but here in the States there seems to be a social stigma among younger graduates attached to manufacturing jobs that sometimes clouds one's financial judgement.
From the article, writing about China:
"Students themselves have not adjusted to the concept of mass education, so students are accustomed to seeing themselves as becoming part of an elite when they enter college"...
China has a millenniums-old Confucian tradition in which educated people do not engage in manual labor.
The US used to be more about manufacturing, and there was no disgrace to being an engineer in a factory. There was a certain contempt for "college men" as impractical and lazy. That lasted through WWII and into the 1950s. Then came the post-war education boom, a vast number of college graduates, and, for a while, jobs for them. Then came information technology, and a huge cutback in paper-pushing.
China is going into their education boom with the paper-pushing era already over.
The original place was sold many, many years ago. The original location on Fair Oaks is under condos now.
No, that was HalTek, a competitor of HalTed. HalTed is still there. They have a faded copy of a receipt from Steve Jobs, buying a used 'scope, on their bulletin board.
Address space randomization is a form of security through obscurity. It's also an admission that your system security really sucks. The concept is that the code is full of exploitable buffer overflows, but address space randomization will make it harder for exploits to patch the right target area.
So low-level exploits tend to crash the system, or at least just mess it up, rather than getting their code executed.
There are now "address spraying" attacks which counter address space randomization, so this is already an obsolete defensive measure.
We're entering an era where more countries have nuclear weapons. They've become too easy to make. Isotope separation used to take huge gaseous-diffusion plants. Entire cities were built just to enrich uranium.
Centrifuge plants are now medium sized industrial park installations. That's URENCO's plant in New Mexico. It produces enough enriched uranium to power a sizable fraction of US reactors, and it's being expanded. A much smaller plant could enrich enough uranium for a few bombs.
Once you have enriched uranium, making a nuclear bomb isn't a huge job. As a build, it's roughly comparable to making an auto engine from scratch, a job that some auto racing shops can do. Machining uranium isn't that hazardous. Here's a how-to guide from Union Carbide from the 1980s. (Plutonium is a totally different story; there you need glove boxes, remote manipulators, and huge precautions against dust escaping.) There aren't many secrets left about how ordinary atomic bombs work. It's been almost 70 years, after all. (Some of the tricks of fusion weapons still haven't leaked out.)
We've been very lucky that this was a hard thing to do. But it's not that hard any more, and it keeps getting easier.
"The backdoor accounts are present on in all available versions of Barracuda Spam and Virus Firewall, Web Filter, Message Archiver, Web Application Firewall, Link Balancer, Load Balancer, and SSL VPN appliances."
That cannot have happened by accident.
Barracuda Networks should be charged with material support of terrorism for this.
There are far fewer middle managers than there used to be. Span of control (number of persons reporting to a manager) was typically 4-5 in the 1950s. Now it's typically up to around 8-10. This is a direct result of improved information technology. This implies less upward mobility.
Retail is shrinking. The US has a lot of closed stores and dead malls. They're not coming back. First Wal-Mart clobbered the small town main street, and now Amazon is clobbering what's left.
A less discussed side effect of information technology is that it's now possible to run bigger business units than before. Before heavy use of computers and networks, management and control problems of scaling tended to choke large businesses. Big companies had trouble getting out of their own way. Dividing companies into divisions was necessary just to deal with scaling issues. General Motors was the classic example of the division-based company. Each brand had its own factories.
That's much less true today. Wal-Mart, McDonalds, and Amazon don't seem to have scaling problems. Automakers run as units, with work farmed out to various factories as appropriate. With no real operational limits on business size, (and weak antitrust enforcement) the trend is towards a world where there are only a very few huge businesses in each category.
In the entire history of the Bell System, no electromechanical central office was ever down for more than 30 minutes for any reason other than a natural disaster. Not because the components were reliable, but because the architecture was. If you design high-reliability systems, you should understand the architecture of Number 5 Crossbar.
Not sure how or why anybody would actually leave Apple to go to Palm..
