I can tell you that Hackworth's glowing assessment of MILES (which I have heard of before, although I have never read his book) is not at all conform with my experiences or those of many other soldiers I knew/know.
First, mod parent up.
Second, the MILES enthusiast I've read was not Hackworth (whose career predated MILES) but Lt. Gen. Daniel Bolger. In his junior officer days, he wrote a book ("Dragons at War") about his unit's experience at the National Training Center.
It turned out that a large number of soldiers never fired their weapons.
That "research", by S.L.A. Marshall, has been discredited. Read Col. Dave Hackworth's "About Face". Hackworth was a very good infantry commander and worked with S.L.A. Marshall in Vietnam, where Marshall was a journalist. Marshall made up a lot of what he wrote. His work reads like he was there when, most of the time, he wasn't.
The big breakthrough in training was in the late 1970s, when the U.S. Army developed the Multiple Integrated Laser Engagement System (MILES). This is the militarized version of laser tag. For the first time, soldiers fired their weapons during force-on-force exercises and the hits and misses were tallied. Previously, everybody made lots of noise with blanks and umpires randomly decided who died, like dungeon masters. With MILES, troops had to aim to hit in a combat training situation, because their performance was being measured. They got a lot better at it, and US infantry became much more effective as a result.
The problem was not soldiers failing to fire their weapons. It was firing but not hitting the enemy.
That robot runs QNX. So does BigDog. When the process absolutely, positively has to get control when it's supposed to, QNX is the way to go. It offers a POSIX API, but underneath, it's a tiny message-passing microkernel.
If only the company behind QNX weren't so screwed up. QNX used to be a standalone company. Then one of the founders died, and the company was sold to Harmon, which is mostly an audio company. They then focused on car dashboard systems. Harmon had no clue what to do with a real-time OS company, and sold off QNX to Research in Motion, the Blackberry company.
QNX used to be closed source. Then they issued a free version around 2001 and opened up some of the code. Then they closed the source around 2004. Then they made the whole thing open source (free for noncommercial use) around 2008. Then in 2010, they closed the source again and cancelled the free program. So nobody ports free software to QNX any more, and the developer community is fed up. There's still a non-commercial license, so you can still play with QNX, but few people do.
It used to be possible to run QNX on the desktop, which is useful if you're developing for QNX. But nobody has ported a browser since 2002 or so.
If you're in Northern California, you have the option of switching to Sonic.net. Sonic is an independent ISP which has grandfathered rights to lease AT&T DSL lines at favorable rates. They back-haul your DSL link to Santa Rosa, CA, and then connect to the Internet via Cable and Wireless.They have no usage cap and no intention of adding one. Sonic has been slightly more expensive than AT&T until recently. But if you're faced with AT&T's bandwidth cap, they can now be cheaper.
Sonic just sells a data pipe. They don't sell any content over their DSL lines, so they have no incentive to force you into some "entertainment package". (They do resell DirectTV, but that's via a satellite dish and is mostly a sideline for their rural customers.
There's no "packet inspection" nonsense with Sonic. No caching. No funny DNS rerouting. No custom browser. They just pipe through the bits you send and receive. You pay for bandwidth (and it's not "up to 6 mbps", it's "3.0mbsp to 6.0 mbps download, 512kbps to 768kbps upload."). My own line at in that tier measures at about 4.1mbps.
They also have 20mbps and 40mbps services, but they're available only in limited areas.
Sonic also has better policies than AT&T. "Sonic.net, Inc. functions as a common carrier and does not censor." They don't require arbitration; you can go to Small Claims Court if you have to.
Specialized CPU elements have been tried. The track record to date is roughly this:
Floating point - huge win.
Graphics support - huge win, mostly because graphics parallelizes very well.
Multiple parallel integer operations on bytes of a word - win, but not not a huge win.
Hardware subroutine call support, such as register saving - marginal to negative. Many CPUs have had fancy CALL instructions that were slower than writing out the register saves, etc.
Hardware call gates - in x86, not used much.
Hardware context switching - in some CPUs, not used much.
Array instructions - once popular at the supercomputer level, now rare.
Compression/decompression (MPEG, etc.) - reasonable win, but may be more useful as part of a graphics device than a CPU.
List manipulation, LISP support, Java stack support - usually a lose over straight code.
