Where would one purchase the air valves to make one of these ? Are these low cost items or expensive specialist things.
Ordinary solenoid on/off air valves are widely used, easily available and cheap. Grainger stocks them, and they show up in surplus stores. Proportional air valves are somewhat exotic items.
From the video, it looks like the robot's legs are controlled through simple on-off valves (notice the clicks and the jerky leg motion) but the trunk has full proportional control and feedback.
Precision pneumatics is effective but not used all that much. Most pneumatic actuators are a simple solenoid valve controlling air to a cylinder, with no feedback. Precision control of a pneumatic actuator requires a pair of proportional valves (which themselves require internal feedback on their valve spool) and a double acting air cylinder or other actuator with position and pressure feedback. There are at least two analog outputs (the valve solenoids) and five inputs (two valve positions, two pressures, and one actuator position) to deal with for each actuator. The control theory for this is complex. If you do all that, though, you get an actuator that behaves like a spring with adjustable spring constant, zero position, and damping. Which is what a muscle pair has.
The problem with pneumatic mobile robots, of course, is lugging the air compressor around.
We get at least one of these overhyped materials science articles each month. This time, someone has figured out how to deposit a garnet layer in a wafer fab. This is blown up into "photonic computing real soon now". It's not.
There's a lot of work in progress (PowerPoint) on optical on-chip interconnects. This is not "photonic computing". It's clusters of CPUs with a network of optical interconnects, all on one IC. The CPUs are still made of transistors. IBM has a very active research program in this area. But it's a long way from working. There are optical switching elements that work experimentally, but nothing ready for volume manufacturing yet. The optical interconnects themselves aren't considered to be the big problem.
So far, most of the proposed approaches involve un-buffered circuit switched networks. An optical connection is set up from CPU 1 to CPU 2 by electrical means, and then data is blasted across it. Circuit setup time is long compared to the data rate. So this is for long messages within a cluster, not cache synchronization. Think (inevitably) Beowulf cluster on a chip, not thousand-CPU shared memory microprocessors. The technology may also be useful as a network optical switch.
Short version: when this all works, servers get more densely packed.
Google's stock peaked in 2007. Google is making money, but it's keeping all of it. They don't pay a dividend. Investors would like to see some return on their investment.
Google still has the big problem that they only have one revenue source, ads. Nothing else they've tried makes much money. 96% is still from ads. Google's flings with telephony and social networking aren't contributing significant revenue. Being #1 in giving stuff away isn't paying off as well as expected.
Here's the DEA's list. Those marked as "List 1" are the most restricted. It's not that long a list. Iodine is the only chemical on List 1 that isn't particularly hazardous.
I know one business which has good results with Groupon. It's a riding school (horses). The Groupon lessons are an introduction to riding, with about an hour on the ground grooming and learning about horses, and an hour on the horse. Groupon insures that enough people sign up to make up a profitable group, and Groupon pays for no-shows in bad weather. This insures that each lesson brings in a known income.
Some of the people discover they like horses and come back for more lessons. But even without that, it's profitable.
This is the right way to use Groupon - not as a traffic builder, but for something that's done in a group.
Also, not being powered by anthing more than the waves themselves, I am sure they are subject to the ocean currents.
Actually, they tend to stay within 50 meters of their programmed track. I've seen a presentation by the developers. They have a GPS and follow waypoints, and they have an Iridium satellite phone link. As long as there's a little wind to produce even light chop, they get enough energy from wave motion to overcome ocean currents.
Itanium was a joint Intel-HP project, remember? HP might well pay Intel to keep it alive.
The idea behind Itanium was that it had lots of new, different, patentable technology, so Intel didn't have to worry about clones. The problem was that it wasn't better technology. Just different.
Classic bad CPU architecture ideas of the "build it and they will come" variety:
"Hey, let's build a machine with lots of little CPUs that don't share memory and intercommunicate via I/O!" Examples are the nCube, the Connection Machine, and the Cell processor. There's no problem building such machines, but chopping the problem up into communicating bite-sized pieces is very tough, and very closely tied to the specific hardware.