Because Steve Jobs was a bully. I know someone who left Apple for Palm. He worked on the iPhone, stayed until it shipped, then left. He was a quiet guy and didn't like being yelled at by Steve Jobs.
Automatic burger machines date back to the 1950s. Back then, everybody ate the same thing, so assembly-like type systems were useful. American Machine and Foundry built an automated fast-food outlet in the 1960s, but it wasn't cost-effective. McDonalds tried this out back in 2003.
It's not that it's technically difficult. It's that the volume required to make it profitable is higher than most fast food outlets can sell.
We've heard this before from the top-down AI crowd. I went through Stanford CS in the 1980s when that crowd was running things, so I got the full pitch. The Cyc project is, amazingly, still going on after 29 years. The classic disease of the academic AI community was acting like strong AI was just one good idea away. It's harder than that.
On the other hand, it's quite likely that Google can come up with something that answers a large fraction of the questions people want to ask Google. Especially if they don't actually have to answer them, just display reasonably relevant information. They'll probably get a usable Siri/Wolfram Alpha competitor.
The long slog to AI up from the bottom is going reasonably well. We're through the "AI Winter". Optical character recognition works quite well. Face recognition works. Automatic driving works. (DARPA Grand Challenge) Legged locomotion works. (BigDog). This is real progress over a decade ago.
Scene understanding and manipulation in uncontrolled environments, not so much. Willow Garage has towel-folding working, and can now match and fold socks. The DARPA ARM program is making progress very slowly. Watch their videos to see really good robot hardware struggling to slowly perform very simple manipulation tasks. DARPA is funding the DARPA Humanoid Challenge to kick some academic ass on this. (The DARPA challenges have a carrot and a stick component. The prizes get the attention, but what motivates major schools to devote massive efforts to these projects are threats of a funding cutoff if they can't get results. Since DARPA started doing this under Tony Tether, there's been a lot more progress.)
Slowly, the list of tasks robots can do increases. More rapidly, the cost of the hardware decreases, which means more commercial applications. The Age of Robots isn't here yet, but it's coming. Not all that fast. Robots haven't reached the level of even the original Apple II in utility and acceptance. Right now, I think we're at the level of the early military computer systems, approaching the SAGE prototype stage. (SAGE was an 1950s air defense system. It had real time computers, data communication links, interactive graphics, light guns, and control of remote hardware. The SAGE prototype was the first system to have all that. Now, everybody has all that on their phone. It took half a century to get here from there.)
The obvious company to buy the XBox line would be Hon Hai Precision Industries, the parent of Foxconn. They already make the XBox. Hon Hai's CEO wants to develop a global brand of their own. It would just mean Hon Hai taking over a slightly larger portion of the supply and marketing chain for something they already make.
the printer's website is long on hype, short on real information or accomplishments.
I noticed that. The site makes claims about how strong the material is, but they compare it to marble. Marble is brittle. If you drop a standard 0.400" marble tile on a hard surface, it will probably shatter. The tensile strength of stone is poor. Stone structures are usually designed to have almost everything in compression. Stone beams are limited to narrow windows and doors.
A machine for turning out heavy cardboard forms for concrete might be more useful. Concrete columns are often cast inside heavy cardboard tubes. This is called Sonotube construction. Once the concrete sets, the cardboard is peeled off. More complex surfaces have to be built up as metal or wooden forms. There are sets of metal forms for complex, but usually rather boring standard designs. "Manhole" and "stairs" form sets are popular. There are sets of standard form parts which are bolted together to make the forms for concrete buildings. The end result tends to be rather Lego-like, but it's fast and cheap.
A system for making concrete forms with organic curves would be valuable to architects. Decorative surfaces are sometimes made by machining custom form inserts to put designs on a surface. But that's limited to surface decoration, signs, and such. A more general 3D system would be useful. But it doesn't have to make stone. Heavy cardboard or dense foam would be enough for many concrete form applications.
A surprisingly large number of key data centers and control points have been relocated to locations in Northern Virginia near CIA HQ. AOL is there. The Iridium satellite control center is there. (It used to be in Schaumburg, IL, near Motorola HQ) Ashburn alone has four Equnix colo facilities, two AT&T data centers, two Net2EZ facilities, and a few other major centers.