Explicit parallelism, as in Itanium - usually a lose over superscalar machines.
Filter-type operations (Fourier transform, convolution, wavelets, etc.) - very successful, but usually more useful as part of a signal processing part than as part of a CPU.
Inter-CPU communication - useful, but very hard to get right. DMA to shared memory (as in the Cell) seems to be the wrong way. Infiniband, which is message passing hardware, is useful but so far only seen in high end machines.
Hardware CPU dispatching - has potential if connected to "hyperthreading", but historically not too successful.
Capability-based machines. - successful a few times (IBM System/38 being the best example) but never made it in microprocessors.
A lot of things which you might think would help turn out to be a lose. Superscalar machines and optimizing compilers do a good job on inner loops today. (If it's not in an inner loop, it's probably not being executed enough to justify custom hardware.)
Right. The current record for an experimental solar cell technology is around 42%. Mainstream commercial products run 12 to 20%. That's up from 6% in the original Bell Labs solar cells in the 1950s. Single-layer cells have a theoretical limit at 34%, but multi-layer cells can beat that somewhat.
Easy way to boost your conviction rate, with very little man power, and the people will love you for protecting the children.
That's the real reason. Real online fraud investigations are hard. The trail may lead through multiple countries, and serious investigative work is required. Even then, the investigation may dead-end.
Still, devoting 41% of FBI computer resources to kiddie porn, compared to 4% on online fraud, is way out of line. That's probably contributed to the growth in online fraud.
Joel-on-Software Spolsky promotes himself as an authority on software development, but he only runs a tiny company that makes applications for a relatively simple problem. It's still a tiny company, after over a decade of operation. I'd rather hear from the people who managed the software for Voyager. Or the vehicle stabilization system for a modern car. Or the radio inside the iPhone. (I know the guy who headed that team; he waited until the iPhone shipped, and then quit Apple in disgust with having to work for Steve Jobs.) Or the file system that keeps Google working even when machines fail.
That's so Apple. A case designer needs some basic info about the outside dimensions, like height, width, thickness, and corner radius. Maybe a bit of info about areas that need to to be uncovered for connector or button access. Only Apple would consider those to be significant trade secrets. Especially since this is Model #2 of the product, which is probably going to be a lot like Model #1.
No, that's Facebook. Google was like that once, but that was years ago.
The ad side of the business has been slowly taking over. A big fraction of their head count is in ad sales now. 97% of revenue is still from ads. All those smart people on the engineering side haven't been able to come up with a second big money-making product. That's Google's killer problem.
Google's stock peaked in 2007, around 714, and it's 538 today. Google is no longer a growth company. They're quite profitable, but since they don't pay a dividend, that does nothing for the stockholders.
I just posed the same question in another topic, and wrote this:
WiFi routers should have the option of putting the air link on the outside of the local firewall. Actually, it would make sense if, by default, open WiFi links gave guest access to the outside Internet world, but not the inside LAN world, while encrypted links offered access to the inside world. This allows opening up guest access without exposing local servers and Windows shares.
A router should support both modes simultaneously, offering itself as two access points. Encrypted links should have higher packet priority over nonencrypted links, so that guest access can't starve out authorized users.
This seems obvious enough that some routers probably implement it already. Anyone know of one?
Right. I have access to all that stuff at TechShop. Costs me $100 a month. Cheaper than some cable TV packages.
Now, if they had a water jet cutter (which TechShop is installing right now in San Francisco) or a 5-axis CNC mill with HyperMill software (TechShop only has a 3-axis plus a rotary table, and midrange Vectrix software) or a full surface mount assembly setup (Hacker Dojo has one, but nobody uses it) I'd be impressed.
WiFi routers should have the option of putting the air link on the outside of the local firewall. Actually, it would make sense if, by default, open WiFi links gave guest access to the outside Internet world, but not the inside LAN world, while encrypted links offered access to the inside world. This allows opening up guest access without exposing local servers and Windows shares.
A router should support both modes simultaneously, offering itself as two access points. Encrypted links should have higher packet priority over nonencrypted links, so that guest access can't starve out authorized users.
This seems obvious enough that some routers probably implement it already. Anyone know of one?
