All the humans can do is operate some instruments for the brief period they're slingshotting around the planet.
This has been discussed in the US, but, as in Switzerland, as a "veto" by shareholders. Shareholders should determine CEO compensation. The US SEC tracks the total compensation of the top 5 employees of publicly held companies, and those should be the ones the shareholders set salaries for.
The highest paid person in the company should have a total compensation set by a shareholder vote. Each shareholder puts in a number, and the share-weighted median is the CEO salary. The CEO can suggest a number, but it's not the default.
Who gets to vote the shares? The party that pays the taxes on them. Funds that are now tax pass-throughs would have to pass through voting rights to avoid being taxed. No taxation without representation.
For funds with a big portfolio, the shareholders can pick a compensation number for the whole fund, and the fund managers can set the compensation for each CEO. But the total can't exceed the shareholder-chosen value.
I found a Forbes article [forbes.com] which estimates that Google makes a profit of 350k per head, while Valve's is in the 87.5 million per head...
No, that's not right. Read the article again. " More specifically, Newell says of the 250-person company that on a per-employee basis, Valve is more profitable than tech giants like Google and Apple. Google made an average $350,000 in profits per employee in 2010. That means Valve sees profits of around $87.5 million at least." The $87.5 million number is total profit, not per employee.
The Dragon spacecraft/capsule is partially reusable. So far, the Falcon boosters are single-use. Space-X hopes to start recovering the first stage boosters, but that isn't working yet.
Meanwhile, they have 9 Merlin engines per Falcon first stage, one per second stage, and they're building about 400 per year. So they get manufacturing economies of scale. That's more valuable than reusability with heavy refurbishing, which tends to be a labor-intensive custom job. Refurbishing was the big cost problem with the US Space Shuttle - the amount of labor required for each turnaround was very high.
Slashdot Mirror
Blocking at the web browser level, where the blocking program has an idea of what's going on, works best. Blocking at the IP level will stall out some sites. It's technically possible to block in the browser in such a way that the site can't figure out that it's being blocked. Few sites detect ad blockers yet, but more could. It may be worthwhile to delay loads of ad sites and see if this stalls the loading of the real content. For mobile, it would be amusing to have an ad-blocking proxy site which reads the ads into the proxy machine but never sends them over the air link.
We need a new level of popup-blocking technology, one that understands HTML layers and decides which ones get to appear. Anybody working on this? Also, most of the existing ad blockers run off of big lists of regular expressions, which are manually updated. That's rather retro technology. They should be using classifiers.
Blocking tracking sites is usually a win. For this page, I'm blocking Google Analytics and Comscore Beacon, using Abine's DoNotTrackMe Firefox add-on. This blocking has the amusing side effect that CBS shows will run without showing any ads.
Of course, with "apps", it's much tougher to block. It may be necessary to run apps under a virtual machine that prevents the app from doing certain things. An ad-hostile version of Flash might be worth constructing.
Should some ads get through? We offer Ad Limiter, which declutters Google search result pages by removing all but one ad. We pick the one ad based on our ratings of site legitimacy. Interestingly, most users of that add-on seem to be business sites - usage is high on weekdays and drops off on weekends. There may be a market for business-based ad blocking products.
It's hard because the space guys want exclusive use of a chunk of spectrum over most of the planet. Then they don't use it that often.
If there was enough space travel to justify it, there would be something like Aeronautical Radio, Inc. to handle it, with ground stations used by many parties. Or Iridium would offer links that were aimed at other directions than the ground. Commercial space isn't big enough yet to justify the investment.
X10! X10! X10! X10! X10! X10! X10! X10!
"Home control" has been around since the 1990s. It was once promoted with some really annoying blinking pop-up ads for the X10 wireless control system. Around 2001, X10 was the fourth most popular property on the web. You can still buy X10 gear. It works fine. Nobody cares.
Then there was Echelon LonWorks. This was a technically better system than X10 (which was mostly one-way), and it's widely used in commercial buildings. It has really good noise immunity, which has resulted in it being used to control auxiliary systems (lights, HVAC, destination signs, etc.) in subway trains. As a home control system, which was the original plan, it went nowhere.