"Hey, let's build a Very Long Instruction Word machine so we can run several instructions at once!". A success for some signal processing chips, but general purpose CPUs based on VLIW technology, the i860 and the Itanium, didn't do so well. Intel tried to deny that the Itanium was a VLIW machine, but it is. Optimizing compilers for such machines are very hard. (I met the HP guys trying to do the Itanium compiler once. It was not going well.)
"Hey, let's build a shared-memory multiprocessor with non-synchronized caches!" This has been tried a few times. The usual result is software race conditions which are very tough to find, and an extremely painful programming model.
In the spectrum of concurrency, shared-memory mulitprocessors with synchronized caches work, and clusters of powerful machines which communicate over networks work. Those are the extremes of the concurrency range. With the notable exception of graphics processors, no machine in the middle of that range has been a success. Such machines can be built, but are so hard to program they're always behind the classical architectures. The Cell in the PS/3 is the only example ever deployed in volume, and that nearly killed Sony.
This is a form of progress. However, one would expect the OS itself to have much stronger defenses against anything from the outside running at kernel level. There should be no way to get a "boot sector virus" onto a machine while running under an operating system.
I had the unfortunate experience of going through Stanford for a MSCS just before the "AI Winter". The "expert systems" (remember "expert systems"? ) profs were running the department. It was becoming clear that expert systems weren't going anywhere, and the faculty was in denial about that. They'd set up a 5-year "knowledge engineer" program, with a combination of computer science theory, philosophy, and psychological interviewing technique to write rules for expert systems (Where are those people now?) I had one exam where a question was "Does a rock have intentions"?
It took over a decade for the CS department to recover. After I graduated, the CS department was moved from Arts and Sciences, where it had been mostly autonomous, to Engineering, where it had adult supervision. It wasn't until the DARPA Grand Challenge forced Stanford to bring in machine learning people from CMU that the department really started moving forward again. Now they're making real progress.
(This is not well known, but Tony Tether, the director of DARPA, used the Grand Challenge to kick some ass in academic AI. The schools receiving funding from DARPA were told that if the private sector did better than they did, DARPA was turning off their grant money in AI. That's why the big schools put entire CS departments on the Grand Challenge.)
Seriously, though, I'd like to read more about Sears and the distribution solution. Wikipedia didn't really have anything. Any links?
No one seems to have described the "schedule system" in detail on line. It gets a brief mention in the Sears archives. Not much detail, though.
The obvious way to do fulfillment is to have order pickers, each with a few orders to pick, going through the storage aisles picking items, then delivering them to the packing and shipping area. That works if the inventory isn't too big. Safeway, for example, does on line shopping that way, with pickers running around retail grocery stories.
But the time to pick goes up with the size of the inventory, as the pickers have to travel further. The next idea is to divide up the orders by section, so that the items from each order are fanned out to different departments and picked by pickers in those departments. Then the picked partial orders have to be brought together for assembly. That creates a sorting problem, and as the volume goes up, the order assembly area tends to choke with work in progress.
The "schedule system" is a variant on picking by department. Orders are divided up by department at the front end of the process, where orders are read and pick slips produced. (Sears had to do this by hand in 1895, of course.) The pick slips specify a time slot and a bin number. Time slots were originally 45 minutes long. During a time slot, the pickers in each department work only on orders assigned to that time slot, picking items and putting them in small bins which travel on chutes and conveyors to the order assembly area, which has a receiving bin for each order being processed in that time slot. At the end of the time slot, the pickers switch to the next set of orders, even if something didn't get picked in time.
At the end of the time slot, all the bins in the order assembly area are replaced with empty bins, and the filled bins go to order checking and shipping. The original order is checked against the bin contents, anything missing is deducted from the charges and perhaps queued for another try on a later day, and the order is packed and shipped. Meanwhile, the next set of orders is being picked.
With this system, the pickers are only working on a moderate number of orders at a time, and only have to look within their own department. If they get behind during a time slot, some orders will be partially filled, and that gets caught in order assembly and retried. Order checking, packing, and shipping can be fanned out to as many assembly stations as necessary, and more stations can be staffed and brought on line if there's a backlog.