It's not a bad idea, provided that people are actually willing to work for you for free. Usually, they aren't. It's been tried before, for spam filtering, but the reviewers were overwhelmed. (You'd get random messages containing spam which you have to rate. Right.) This approach sometimes works when the number of items to rate is much smaller than the number of raters, and when the user has to read the thing they're rating anyway, as with Slashdot.
Also, the annoyance level of a message may depend on the recipient. "Hey. let's do lunch at Pizza Hut" can produce reactions ranging from "Great, see you there!" to "Who the hell is this?" to "Another Pizza Hut spam." Then there's the problem of the confidentiality of messages.
The Slashdot story misreports the data, as usual.
The actual report says that 36% of the agents who were assigned to national security related cyber investigations self-reported that they did not have the necessary expertise for the job they were doing.
And those are the national-security related cases, which the FBI considers to be the most important category. It's probably worse at the regular computer-related crime level.
They're trying. The FBI actually runs agents through "A+" training, and "Linux for Law Enforcement". After 5 years as an FBI agent on the "cyber" side, agents should be able to configure a Linux kernel and have an in-depth knowledge of the Windows registry. Those agents also have to learn all the regular FBI agent skills.
The report points out that 41% of the FBI's "cyber" force is tied up investigating child pornography, while only 4% work on Internet fraud. That's why they're doing so badly on online crime.
you might as well just put all the troll flagged accounts into the same bubble, so they could see each other's posts, but they would all be invisible to everyone else.
That was seriously considered for an early MMORPG. Annoying players would be dumped into a dungeon level full of NPCs and other annoying players, where they could flame war and player kill as long as they wanted, without bothering anybody else. It wasn't done due to resource constraints, but it remains a good idea.
Wiretapping is just too easy now. It used to be quite difficult. Before electronic central offices, wiretapping required either a tap near the phone end of the circuit, or wiring to the appropriate circuit at the central office's main distributing frame. Telcos charged law enforcement for central office taps. Guliani writes, in his book about his days prosecuting the New York Mafia, that they were paying about a million dollars a year to New York Telephone for wiretaps, which were charged as private line extensions. On one occasion, the FBI didn't pay a bill on time, and the billing system billed the other party on the circuit, the one being wiretapped. This was a major motivation for CALEA.
In the crossbar era, it was possible to use the Automatic Line Insulation Test (ALIT) gear for wiretapping. This was a system that automatically tested each line every night, applying a test voltage and measuring leakage between the lines and to ground. Lines could also be tested remotely, on request, and the gear allowed listening in. But a central office would typically only have two sets of ALIT gear (three racks each), and using it for wiretapping interfered with routine maintenance. The FBI could sometimes get access to ALIT gear, but not local law enforcement. Only for short periods, too; the telco would keep demanding their test gear back.
All this was such a headache that wiretaps weren't used much. Now, all CO gear has remote wiretapping of large numbers of lines on demand at all times. It's also much easier to record and to monitor the recordings. Orwell would be so impressed.
Telautographs were used well into the 1970s. You write or draw at one end, and the pen at the other end follows. That's all they do. Railroads used to use them for train orders, which had to be signed. They have zero relationship with the iPad. (The Newton, which had pen input, maybe.)
Early telautographs suffered from the usual problem of pre-vacuum tube electrical devices - they needed signal amplification. That was really hard to do before tubes, let alone transistors. There's a long history of early amplifying devices, all of them awful. Grey's patent shows one mechanical approach. Later (tube) versions used analog audio tones, so they could transmit over phone lines.
See "The Future of the PhD. Basically, the entire world is producing more PhDs than jobs for them.
In some countries, including the United States and Japan, people who have trained at great length and expense to be researchers confront a dwindling number of academic jobs, and an industrial sector unable to take up the slack. Supply has outstripped demand and, although few PhD holders end up unemployed, it is not clear that spending years securing this high-level qualification is worth it for a job as, for example, a high-school teacher. In other countries, such as China and India, the economies are developing fast enough to use all the PhDs they can crank out, and more â" but the quality of the graduates is not consistent. Only a few nations, including Germany, are successfully tackling the problem by redefining the PhD as training for high-level positions in careers outside academia.
Germany seems to do well on labor issues. Not just for academics, either. The country has an organized apprenticeship system turning out good technicians.