A few miles away in Vienna, VA, even closer to CIA HQ in McLean and less than a mile from "Liberty Crossing" (Homeland Security HQ) there are six more big data centers.
On the consumer side, where the contracts are heavily biased towards the vendor, it's worse. Apple dropped MobileMe, and Google dropped a long list of products. Windows Live Mesh shuts down February 13, 2013.
When cloud services die, they tend to die fast. A business which relied on a "cloud" service can be in big trouble. The best case is a frantic effort to get the data off and move to some alternative. Worst case is the data gets lost.
When a page fault occurs part-way through an instruction, the CPU has to interrupt execution. After the page has been brought in from disk, execution can resume. But it must resume as if a page fault hadn't occurred. The usual approach is to restart from the instruction that failed, which means that instruction gets done twice.
The problem is that some instructions aren't idempotent - doing them twice has effects different than doing them once.
On some CPUs, an instruction can call for both a memory access and a register increment. If the memory access faults, the register must not be incremented twice. So either the instruction has to be backed out to the state just before it started, or the state of the partially executed instruction has to be saved in the interrupted state. (The M68010 actually did the latter; there were extra words in the state saved on an interrupt to hold data about partially finished instructions.)
This gets much more complicated in superscalar machines, where multiple instructions have to be undone.
See
these lecture notes from a CS course at U. Vermont, which discusses "back-out", and its successors. In machines with out-of-order execution superscalar processors, you can't just back up; undoing the state of the CPU on a page fault is a big deal. It works, but it took Intel 3,000 engineers to design the Pentium Pro to do out of order x86 code.
The Lisa got so many things right. A good GUI, a protected-memory operating system, and a hard drive file system. The problem was price.
The price problem was due to trouble at Motorola. The Motorola 68000 didn't do instruction backout properly, so it couldn't handle page faults correctly. That was corrected in the Motorola 68010, but the 68010 was too late for the Lisa. So the Lisa had to use a compiler hack to work around the lack of instruction backout.
Because the 68000 couldn't do instruction backout, Motorola didn't make an MMU chip for it. So the Lisa had a custom MMU built out of a large number of ICs. This pushed the parts count and cost way up.
Because good hard drives weren't available for personal computers when the Lisa was designed, Apple built their own, the LisaFile. Apple's attempt at hard drive manufacturing produced a slow, expensive, unreliable drive.
By the time the Lisa shipped, Sun was shipping the Sun I, and the UNIX workstation era had started. The Lisa was in the same price range as UNIX workstations, but the Sun I had a 68010, Ethernet, and hard drives that were expensive but worked.
If it weren't for the instruction backout problem on the 68000, the history of computing could have been completely different. The Lisa was usable, but overpriced. The original Macintosh was an appallingly weak machine - one or two floppies, a slow CPU, and very little memory. This tends to be forgotten, but the original Mac was a commercial failure. Not until the hardware was built up to 512K and a hard drive was supported did it become profitable. (Or usable.) But it was saddled with an OS designed for 64K of RAM. (The original MacOS had a good GUI, but under the hood, it was a lot like DOS - not only was there no memory protection, there wasn't even a CPU dispatcher. The original Mac was supposed to have only 64K of RAM (most of the OS was in ROM) but shortly before shipment, it was increased to 128K.)
Do you get 500GB of storage if it's not pirated? For pirated content, they can usually merge duplicates with the other copies of the same pirated content. If it's unique content, they really have to store it.
Will they support Mecurial or some other revision control system? I'd like to store Autodesk Inventor engineering design files for my personal projects, which are many gigabytes of binary files if you keep all the revisions. 500GB on Dropbox is $500 per year, and Github gets upset if you store big binary files on their system.
I use the files on different machines at different TechShop locations, and have to haul them around on USB sticks and manage backups. So I have a use case for this.