There's no problem doing this, and plenty of products are available. Remote off/on control of home lights and appliances just isn't that useful.
Hello? Why is a web browser launching other applications without explicit user consent? Ever?
This was the classic Microsoft security hole - executing anything that came in which could possibly be executed - Word documents, spreadsheets, autoplay files, Universal Plug and Play. Microsoft has now turned most of that off. Apple is replicating a classic Microsoft mistake here.
Those aren't the phishers you're really worried about. There seem to be about ten "usual suspects" we keep seeing on our phishing reports. The low-end ones are trolling for Habbo Hotel accounts. A few notches up are phony logins for bank accounts (PayPal and HSBC are popular targets. New this week: Swedish tax refunds. And, for some reason, several new phish sites for AOL 9.0 accounts.) We track these, but they're more of a nuisance than a real threat.
The ones to worry about are better targeted and are of better quality. Those are aimed at corporate login info. Those won't be seen by broad-based phishing detection services because they're only sent to people who might have those logins. So they tend not to be blacklisted.
NASA didn't produce plutonium. DoE's Oak Ridge, TN facility did. NASA just issued the press release. NASA is good at that.
This is the top positive review on Amazon:
You'd think I'd be mega unhappy like everyone else at the constant waiting and lack of actually being able to play a game I purchased.
Well, you'd be wrong.
The hours upon hours since launch that I haven't been able to log in, whether it be sitting in queues, or server busy messages, or just plain old not working screens, I've managed to do a heap of things that I never do when I'm locked in my man cave playing video games.
I've washed the dishes, the laundry, changed the oil in the car, mopped the floors, dusted, did a spot of gardening, greeted my children who I hadn't really seen since Christmas, walked the dog, asked how my wife's day has been and listened to the entire response, restocked the groceries and many more things! My family has never been happier that they've got a father and husband again.
In fact, I feel like Simcity has given me a new lease on life. This wouldn't have been possible without the seemingly crazy decision to have constant online connections and server side save points even for single player.
So I can only thank EA and Maxis. Your failures have been my rewards. 5 stars!
Fastenal uses the exact same vending machine (minus all the photos of course!).
Right. It's an Edge 5000 Industrial Vending Machine from Apex Supply Chain Technologies. Fastenall has about 30,000 installed. Facebook, not so many.
These industrial vending machines look like candy machines, but they're more versatile. They can be configured for a wide range of item sizes, they have an IR beam system to make sure the item actually is dispensed, and there are options available for soft handling of fragile items. The machines have an Internet connection, report who got what, and tell the supplier what needs refilling.
Facebook isn't being original here. Fastenall, which sells cutting tools, bolts, and other useful things used by people who make Real Stuff, has special vending machines for industrial plants. Employees use their employee badge or a PIN to get tools and supplies. The machines report back to Fastenall, and they restock the machines. The customer only pays for items when they're vended.
Here's the Youtube video. Fastenall vends electric drills and WD-40, rather than keyboards and cables. They have little machines for things like drill bits, and locker-sized bins for big items. So they're already doing what Facebook is only talking about.
Carol Barth did well running Autodesk. Not so well at Yahoo, but that was Yahoo's problem. Nobody else has been able to turn around Yahoo either.
Destructoid.com - stuck trying to read assets from "craveonline.com", "bulk2.destructoid.com", etc. When it finally comes up, we get a giant picture from Teenage Pokemon, followed by clips from stories. Plus lots of ads.
Their RSS feed is more readable and loads quickly. Now we get to see the content.
It's just some gamer's blog. "This is my favorite episode so far." "There's not a whole lot of information disclosed on how and when the game will released". "Ten golden rules of online gaming." (the usual excuse for hanging ads on every paragraph.) "We had a delightful little Saturday Morning Hangover this morning, playing the recently released Phantom Breaker: Battle Grounds." No insights. No inside information. Not even good trip reports.
Why should this guy expect to make money for writing a personal blog about his hobby?
The image shown as an example has most of the screen real estate tied up with a useless background of car images. Then there's a tiny map. The screen contains no useful information about bypassing the delay. The actual info is less useful than what 511.org or calling 511 provides.