In the pre-computer era, this was a good way to coordinate an operation spread across acres of multi-story buildings. The order-checking phase of order assembly generates a ticket for each error, and those indicate what needed to be adjusted - too few pickers in one department, or too many out-of-stock reports from one department. It also provides a retry mechanism which doesn't stall out picking. This makes a huge operation manageable.
The biggest difference in modern fulfillment is that today, the inventory is known at the front end of order processing. If something is out of stock, no attempt is made to pick it. Manual systems have to carry more inventory to avoid pick fails. Systems today aren't tied to a rigid timetable, and there's a lot of bar coding and RFID tagging to track products and bins as they move around. But the fan-out-to-department and fan-in-to-assembly structure remains, since that's what gives the improvement from O(N*M) to O(N*log(M)). This is just like converting from a bubble sort to a merge sort.
This field is called "industrial engineering", which is about how to organize work so that it gets done efficiently. Anybody who supervises more than about 10 people needs to know the basics of this. Unfortunately, too many managers don't.
I've often wondered with McMaster's warehouses (what are now called fulfillment centers) are like.
Today, there are warehouses, fulfillment centers, and distribution centers, plus many other types of logistic facilities. A warehouse is mostly storage. A distribution center is an intermediate stop between suppliers and retail stores. A fulfillment center does order picking for customers.
Sears invented the fulfillment center between 1896 and 1906. Their mail order business was successful, but as the business grew the order handing process choked. They figured out how to do order fulfillment efficiently from a broad inventory in huge volume, without computers. They built a 40-acre facility in Chicago, called "The Works", which operated until 1993 when Sears finally exited catalog sales. The "schedule system" which did that is quite clever. In fulfillment, the obvious solution is O(N*M), where N is the number of orders and M is the number of orderable items. This does not scale well. Sears got that down to O(N*log(M)) and dominated mail order for most of a century.
I just bought four SanDisk USB drives, in original packaging, at Costco. I had to clean them of junk before using them. They even had autorun files and some kind of installer.
Send the guy an empty drive that's really empty. That's a real gift today.
What I'm seeing lately are water operators, IT people, and system integrators who are overzealous when it comes to connectivity and all the "neat" things that can be done remotely via technology.
Yes. Read "Access Your
Embedded Controller
with Ease through a
Web Server", from Texas Instruments, which ought to know better. "The designer should also make it as easy as possible to change the settings on
a piece of equipment, reconfigure its operation, or fine-tune the system. The
more intuitive and explicit that activity is, the more likely the result will be what
the operator desires. Losing the instruction manual can seriously impair the
user's operation of many systems."
What that paper describes is a family of embedded controllers with a web server in each controller and no security. What's wrong with this picture?
OK, Clueless guy visits an order fulfillment center. Not even a very interesting one. No Kiva robots like "soap.com", no incredibly fast processing on long orders for many tiny items like "digikey.com", no unusual outsourcing like UPS's laptop repair center.. Just an ordinary fulfillment center.
Maybe next he'll get out of Manhattan and visit a factory.
(Then again, "Pawn Stars" and "Storage Wars" are actual reality shows.)
It's going to be embarrassing if this turns out just to be a failed real estate development. China has those too.
Compare
this image from Google Maps. There's a nice "alignment target" in the middle of nowhere. It was supposed to be an industrial park near Dubai, but never got beyond road building. The China one looks like a project that never got beyond bulldozer stage.
Game programming is "cool". It also sucks as a job. Too many people think it's cool to be a game programmer and are willing to put up with miserable working hours and low pay. (Much the same is true of Hollywood, home of the actress/model/waitress. Median acting income of SAG members is below poverty line.)
"App" programming is in some ways worse. Most apps generate little or no revenue. Writing "apps" on speculation is a sucker bet. One Apple wants to encourage.