Hacker Dojo isn'[t a "hacker space". It's more like an incubator for startups. I've taken a machine learning class there, where we did homework from the Stanford class and were taught by a quant from Blackstone Capital. During the day, there are people working quietly with laptops, and a few people rent offices there. It's more like a Starbucks. They have a surface-mount workstation, microscope and all, but I've never seen it used.
TechShop in Menlo Park is more like a hacker space. Most of the members have engineering or technician backgrounds (this is Silicon Valley, after all), and the big, serious machine tools get heavy use. During the day, there are people who are sent there by their employers to get machining done.
TechShop in San Francisco is still trying to find its niche. Mostly I see people cutting art objects on the laser cutters. The big machine tools aren't used much; looking at the equipment sign up calendar, nobody has signed up for the big manual mills all week. The CNC mill (a Tormach) is reasonably busy. SF has enough sewing and embroidery equipment for a sweatshop, but it's not being used.
Both locations have electronics tools, but they don't get heavy use. Sometimes someone will be building some industrial electronics equipment, but that's not a hacker project.
San Francisco does have a "hacker space", Noisebridge. They do everything from Vegan cooking lessons to lockpicking to a high-altitude balloon space program. That's more like NYC Resistors.
We need a forced migration of the mobile world to IPv6. The mobile people have the advantage that the carrier controls both the phone and the ISP, so they can upgrade them compatibly. Most of the growth is in mobile, after all.
Many people running medical marijuana dispensaries aren't used to running legitimate businesses.
Right. We'd be better off if the stuff was just made a class II or III prescription drug and dispensed at real pharmacies. Some people have a medical need for it, but nowhere as many as the number of people wanting it.
Having lived near San Francisco for years, I'm not impressed with the stoner community. There are way too many burnout cases on the streets.
In 1970, engineering and law paid about the same. The IEEE tracks this. Dentistry paid better. Real estate sales paid worse, on a par with auto sales.
What happened? Something few want to admit. Major parts of science are mined out. The return on investment from pure research has dropped since the 1970s. There was a long period when a small team might produce something like the tungsten-filament light bulb or the transistor. Now, it takes an army of researchers to get a minor improvement. That's why the big corporate research labs went into decline in the 1980s and are now mostly gone. Notice that high-energy physics hasn't produced much in the way of products in half a century. (Low-energy physics has produced substantial results, though.) Semiconductors have made huge progress, but huge resources were required to accomplish that. A modern wafer fab costs billions. The payoff for cleverness has declined, and salaries have declined accordingly.
(Biology is still making real progress, and has plenty of work ahead. Outside bio, though, things are slow.)
Slashdot Mirror
Why should it take 512 megabytes of RAM to display a page of HTML? Really.
I can tell you that Hackworth's glowing assessment of MILES (which I have heard of before, although I have never read his book) is not at all conform with my experiences or those of many other soldiers I knew/know.
First, mod parent up.
Second, the MILES enthusiast I've read was not Hackworth (whose career predated MILES) but Lt. Gen. Daniel Bolger. In his junior officer days, he wrote a book ("Dragons at War") about his unit's experience at the National Training Center.
It turned out that a large number of soldiers never fired their weapons.
That "research", by S.L.A. Marshall, has been discredited. Read Col. Dave Hackworth's "About Face". Hackworth was a very good infantry commander and worked with S.L.A. Marshall in Vietnam, where Marshall was a journalist. Marshall made up a lot of what he wrote. His work reads like he was there when, most of the time, he wasn't.
The big breakthrough in training was in the late 1970s, when the U.S. Army developed the Multiple Integrated Laser Engagement System (MILES). This is the militarized version of laser tag. For the first time, soldiers fired their weapons during force-on-force exercises and the hits and misses were tallied. Previously, everybody made lots of noise with blanks and umpires randomly decided who died, like dungeon masters. With MILES, troops had to aim to hit in a combat training situation, because their performance was being measured. They got a lot better at it, and US infantry became much more effective as a result.
The problem was not soldiers failing to fire their weapons. It was firing but not hitting the enemy.
That robot runs QNX. So does BigDog. When the process absolutely, positively has to get control when it's supposed to, QNX is the way to go. It offers a POSIX API, but underneath, it's a tiny message-passing microkernel.