As for dialog boxes, Apple had a spec for those in the original Macintosh user interface guidelines. Trouble dialogs should be two sentences. The first sentence describes the problem. The second suggests corrective action. And you should never have to tell the machine something it already knows.
What they actually say about their design sounds like the design spec for Metro, except without the emphasis on square flat-shaded icons. Scrollable grids of icons presented in more or less random order do not scale well.
The "Computing Research Association" is a lobbying group. It's not on K Street NW in DC like most lobbyists. It's on L street, one block over. It's a lobby for federal funding for college CS departments.
Here's the actual report. Two charts are upside down. The focus is on race and gender. There's little discussion of CS vs IT vs EE vs CE degrees, although there are some separate table columns. Employment statistics are provided only for PhD graduates.
The data seems to be self-reported by the institutions involved.
The district court ruled that the evidence obtained by the search should be suppressed and not admissible as evidence, which means the defendant could not be convicted. The Court of Appeals for the 9th circuit reversed that decision and held that the search was permissible:
For the above reasons, we conclude that the examination of Cotterman's electronic devices was supported by reasonable suspicion and that the scope and manner of the search were reasonable under the Fourth Amendment. Cotterman's motion to suppress therefore was erroneously granted.
REVERSED.
Cotterman (the defendant) goes to jail.
Yes, there's some dicta there indicating that there are limits on search authority at the border, but the threshold of suspicion is so low as to be meaningless.
Honestly, sometimes it just seems like Google doesn't now how to run themselves in spite of billions in profit. The are succeeding in spite of themselves.
Google's business is pay-per-click ads. They're really good at pay-per-click ads. Most of their other activities lose money. Google has tried a wide range of products and services, looking for the Next Big Moneymaker. So far, fail.
This is their second try at phones. In a year, we'll know if it generates profits.
New Mexico already was screwed by Virgin Galactic on this. The state built Spaceport America for Virgin. Then Virgin demanded a better deal, and got it. Once in a while Armadillo Aerospace launches some test rocket from there. The terminal building is used for bus tours.
This is the first time in decades that any country with nuclear weapons has seriously threatened to start a nuclear war. Iraq never had nuclear weapons. Iran probably doesn't. North Korea has demonstrated both nuclear weapons and an ICBM. Maybe they can't cross the Pacific Ocean with it, but they can certainly hit South Korea, China, or Japan.
This is the real worry in the world today. Afghanistan is irrelevant. Iran/Israel is a sideshow. North Korea has the military power to do real damage.
If you're in South Korea, I would suggest getting a long way from Seoul for a while.
The problem with building trustworthy robots is that the computer industry can't do it. The computer industry has a culture of irresponsibility. Software companies are not routinely held liable for their mistakes.
Automotive companies are held liable for their mistakes. Structural engineering companies are. Aircraft companies are. Engineers who do civil or structural engineering carry liability insurance, take exams, and put seals of approval on their work.
Trustworthy robots are going to require the kinds of measures take in avionics - multiple redundant systems, systems that constantly check other systems, and backup systems which are completely different from the primary system. None of the advanced robot projects of which I am aware do any of that. The Segway is one of the few consumer robotic-like products with any real redundancy and checking.
The software industry is used to sleazing by on these issues. Much medical equipment runs Windows. We're not ready for trustworthy robots.
What killed the RISC alternatives to x86 was the Pentium Pro. Before the Pentium Pro, the industry consensus was that the way to faster machines was RISC. Then Intel developed a superscalar x86 machine and beat out RISC hardware.
It was an incredible technical achievement to make an instruction set designed for zero parallelism go superscalar. All previous superscalar machines, from the IBM 7030 and CDC 6600 of the 1960s, had imposed restrictions on what programs could do to accommodate the problems of concurrency.
The Pentium Pro didn't do that. All the awful cases were handled. Exceptions were exact. Storing into code just ahead of execution was allowed. It took Intel 3,000 engineers to make that work. Nobody had ever put that level of effort into a CPU design before. The design team for a MIPS processor was about 15 people.