This isn't a problem for animators. Getting the models and motion right requires only the quality of a good graphics card. It's tweaking lighting, shaders, and post effects, the "look", that needs many repeats of full rendering. That's done by colorists and post people.
Microsoft pays a dividend. Not a big one, but about 3%. Compare Google, whose stock peaked in 2007, does not pay a dividend, and has a two-tier stock structure which prevents investors from controlling the company.
IBM stock is at an all-time high, and they pay a dividend. They're a mature company that performs like a growth company.
However, he does forget that phones need holes for audio and mic. And if you have any holes, what's the point?
Kyrocera, Casio, and Motorola have all solved that problem. They have waterproof speakers and mics, and their ruggedized phones can withstand immersion in water.
Kyocera makes a phone that can be run over by a car and still work. Casio makes a phone which can be dropped and submerged for hours and still work. Motorola makes a Android smartphone which is dustproof and water-resistant.
Apple makes a phone that breaks if it falls of a table.
There's no reason that Apple's products need to be so fragile.
The first generation of rugged phones was based on a rigid frame surrounded by rubber. That adds bulk. The future may be to give the printed circuit board, display, and battery some flexibility and make the case out of Kevlar, with sapphire-over-polycarbonate for the screen. Then the whole phone can flex a bit without damage.
The next step is to get rid of the holes. It's time for connectorless phones. Modern phones have Bluetooth, WiFi, and CDMA/GSM radios. Add inductive charging and you don't need any connectors. Then the whole unit can be watertight. This beats putting in water detectors that invalidate the warranty.
Microsoft's "strategic" moves over the past several years...
The difference is that Microsoft stays with something until they dominate the industry. The original XBox lost money from beginning to end. Now Microsoft's game operation is profitable, and they and Nintendo are on top, Sony is in trouble and Sega is forgotten.
Recently, a Microsoft exec made the comment that Microsoft is happy with Bing's progress. They gained 4% market share in search last year, and are now at 30%. Five more years and they might pass Google. Once Bing passes Google. they become the "must be on" ad network.
(Take that threat seriously. Twelve years ago, the top search engine was Lycos, "the catalog of the Internet". Where are they now? Myspace and Yahoo have tanked. Microsoft is still here. The one-product companies haven't done so well. And Google is a one-product company - ads are 96% of revenue. Despite many attempts, Google has never had a second winning product that generates serious revenue. The free stuff doesn't count.)
In the 1990s, Warner Music was the largest record company. Now they're third.
Warner Music is owned by a Russian oligarch, Leonard Blavatnik, who bought it last July. If Google had wanted Warner Music, they could have bought it then. It sold for $3 billion (actually only $320 million in cash plus the assumption of debt) a few months ago.
Google probably doesn't want to own a record company. It would be a distraction.
No, you've confused the instructions for first responders with the instructions for the dealerships doing post-crash repairs.
Yes, the maintenance instructions say to pull the battery pack. But that's part of repair, not a caution item to be done before even storing the wreckage. What's needed is a quick way for first responders and tow people to check for an internal battery short on a damaged vehicle. Maybe something like temperature-sensitive paint on the battery - "If this square is red, battery may explode; back off and contact HAZMAT". It's going to be a rare problem, but you want to know before moving the wreckage if there's a short.
And here they are, from the 2-hour training session for first responders to Chevy Volt accidents. It's necessary to open the trunk and cut 12V cables at two points with heavy wire cutters. The cut points are marked with yellow tape printed with a firefighter hat and wire cutters. Here is GM's official instruction sheet for this. There's also a battery disconnect switch inside the center console of the vehicle, where a big plug is turned and removed. That's the normal procedure for disconnecting power during service.
So that's the documented "protocol following the crash". That's what GM says to do, and what a first responder or a tow company would have done if they did everything right. It would have had little effect if a battery had an internal short.
The Prius, Civic, and Ford Escape each have completely different battery disconnection procedures. The first responder community is not happy about this. They want a standardized, easy to get at way to quickly disconnect the high voltage battery in an emergency.