Until cost and EROEI figures come out, this is vaporware. There are lots of ways to make fuel from biomass, but most of them are too expensive. Some consume more energy than they produce (EROEI < 1). Any useful process needs an EROEI over 5, and preferably over 10, to be worth the trouble. Photovoltaic is now up to 7, which is encouraging. Ethanol from corn is listed as 1.3, and some studies put it at less than 1. (Ethanol distillation plants, unlike oil refineries, don't run on their own product; they take in natural gas or some other fuel.)
I see the hemp enthusiasts are out in force again. Hemp isn't a good fuel crop. If you just want biomass for cellulose, you use agricultural waste - corn husks and cobs, straw, bagasse from sugar cane, etc. Hemp seed oil is useful, but only a small part of the biomass comes out as oil. There are better plants for direct oil production.
"Baxter" looks like a clone of the Yaskawa Motoman SDA two-armed robot. Brooks quotes a cheaper price, though; the SDA dual-arm is about $63K. Mechanically there's nothing new here.
Brooks claims better safety systems and easier programming, so that the thing doesn't have to be run behind safety fences. That's the claimed innovation. It's about time for that. Industrial robots have been expensive semi-custom products for decades, and there's no good reason for that. Today, it's cheaper to include a vision system and good force feedback than to support both smart and dumb versions. iRobot's experience with the Roomba has taught them how to deploy and service standard robots in quantity. So they have a good chance of bringing this off.
Microsoft discontinued support for Silverlight last year. Why are they using it?
I'm underwhelmed with streaming video. Yesterday, CBS's streaming video kept stalling because one of the anal-probe tracking systems, "adobetag.com", was not responding reliably. Video playback would stall with "Waiting for adobetag.com" in the browser status.
Probably the same way they were in the UK and Italy. Apple was using false and misleading advertising to sell unnecessary "AppleCare" coverage when EU law required a 2-year warranty built into the price of the product.
Slashdot Mirror
but here in the States there seems to be a social stigma among younger graduates attached to manufacturing jobs that sometimes clouds one's financial judgement.
From the article, writing about China:
"Students themselves have not adjusted to the concept of mass education, so students are accustomed to seeing themselves as becoming part of an elite when they enter college" ...
China has a millenniums-old Confucian tradition in which educated people do not engage in manual labor.
The US used to be more about manufacturing, and there was no disgrace to being an engineer in a factory. There was a certain contempt for "college men" as impractical and lazy. That lasted through WWII and into the 1950s. Then came the post-war education boom, a vast number of college graduates, and, for a while, jobs for them. Then came information technology, and a huge cutback in paper-pushing.
China is going into their education boom with the paper-pushing era already over.
The original place was sold many, many years ago. The original location on Fair Oaks is under condos now.
No, that was HalTek, a competitor of HalTed. HalTed is still there. They have a faded copy of a receipt from Steve Jobs, buying a used 'scope, on their bulletin board.
Address space randomization is a form of security through obscurity. It's also an admission that your system security really sucks. The concept is that the code is full of exploitable buffer overflows, but address space randomization will make it harder for exploits to patch the right target area. So low-level exploits tend to crash the system, or at least just mess it up, rather than getting their code executed.
There are now "address spraying" attacks which counter address space randomization, so this is already an obsolete defensive measure.
We're entering an era where more countries have nuclear weapons. They've become too easy to make. Isotope separation used to take huge gaseous-diffusion plants. Entire cities were built just to enrich uranium. Centrifuge plants are now medium sized industrial park installations. That's URENCO's plant in New Mexico. It produces enough enriched uranium to power a sizable fraction of US reactors, and it's being expanded. A much smaller plant could enrich enough uranium for a few bombs.
Once you have enriched uranium, making a nuclear bomb isn't a huge job. As a build, it's roughly comparable to making an auto engine from scratch, a job that some auto racing shops can do. Machining uranium isn't that hazardous. Here's a how-to guide from Union Carbide from the 1980s. (Plutonium is a totally different story; there you need glove boxes, remote manipulators, and huge precautions against dust escaping.) There aren't many secrets left about how ordinary atomic bombs work. It's been almost 70 years, after all. (Some of the tricks of fusion weapons still haven't leaked out.)
We've been very lucky that this was a hard thing to do. But it's not that hard any more, and it keeps getting easier.