If only the company behind QNX weren't so screwed up. QNX used to be a standalone company. Then one of the founders died, and the company was sold to Harmon, which is mostly an audio company. They then focused on car dashboard systems. Harmon had no clue what to do with a real-time OS company, and sold off QNX to Research in Motion, the Blackberry company.
QNX used to be closed source. Then they issued a free version around 2001 and opened up some of the code. Then they closed the source around 2004. Then they made the whole thing open source (free for noncommercial use) around 2008. Then in 2010, they closed the source again and cancelled the free program. So nobody ports free software to QNX any more, and the developer community is fed up. There's still a non-commercial license, so you can still play with QNX, but few people do.
It used to be possible to run QNX on the desktop, which is useful if you're developing for QNX. But nobody has ported a browser since 2002 or so.
If you're in Northern California, you have the option of switching to Sonic.net. Sonic is an independent ISP which has grandfathered rights to lease AT&T DSL lines at favorable rates. They back-haul your DSL link to Santa Rosa, CA, and then connect to the Internet via Cable and Wireless.They have no usage cap and no intention of adding one. Sonic has been slightly more expensive than AT&T until recently. But if you're faced with AT&T's bandwidth cap, they can now be cheaper.
Sonic just sells a data pipe. They don't sell any content over their DSL lines, so they have no incentive to force you into some "entertainment package". (They do resell DirectTV, but that's via a satellite dish and is mostly a sideline for their rural customers.
There's no "packet inspection" nonsense with Sonic. No caching. No funny DNS rerouting. No custom browser. They just pipe through the bits you send and receive. You pay for bandwidth (and it's not "up to 6 mbps", it's "3.0mbsp to 6.0 mbps download, 512kbps to 768kbps upload."). My own line at in that tier measures at about 4.1mbps.
They also have 20mbps and 40mbps services, but they're available only in limited areas.
Sonic also has better policies than AT&T. "Sonic.net, Inc. functions as a common carrier and does not censor." They don't require arbitration; you can go to Small Claims Court if you have to.
Specialized CPU elements have been tried. The track record to date is roughly this:
A lot of things which you might think would help turn out to be a lose. Superscalar machines and optimizing compilers do a good job on inner loops today. (If it's not in an inner loop, it's probably not being executed enough to justify custom hardware.)
Right. The current record for an experimental solar cell technology is around 42%. Mainstream commercial products run 12 to 20%. That's up from 6% in the original Bell Labs solar cells in the 1950s. Single-layer cells have a theoretical limit at 34%, but multi-layer cells can beat that somewhat.
Easy way to boost your conviction rate, with very little man power, and the people will love you for protecting the children.
That's the real reason. Real online fraud investigations are hard. The trail may lead through multiple countries, and serious investigative work is required. Even then, the investigation may dead-end.
Still, devoting 41% of FBI computer resources to kiddie porn, compared to 4% on online fraud, is way out of line. That's probably contributed to the growth in online fraud.
Joel-on-Software Spolsky promotes himself as an authority on software development, but he only runs a tiny company that makes applications for a relatively simple problem. It's still a tiny company, after over a decade of operation. I'd rather hear from the people who managed the software for Voyager. Or the vehicle stabilization system for a modern car. Or the radio inside the iPhone. (I know the guy who headed that team; he waited until the iPhone shipped, and then quit Apple in disgust with having to work for Steve Jobs.) Or the file system that keeps Google working even when machines fail.
That's so Apple. A case designer needs some basic info about the outside dimensions, like height, width, thickness, and corner radius. Maybe a bit of info about areas that need to to be uncovered for connector or button access. Only Apple would consider those to be significant trade secrets. Especially since this is Model #2 of the product, which is probably going to be a lot like Model #1.
Google's a cult of big kids.
No, that's Facebook. Google was like that once, but that was years ago.
The ad side of the business has been slowly taking over. A big fraction of their head count is in ad sales now. 97% of revenue is still from ads. All those smart people on the engineering side haven't been able to come up with a second big money-making product. That's Google's killer problem.
Google's stock peaked in 2007, around 714, and it's 538 today. Google is no longer a growth company. They're quite profitable, but since they don't pay a dividend, that does nothing for the stockholders.