The LEO series continued for a while. A few mergers later there was the English Electric LEO Marconi KDF9, an elegant stack machine closer to a Java VM than anything in current hardware. English Electric Leo Marconi was swallowed up by International Computers Limited, which was formed by the merger of Ferranti and International Computers and Tabulators, which had been formed by the merger of Powers-Samas and the British Tabulating Machine Company.
This mess was partially owned by the British government. ICL was never very successful, and its main customer was the British government. It ground on until 2002, until it was finally sold to Fujitsu.
Slashdot Mirror
Where would one purchase the air valves to make one of these ? Are these low cost items or expensive specialist things.
Ordinary solenoid on/off air valves are widely used, easily available and cheap. Grainger stocks them, and they show up in surplus stores. Proportional air valves are somewhat exotic items.
From the video, it looks like the robot's legs are controlled through simple on-off valves (notice the clicks and the jerky leg motion) but the trunk has full proportional control and feedback.
Precision pneumatics is effective but not used all that much. Most pneumatic actuators are a simple solenoid valve controlling air to a cylinder, with no feedback. Precision control of a pneumatic actuator requires a pair of proportional valves (which themselves require internal feedback on their valve spool) and a double acting air cylinder or other actuator with position and pressure feedback. There are at least two analog outputs (the valve solenoids) and five inputs (two valve positions, two pressures, and one actuator position) to deal with for each actuator. The control theory for this is complex. If you do all that, though, you get an actuator that behaves like a spring with adjustable spring constant, zero position, and damping. Which is what a muscle pair has.
The problem with pneumatic mobile robots, of course, is lugging the air compressor around.
We get at least one of these overhyped materials science articles each month. This time, someone has figured out how to deposit a garnet layer in a wafer fab. This is blown up into "photonic computing real soon now". It's not.
There's a lot of work in progress (PowerPoint) on optical on-chip interconnects. This is not "photonic computing". It's clusters of CPUs with a network of optical interconnects, all on one IC. The CPUs are still made of transistors. IBM has a very active research program in this area. But it's a long way from working. There are optical switching elements that work experimentally, but nothing ready for volume manufacturing yet. The optical interconnects themselves aren't considered to be the big problem.
So far, most of the proposed approaches involve un-buffered circuit switched networks. An optical connection is set up from CPU 1 to CPU 2 by electrical means, and then data is blasted across it. Circuit setup time is long compared to the data rate. So this is for long messages within a cluster, not cache synchronization. Think (inevitably) Beowulf cluster on a chip, not thousand-CPU shared memory microprocessors. The technology may also be useful as a network optical switch.
Short version: when this all works, servers get more densely packed.
Google's stock peaked in 2007. Google is making money, but it's keeping all of it. They don't pay a dividend. Investors would like to see some return on their investment.
Google still has the big problem that they only have one revenue source, ads. Nothing else they've tried makes much money. 96% is still from ads. Google's flings with telephony and social networking aren't contributing significant revenue. Being #1 in giving stuff away isn't paying off as well as expected.
Here's the DEA's list. Those marked as "List 1" are the most restricted. It's not that long a list. Iodine is the only chemical on List 1 that isn't particularly hazardous.
I know one business which has good results with Groupon. It's a riding school (horses). The Groupon lessons are an introduction to riding, with about an hour on the ground grooming and learning about horses, and an hour on the horse. Groupon insures that enough people sign up to make up a profitable group, and Groupon pays for no-shows in bad weather. This insures that each lesson brings in a known income. Some of the people discover they like horses and come back for more lessons. But even without that, it's profitable.
This is the right way to use Groupon - not as a traffic builder, but for something that's done in a group.
Also, not being powered by anthing more than the waves themselves, I am sure they are subject to the ocean currents.
Actually, they tend to stay within 50 meters of their programmed track. I've seen a presentation by the developers. They have a GPS and follow waypoints, and they have an Iridium satellite phone link. As long as there's a little wind to produce even light chop, they get enough energy from wave motion to overcome ocean currents.
Itanium was a joint Intel-HP project, remember? HP might well pay Intel to keep it alive.