"The backdoor accounts are present on in all available versions of Barracuda Spam and Virus Firewall, Web Filter, Message Archiver, Web Application Firewall, Link Balancer, Load Balancer, and SSL VPN appliances."
That cannot have happened by accident. Barracuda Networks should be charged with material support of terrorism for this.
There are far fewer middle managers than there used to be. Span of control (number of persons reporting to a manager) was typically 4-5 in the 1950s. Now it's typically up to around 8-10. This is a direct result of improved information technology. This implies less upward mobility.
Retail is shrinking. The US has a lot of closed stores and dead malls. They're not coming back. First Wal-Mart clobbered the small town main street, and now Amazon is clobbering what's left.
A less discussed side effect of information technology is that it's now possible to run bigger business units than before. Before heavy use of computers and networks, management and control problems of scaling tended to choke large businesses. Big companies had trouble getting out of their own way. Dividing companies into divisions was necessary just to deal with scaling issues. General Motors was the classic example of the division-based company. Each brand had its own factories.
That's much less true today. Wal-Mart, McDonalds, and Amazon don't seem to have scaling problems. Automakers run as units, with work farmed out to various factories as appropriate. With no real operational limits on business size, (and weak antitrust enforcement) the trend is towards a world where there are only a very few huge businesses in each category.
That's where the middle class went.
In the entire history of the Bell System, no electromechanical central office was ever down for more than 30 minutes for any reason other than a natural disaster. Not because the components were reliable, but because the architecture was. If you design high-reliability systems, you should understand the architecture of Number 5 Crossbar.
Not sure how or why anybody would actually leave Apple to go to Palm..
Because Steve Jobs was a bully. I know someone who left Apple for Palm. He worked on the iPhone, stayed until it shipped, then left. He was a quiet guy and didn't like being yelled at by Steve Jobs.
Automatic burger machines date back to the 1950s. Back then, everybody ate the same thing, so assembly-like type systems were useful. American Machine and Foundry built an automated fast-food outlet in the 1960s, but it wasn't cost-effective. McDonalds tried this out back in 2003.
It's not that it's technically difficult. It's that the volume required to make it profitable is higher than most fast food outlets can sell.
We've heard this before from the top-down AI crowd. I went through Stanford CS in the 1980s when that crowd was running things, so I got the full pitch. The Cyc project is, amazingly, still going on after 29 years. The classic disease of the academic AI community was acting like strong AI was just one good idea away. It's harder than that.
On the other hand, it's quite likely that Google can come up with something that answers a large fraction of the questions people want to ask Google. Especially if they don't actually have to answer them, just display reasonably relevant information. They'll probably get a usable Siri/Wolfram Alpha competitor.
The long slog to AI up from the bottom is going reasonably well. We're through the "AI Winter". Optical character recognition works quite well. Face recognition works. Automatic driving works. (DARPA Grand Challenge) Legged locomotion works. (BigDog). This is real progress over a decade ago.
Scene understanding and manipulation in uncontrolled environments, not so much. Willow Garage has towel-folding working, and can now match and fold socks. The DARPA ARM program is making progress very slowly. Watch their videos to see really good robot hardware struggling to slowly perform very simple manipulation tasks. DARPA is funding the DARPA Humanoid Challenge to kick some academic ass on this. (The DARPA challenges have a carrot and a stick component. The prizes get the attention, but what motivates major schools to devote massive efforts to these projects are threats of a funding cutoff if they can't get results. Since DARPA started doing this under Tony Tether, there's been a lot more progress.)
Slowly, the list of tasks robots can do increases. More rapidly, the cost of the hardware decreases, which means more commercial applications. The Age of Robots isn't here yet, but it's coming. Not all that fast. Robots haven't reached the level of even the original Apple II in utility and acceptance. Right now, I think we're at the level of the early military computer systems, approaching the SAGE prototype stage. (SAGE was an 1950s air defense system. It had real time computers, data communication links, interactive graphics, light guns, and control of remote hardware. The SAGE prototype was the first system to have all that. Now, everybody has all that on their phone. It took half a century to get here from there.)