I just posed the same question in another topic, and wrote this:
WiFi routers should have the option of putting the air link on the outside of the local firewall. Actually, it would make sense if, by default, open WiFi links gave guest access to the outside Internet world, but not the inside LAN world, while encrypted links offered access to the inside world. This allows opening up guest access without exposing local servers and Windows shares.
A router should support both modes simultaneously, offering itself as two access points. Encrypted links should have higher packet priority over nonencrypted links, so that guest access can't starve out authorized users.
This seems obvious enough that some routers probably implement it already. Anyone know of one?
Right. I have access to all that stuff at TechShop. Costs me $100 a month. Cheaper than some cable TV packages.
Now, if they had a water jet cutter (which TechShop is installing right now in San Francisco) or a 5-axis CNC mill with HyperMill software (TechShop only has a 3-axis plus a rotary table, and midrange Vectrix software) or a full surface mount assembly setup (Hacker Dojo has one, but nobody uses it) I'd be impressed.
WiFi routers should have the option of putting the air link on the outside of the local firewall. Actually, it would make sense if, by default, open WiFi links gave guest access to the outside Internet world, but not the inside LAN world, while encrypted links offered access to the inside world. This allows opening up guest access without exposing local servers and Windows shares.
A router should support both modes simultaneously, offering itself as two access points. Encrypted links should have higher packet priority over nonencrypted links, so that guest access can't starve out authorized users.
This seems obvious enough that some routers probably implement it already. Anyone know of one?
OP just reinvented the jury system.
It's not a bad idea, provided that people are actually willing to work for you for free. Usually, they aren't. It's been tried before, for spam filtering, but the reviewers were overwhelmed. (You'd get random messages containing spam which you have to rate. Right.) This approach sometimes works when the number of items to rate is much smaller than the number of raters, and when the user has to read the thing they're rating anyway, as with Slashdot.
Also, the annoyance level of a message may depend on the recipient. "Hey. let's do lunch at Pizza Hut" can produce reactions ranging from "Great, see you there!" to "Who the hell is this?" to "Another Pizza Hut spam." Then there's the problem of the confidentiality of messages.
Probably not a winner here.
The Slashdot story misreports the data, as usual. The actual report says that 36% of the agents who were assigned to national security related cyber investigations self-reported that they did not have the necessary expertise for the job they were doing.
And those are the national-security related cases, which the FBI considers to be the most important category. It's probably worse at the regular computer-related crime level.
They're trying. The FBI actually runs agents through "A+" training, and "Linux for Law Enforcement". After 5 years as an FBI agent on the "cyber" side, agents should be able to configure a Linux kernel and have an in-depth knowledge of the Windows registry. Those agents also have to learn all the regular FBI agent skills.
The report points out that 41% of the FBI's "cyber" force is tied up investigating child pornography, while only 4% work on Internet fraud. That's why they're doing so badly on online crime.
you might as well just put all the troll flagged accounts into the same bubble, so they could see each other's posts, but they would all be invisible to everyone else.
That was seriously considered for an early MMORPG. Annoying players would be dumped into a dungeon level full of NPCs and other annoying players, where they could flame war and player kill as long as they wanted, without bothering anybody else. It wasn't done due to resource constraints, but it remains a good idea.
Wiretapping is just too easy now. It used to be quite difficult. Before electronic central offices, wiretapping required either a tap near the phone end of the circuit, or wiring to the appropriate circuit at the central office's main distributing frame. Telcos charged law enforcement for central office taps. Guliani writes, in his book about his days prosecuting the New York Mafia, that they were paying about a million dollars a year to New York Telephone for wiretaps, which were charged as private line extensions. On one occasion, the FBI didn't pay a bill on time, and the billing system billed the other party on the circuit, the one being wiretapped. This was a major motivation for CALEA.
In the crossbar era, it was possible to use the Automatic Line Insulation Test (ALIT) gear for wiretapping. This was a system that automatically tested each line every night, applying a test voltage and measuring leakage between the lines and to ground. Lines could also be tested remotely, on request, and the gear allowed listening in. But a central office would typically only have two sets of ALIT gear (three racks each), and using it for wiretapping interfered with routine maintenance. The FBI could sometimes get access to ALIT gear, but not local law enforcement. Only for short periods, too; the telco would keep demanding their test gear back.