The idea behind Itanium was that it had lots of new, different, patentable technology, so Intel didn't have to worry about clones. The problem was that it wasn't better technology. Just different.
Classic bad CPU architecture ideas of the "build it and they will come" variety:
In the spectrum of concurrency, shared-memory mulitprocessors with synchronized caches work, and clusters of powerful machines which communicate over networks work. Those are the extremes of the concurrency range. With the notable exception of graphics processors, no machine in the middle of that range has been a success. Such machines can be built, but are so hard to program they're always behind the classical architectures. The Cell in the PS/3 is the only example ever deployed in volume, and that nearly killed Sony.
This is a form of progress. However, one would expect the OS itself to have much stronger defenses against anything from the outside running at kernel level. There should be no way to get a "boot sector virus" onto a machine while running under an operating system.
I had the unfortunate experience of going through Stanford for a MSCS just before the "AI Winter". The "expert systems" (remember "expert systems"? ) profs were running the department. It was becoming clear that expert systems weren't going anywhere, and the faculty was in denial about that. They'd set up a 5-year "knowledge engineer" program, with a combination of computer science theory, philosophy, and psychological interviewing technique to write rules for expert systems (Where are those people now?) I had one exam where a question was "Does a rock have intentions"?
It took over a decade for the CS department to recover. After I graduated, the CS department was moved from Arts and Sciences, where it had been mostly autonomous, to Engineering, where it had adult supervision. It wasn't until the DARPA Grand Challenge forced Stanford to bring in machine learning people from CMU that the department really started moving forward again. Now they're making real progress.
(This is not well known, but Tony Tether, the director of DARPA, used the Grand Challenge to kick some ass in academic AI. The schools receiving funding from DARPA were told that if the private sector did better than they did, DARPA was turning off their grant money in AI. That's why the big schools put entire CS departments on the Grand Challenge.)
Seriously, though, I'd like to read more about Sears and the distribution solution. Wikipedia didn't really have anything. Any links?
No one seems to have described the "schedule system" in detail on line. It gets a brief mention in the Sears archives. Not much detail, though.
The obvious way to do fulfillment is to have order pickers, each with a few orders to pick, going through the storage aisles picking items, then delivering them to the packing and shipping area. That works if the inventory isn't too big. Safeway, for example, does on line shopping that way, with pickers running around retail grocery stories.
But the time to pick goes up with the size of the inventory, as the pickers have to travel further. The next idea is to divide up the orders by section, so that the items from each order are fanned out to different departments and picked by pickers in those departments. Then the picked partial orders have to be brought together for assembly. That creates a sorting problem, and as the volume goes up, the order assembly area tends to choke with work in progress.
The "schedule system" is a variant on picking by department. Orders are divided up by department at the front end of the process, where orders are read and pick slips produced. (Sears had to do this by hand in 1895, of course.) The pick slips specify a time slot and a bin number. Time slots were originally 45 minutes long. During a time slot, the pickers in each department work only on orders assigned to that time slot, picking items and putting them in small bins which travel on chutes and conveyors to the order assembly area, which has a receiving bin for each order being processed in that time slot. At the end of the time slot, the pickers switch to the next set of orders, even if something didn't get picked in time.
At the end of the time slot, all the bins in the order assembly area are replaced with empty bins, and the filled bins go to order checking and shipping. The original order is checked against the bin contents, anything missing is deducted from the charges and perhaps queued for another try on a later day, and the order is packed and shipped. Meanwhile, the next set of orders is being picked.
With this system, the pickers are only working on a moderate number of orders at a time, and only have to look within their own department. If they get behind during a time slot, some orders will be partially filled, and that gets caught in order assembly and retried. Order checking, packing, and shipping can be fanned out to as many assembly stations as necessary, and more stations can be staffed and brought on line if there's a backlog.
In the pre-computer era, this was a good way to coordinate an operation spread across acres of multi-story buildings. The order-checking phase of order assembly generates a ticket for each error, and those indicate what needed to be adjusted - too few pickers in one department, or too many out-of-stock reports from one department. It also provides a retry mechanism which doesn't stall out picking. This makes a huge operation manageable.