The Pentium Pro was designed for 32-bit code, but still ran 16-bit code. Intel thought that by the time the thing shipped in 1995, the desktop world would be 32-bit. After all, it had been 10 years since the 386 introuced 32-bit mode. The desktop world still wasn't ready. Many users ran Windows 3.1/DOS on the Pentium Pro and complained of slow performance. It ran Windows NT quite well, but NT hadn't achieved much market share yet, much to Microsoft's annoyance. So the Pentium II had more transistors devoted to 16-bit support, fixing that problem. The Pentium II and III use modified Pentium Pro architecture. The Pentium 4 (late 2000) was the next new design.
That was the beginning of the end for RISC. RISC could get a simple CPU to one instruction per clock. Superscalar machines could beat one instruction per clock. Superscalar RISC machines had all the complexity of superscalar CISC machines, combined with a lower code density and thus higher demands on memory bandwidth.
As it turned out, x86 wasn't a bad instruction set to make go fast. RISC thinking was that having lots of registers would help. It doesn't. On a superscalar machine, commits to memory are deferred, and most stack accesses are really coming from registers within the execution units. So there's no win in having lots of user-visible registers. Also, if you have a huge number of registers like a SPARC does, time is wasted saving and restoring them. On the stack, you just move the stack pointer.
Also, RISC code is about twice as large as x86 code. Making all the instructions the same length bloats all the small ones.
The Itanium was an attempt to introduce a proprietary architecture that couldn't be cloned. The Itanium has lots of original, patented technology. It was very different from other CPUs. However, it wasn't better. Just different. Compiling fast code for it was really hard. It was a "build it and they will come" architecture, like the Cell. Except they didn't come.
very little of Coursera's network requirement is "scripts and redirects and google metrics."
Yes, it is. Look at the source of their home page. They load stuff from "ogp.me", lots of stuff from "cloudfront.net", jQuery, then they "boot up Coursera", as a comment in the code says. The home page even has a GIF animation of a loading icon for when this is taking some time. The loader creates a new document which is forced in with "document.write()". The new document loads still more Javascript, chosen by the loader.
Amusingly, there are special cases not only for Opera but for the Playstation 3 browser.
What's the point of a manned ballistic fly-by? All the humans can do is operate some instruments for the brief period they're slingshotting around the planet.
This has been discussed in the US, but, as in Switzerland, as a "veto" by shareholders. Shareholders should determine CEO compensation. The US SEC tracks the total compensation of the top 5 employees of publicly held companies, and those should be the ones the shareholders set salaries for.
The highest paid person in the company should have a total compensation set by a shareholder vote. Each shareholder puts in a number, and the share-weighted median is the CEO salary. The CEO can suggest a number, but it's not the default.
Who gets to vote the shares? The party that pays the taxes on them. Funds that are now tax pass-throughs would have to pass through voting rights to avoid being taxed. No taxation without representation.
For funds with a big portfolio, the shareholders can pick a compensation number for the whole fund, and the fund managers can set the compensation for each CEO. But the total can't exceed the shareholder-chosen value.
Now that's shareholder empowerment.
NASA PR must have wanted to get this out before the budget cuts kicked in.
I found a Forbes article [forbes.com] which estimates that Google makes a profit of 350k per head, while Valve's is in the 87.5 million per head...
No, that's not right. Read the article again. " More specifically, Newell says of the 250-person company that on a per-employee basis, Valve is more profitable than tech giants like Google and Apple. Google made an average $350,000 in profits per employee in 2010. That means Valve sees profits of around $87.5 million at least." The $87.5 million number is total profit, not per employee.
The Dragon spacecraft/capsule is partially reusable. So far, the Falcon boosters are single-use. Space-X hopes to start recovering the first stage boosters, but that isn't working yet.
Meanwhile, they have 9 Merlin engines per Falcon first stage, one per second stage, and they're building about 400 per year. So they get manufacturing economies of scale. That's more valuable than reusability with heavy refurbishing, which tends to be a labor-intensive custom job. Refurbishing was the big cost problem with the US Space Shuttle - the amount of labor required for each turnaround was very high.