"Leveraged support model" does not mean what you think it means.
Right. The current record for all solar cells is 44%. 27% has been achieved without rare materials.
When you see indium and gallium in the materials list, it's not going to be a high-volume product.
The obvious company to buy the XBox line would be Hon Hai Precision Industries, the parent of Foxconn. They already make the XBox. Hon Hai's CEO wants to develop a global brand of their own. It would just mean Hon Hai taking over a slightly larger portion of the supply and marketing chain for something they already make.
the printer's website is long on hype, short on real information or accomplishments.
I noticed that. The site makes claims about how strong the material is, but they compare it to marble. Marble is brittle. If you drop a standard 0.400" marble tile on a hard surface, it will probably shatter. The tensile strength of stone is poor. Stone structures are usually designed to have almost everything in compression. Stone beams are limited to narrow windows and doors.
A machine for turning out heavy cardboard forms for concrete might be more useful. Concrete columns are often cast inside heavy cardboard tubes. This is called Sonotube construction. Once the concrete sets, the cardboard is peeled off. More complex surfaces have to be built up as metal or wooden forms. There are sets of metal forms for complex, but usually rather boring standard designs. "Manhole" and "stairs" form sets are popular. There are sets of standard form parts which are bolted together to make the forms for concrete buildings. The end result tends to be rather Lego-like, but it's fast and cheap.
A system for making concrete forms with organic curves would be valuable to architects. Decorative surfaces are sometimes made by machining custom form inserts to put designs on a surface. But that's limited to surface decoration, signs, and such. A more general 3D system would be useful. But it doesn't have to make stone. Heavy cardboard or dense foam would be enough for many concrete form applications.
A surprisingly large number of key data centers and control points have been relocated to locations in Northern Virginia near CIA HQ. AOL is there. The Iridium satellite control center is there. (It used to be in Schaumburg, IL, near Motorola HQ) Ashburn alone has four Equnix colo facilities, two AT&T data centers, two Net2EZ facilities, and a few other major centers.
A few miles away in Vienna, VA, even closer to CIA HQ in McLean and less than a mile from "Liberty Crossing" (Homeland Security HQ) there are six more big data centers.
For a less clueless article, see "France Telecom and Google entangled in peering fight".
"Google Health has been discontinued"
2012 was the Year of the Cloud Going Away. Several major service vendors bailed completely. GoDaddy dropped their cloud service last October, Dell discontinued their Quest Cloud Automation Platform and Harris dropped theirs last February.
On the consumer side, where the contracts are heavily biased towards the vendor, it's worse. Apple dropped MobileMe, and Google dropped a long list of products. Windows Live Mesh shuts down February 13, 2013.
When cloud services die, they tend to die fast. A business which relied on a "cloud" service can be in big trouble. The best case is a frantic effort to get the data off and move to some alternative. Worst case is the data gets lost.
What is instruction backout?
When a page fault occurs part-way through an instruction, the CPU has to interrupt execution. After the page has been brought in from disk, execution can resume. But it must resume as if a page fault hadn't occurred. The usual approach is to restart from the instruction that failed, which means that instruction gets done twice.
The problem is that some instructions aren't idempotent - doing them twice has effects different than doing them once. On some CPUs, an instruction can call for both a memory access and a register increment. If the memory access faults, the register must not be incremented twice. So either the instruction has to be backed out to the state just before it started, or the state of the partially executed instruction has to be saved in the interrupted state. (The M68010 actually did the latter; there were extra words in the state saved on an interrupt to hold data about partially finished instructions.)
This gets much more complicated in superscalar machines, where multiple instructions have to be undone. See these lecture notes from a CS course at U. Vermont, which discusses "back-out", and its successors. In machines with out-of-order execution superscalar processors, you can't just back up; undoing the state of the CPU on a page fault is a big deal. It works, but it took Intel 3,000 engineers to design the Pentium Pro to do out of order x86 code.