The biggest difference in modern fulfillment is that today, the inventory is known at the front end of order processing. If something is out of stock, no attempt is made to pick it. Manual systems have to carry more inventory to avoid pick fails. Systems today aren't tied to a rigid timetable, and there's a lot of bar coding and RFID tagging to track products and bins as they move around. But the fan-out-to-department and fan-in-to-assembly structure remains, since that's what gives the improvement from O(N*M) to O(N*log(M)). This is just like converting from a bubble sort to a merge sort.
This field is called "industrial engineering", which is about how to organize work so that it gets done efficiently. Anybody who supervises more than about 10 people needs to know the basics of this. Unfortunately, too many managers don't.
I've often wondered with McMaster's warehouses (what are now called fulfillment centers) are like.
Today, there are warehouses, fulfillment centers, and distribution centers, plus many other types of logistic facilities. A warehouse is mostly storage. A distribution center is an intermediate stop between suppliers and retail stores. A fulfillment center does order picking for customers.
Sears invented the fulfillment center between 1896 and 1906. Their mail order business was successful, but as the business grew the order handing process choked. They figured out how to do order fulfillment efficiently from a broad inventory in huge volume, without computers. They built a 40-acre facility in Chicago, called "The Works", which operated until 1993 when Sears finally exited catalog sales. The "schedule system" which did that is quite clever. In fulfillment, the obvious solution is O(N*M), where N is the number of orders and M is the number of orderable items. This does not scale well. Sears got that down to O(N*log(M)) and dominated mail order for most of a century.
I just bought four SanDisk USB drives, in original packaging, at Costco. I had to clean them of junk before using them. They even had autorun files and some kind of installer.
Send the guy an empty drive that's really empty. That's a real gift today.
What I'm seeing lately are water operators, IT people, and system integrators who are overzealous when it comes to connectivity and all the "neat" things that can be done remotely via technology.
Yes. Read "Access Your Embedded Controller with Ease through a Web Server", from Texas Instruments, which ought to know better. "The designer should also make it as easy as possible to change the settings on a piece of equipment, reconfigure its operation, or fine-tune the system. The more intuitive and explicit that activity is, the more likely the result will be what the operator desires. Losing the instruction manual can seriously impair the user's operation of many systems."
What that paper describes is a family of embedded controllers with a web server in each controller and no security. What's wrong with this picture?
OK, Clueless guy visits an order fulfillment center. Not even a very interesting one. No Kiva robots like "soap.com", no incredibly fast processing on long orders for many tiny items like "digikey.com", no unusual outsourcing like UPS's laptop repair center.. Just an ordinary fulfillment center.
Maybe next he'll get out of Manhattan and visit a factory.
(Then again, "Pawn Stars" and "Storage Wars" are actual reality shows.)
It's going to be embarrassing if this turns out just to be a failed real estate development. China has those too.
Compare this image from Google Maps. There's a nice "alignment target" in the middle of nowhere. It was supposed to be an industrial park near Dubai, but never got beyond road building. The China one looks like a project that never got beyond bulldozer stage.
Game programming is "cool". It also sucks as a job. Too many people think it's cool to be a game programmer and are willing to put up with miserable working hours and low pay. (Much the same is true of Hollywood, home of the actress/model/waitress. Median acting income of SAG members is below poverty line.)
"App" programming is in some ways worse. Most apps generate little or no revenue. Writing "apps" on speculation is a sucker bet. One Apple wants to encourage.
This isn't a problem for animators. Getting the models and motion right requires only the quality of a good graphics card. It's tweaking lighting, shaders, and post effects, the "look", that needs many repeats of full rendering. That's done by colorists and post people.
Microsoft pays a dividend. Not a big one, but about 3%. Compare Google, whose stock peaked in 2007, does not pay a dividend, and has a two-tier stock structure which prevents investors from controlling the company.
IBM stock is at an all-time high, and they pay a dividend. They're a mature company that performs like a growth company.
However, he does forget that phones need holes for audio and mic. And if you have any holes, what's the point?
Kyrocera, Casio, and Motorola have all solved that problem. They have waterproof speakers and mics, and their ruggedized phones can withstand immersion in water.
Kyocera makes a phone that can be run over by a car and still work. Casio makes a phone which can be dropped and submerged for hours and still work. Motorola makes a Android smartphone which is dustproof and water-resistant. Apple makes a phone that breaks if it falls of a table. There's no reason that Apple's products need to be so fragile.
The first generation of rugged phones was based on a rigid frame surrounded by rubber. That adds bulk. The future may be to give the printed circuit board, display, and battery some flexibility and make the case out of Kevlar, with sapphire-over-polycarbonate for the screen. Then the whole phone can flex a bit without damage.
The next step is to get rid of the holes. It's time for connectorless phones. Modern phones have Bluetooth, WiFi, and CDMA/GSM radios. Add inductive charging and you don't need any connectors. Then the whole unit can be watertight. This beats putting in water detectors that invalidate the warranty.
Microsoft's "strategic" moves over the past several years...
The difference is that Microsoft stays with something until they dominate the industry. The original XBox lost money from beginning to end. Now Microsoft's game operation is profitable, and they and Nintendo are on top, Sony is in trouble and Sega is forgotten.
Recently, a Microsoft exec made the comment that Microsoft is happy with Bing's progress. They gained 4% market share in search last year, and are now at 30%. Five more years and they might pass Google. Once Bing passes Google. they become the "must be on" ad network.
(Take that threat seriously. Twelve years ago, the top search engine was Lycos, "the catalog of the Internet". Where are they now? Myspace and Yahoo have tanked. Microsoft is still here. The one-product companies haven't done so well. And Google is a one-product company - ads are 96% of revenue. Despite many attempts, Google has never had a second winning product that generates serious revenue. The free stuff doesn't count.)
In the 1990s, Warner Music was the largest record company. Now they're third. Warner Music is owned by a Russian oligarch, Leonard Blavatnik, who bought it last July. If Google had wanted Warner Music, they could have bought it then. It sold for $3 billion (actually only $320 million in cash plus the assumption of debt) a few months ago.
Google probably doesn't want to own a record company. It would be a distraction.
No, you've confused the instructions for first responders with the instructions for the dealerships doing post-crash repairs.
Yes, the maintenance instructions say to pull the battery pack. But that's part of repair, not a caution item to be done before even storing the wreckage. What's needed is a quick way for first responders and tow people to check for an internal battery short on a damaged vehicle. Maybe something like temperature-sensitive paint on the battery - "If this square is red, battery may explode; back off and contact HAZMAT". It's going to be a rare problem, but you want to know before moving the wreckage if there's a short.
And here they are, from the 2-hour training session for first responders to Chevy Volt accidents. It's necessary to open the trunk and cut 12V cables at two points with heavy wire cutters. The cut points are marked with yellow tape printed with a firefighter hat and wire cutters. Here is GM's official instruction sheet for this. There's also a battery disconnect switch inside the center console of the vehicle, where a big plug is turned and removed. That's the normal procedure for disconnecting power during service.
So that's the documented "protocol following the crash". That's what GM says to do, and what a first responder or a tow company would have done if they did everything right. It would have had little effect if a battery had an internal short.
The Prius, Civic, and Ford Escape each have completely different battery disconnection procedures. The first responder community is not happy about this. They want a standardized, easy to get at way to quickly disconnect the high voltage battery in an emergency.
The LEO series continued for a while. A few mergers later there was the English Electric LEO Marconi KDF9, an elegant stack machine closer to a Java VM than anything in current hardware. English Electric Leo Marconi was swallowed up by International Computers Limited, which was formed by the merger of Ferranti and International Computers and Tabulators, which had been formed by the merger of Powers-Samas and the British Tabulating Machine Company.
This mess was partially owned by the British government. ICL was never very successful, and its main customer was the British government. It ground on until 2002, until it was finally sold to Fujitsu.