The Lisa got so many things right. A good GUI, a protected-memory operating system, and a hard drive file system. The problem was price. The price problem was due to trouble at Motorola. The Motorola 68000 didn't do instruction backout properly, so it couldn't handle page faults correctly. That was corrected in the Motorola 68010, but the 68010 was too late for the Lisa. So the Lisa had to use a compiler hack to work around the lack of instruction backout.
Because the 68000 couldn't do instruction backout, Motorola didn't make an MMU chip for it. So the Lisa had a custom MMU built out of a large number of ICs. This pushed the parts count and cost way up.
Because good hard drives weren't available for personal computers when the Lisa was designed, Apple built their own, the LisaFile. Apple's attempt at hard drive manufacturing produced a slow, expensive, unreliable drive.
By the time the Lisa shipped, Sun was shipping the Sun I, and the UNIX workstation era had started. The Lisa was in the same price range as UNIX workstations, but the Sun I had a 68010, Ethernet, and hard drives that were expensive but worked.
If it weren't for the instruction backout problem on the 68000, the history of computing could have been completely different. The Lisa was usable, but overpriced. The original Macintosh was an appallingly weak machine - one or two floppies, a slow CPU, and very little memory. This tends to be forgotten, but the original Mac was a commercial failure. Not until the hardware was built up to 512K and a hard drive was supported did it become profitable. (Or usable.) But it was saddled with an OS designed for 64K of RAM. (The original MacOS had a good GUI, but under the hood, it was a lot like DOS - not only was there no memory protection, there wasn't even a CPU dispatcher. The original Mac was supposed to have only 64K of RAM (most of the OS was in ROM) but shortly before shipment, it was increased to 128K.)
Do you get 500GB of storage if it's not pirated? For pirated content, they can usually merge duplicates with the other copies of the same pirated content. If it's unique content, they really have to store it.
Will they support Mecurial or some other revision control system? I'd like to store Autodesk Inventor engineering design files for my personal projects, which are many gigabytes of binary files if you keep all the revisions. 500GB on Dropbox is $500 per year, and Github gets upset if you store big binary files on their system.
I use the files on different machines at different TechShop locations, and have to haul them around on USB sticks and manage backups. So I have a use case for this.
Until cost and EROEI figures come out, this is vaporware. There are lots of ways to make fuel from biomass, but most of them are too expensive. Some consume more energy than they produce (EROEI < 1). Any useful process needs an EROEI over 5, and preferably over 10, to be worth the trouble. Photovoltaic is now up to 7, which is encouraging. Ethanol from corn is listed as 1.3, and some studies put it at less than 1. (Ethanol distillation plants, unlike oil refineries, don't run on their own product; they take in natural gas or some other fuel.)
I see the hemp enthusiasts are out in force again. Hemp isn't a good fuel crop. If you just want biomass for cellulose, you use agricultural waste - corn husks and cobs, straw, bagasse from sugar cane, etc. Hemp seed oil is useful, but only a small part of the biomass comes out as oil. There are better plants for direct oil production.
"Baxter" looks like a clone of the Yaskawa Motoman SDA two-armed robot. Brooks quotes a cheaper price, though; the SDA dual-arm is about $63K. Mechanically there's nothing new here.
Brooks claims better safety systems and easier programming, so that the thing doesn't have to be run behind safety fences. That's the claimed innovation. It's about time for that. Industrial robots have been expensive semi-custom products for decades, and there's no good reason for that. Today, it's cheaper to include a vision system and good force feedback than to support both smart and dumb versions. iRobot's experience with the Roomba has taught them how to deploy and service standard robots in quantity. So they have a good chance of bringing this off.
Microsoft discontinued support for Silverlight last year. Why are they using it?
I'm underwhelmed with streaming video. Yesterday, CBS's streaming video kept stalling because one of the anal-probe tracking systems, "adobetag.com", was not responding reliably. Video playback would stall with "Waiting for adobetag.com" in the browser status.
Most of those "apps" are probably some kind of web content that was run through a packaging system to turn it into an "app".
how is Apple misleading customers?
Probably the same way they were in the UK and Italy. Apple was using false and misleading advertising to sell unnecessary "AppleCare" coverage when EU law required a 2-year warranty built into the price of the product.