All this was such a headache that wiretaps weren't used much. Now, all CO gear has remote wiretapping of large numbers of lines on demand at all times. It's also much easier to record and to monitor the recordings. Orwell would be so impressed.
Telautographs were used well into the 1970s. You write or draw at one end, and the pen at the other end follows. That's all they do. Railroads used to use them for train orders, which had to be signed. They have zero relationship with the iPad. (The Newton, which had pen input, maybe.)
Early telautographs suffered from the usual problem of pre-vacuum tube electrical devices - they needed signal amplification. That was really hard to do before tubes, let alone transistors. There's a long history of early amplifying devices, all of them awful. Grey's patent shows one mechanical approach. Later (tube) versions used analog audio tones, so they could transmit over phone lines.
I've been looking for one. I sometimes restore antique Teletype equipment, especially pre-1930 machines.
See "The Future of the PhD. Basically, the entire world is producing more PhDs than jobs for them.
In some countries, including the United States and Japan, people who have trained at great length and expense to be researchers confront a dwindling number of academic jobs, and an industrial sector unable to take up the slack. Supply has outstripped demand and, although few PhD holders end up unemployed, it is not clear that spending years securing this high-level qualification is worth it for a job as, for example, a high-school teacher. In other countries, such as China and India, the economies are developing fast enough to use all the PhDs they can crank out, and more â" but the quality of the graduates is not consistent. Only a few nations, including Germany, are successfully tackling the problem by redefining the PhD as training for high-level positions in careers outside academia.
Germany seems to do well on labor issues. Not just for academics, either. The country has an organized apprenticeship system turning out good technicians.
Hacker Dojo isn'[t a "hacker space". It's more like an incubator for startups. I've taken a machine learning class there, where we did homework from the Stanford class and were taught by a quant from Blackstone Capital. During the day, there are people working quietly with laptops, and a few people rent offices there. It's more like a Starbucks. They have a surface-mount workstation, microscope and all, but I've never seen it used.
TechShop in Menlo Park is more like a hacker space. Most of the members have engineering or technician backgrounds (this is Silicon Valley, after all), and the big, serious machine tools get heavy use. During the day, there are people who are sent there by their employers to get machining done.
TechShop in San Francisco is still trying to find its niche. Mostly I see people cutting art objects on the laser cutters. The big machine tools aren't used much; looking at the equipment sign up calendar, nobody has signed up for the big manual mills all week. The CNC mill (a Tormach) is reasonably busy. SF has enough sewing and embroidery equipment for a sweatshop, but it's not being used.
Both locations have electronics tools, but they don't get heavy use. Sometimes someone will be building some industrial electronics equipment, but that's not a hacker project.
San Francisco does have a "hacker space", Noisebridge. They do everything from Vegan cooking lessons to lockpicking to a high-altitude balloon space program. That's more like NYC Resistors.
We need a forced migration of the mobile world to IPv6. The mobile people have the advantage that the carrier controls both the phone and the ISP, so they can upgrade them compatibly. Most of the growth is in mobile, after all.
"ChinaMobile" seems to be doing this already.
Many people running medical marijuana dispensaries aren't used to running legitimate businesses.
Right. We'd be better off if the stuff was just made a class II or III prescription drug and dispensed at real pharmacies. Some people have a medical need for it, but nowhere as many as the number of people wanting it.
Having lived near San Francisco for years, I'm not impressed with the stoner community. There are way too many burnout cases on the streets.
This is either a phony message or a lie by Steve Jobs. Both are possible.
In 1970, engineering and law paid about the same. The IEEE tracks this. Dentistry paid better. Real estate sales paid worse, on a par with auto sales.
What happened? Something few want to admit. Major parts of science are mined out. The return on investment from pure research has dropped since the 1970s. There was a long period when a small team might produce something like the tungsten-filament light bulb or the transistor. Now, it takes an army of researchers to get a minor improvement. That's why the big corporate research labs went into decline in the 1980s and are now mostly gone. Notice that high-energy physics hasn't produced much in the way of products in half a century. (Low-energy physics has produced substantial results, though.) Semiconductors have made huge progress, but huge resources were required to accomplish that. A modern wafer fab costs billions. The payoff for cleverness has declined, and salaries have declined accordingly.
(Biology is still making real progress, and has plenty of work ahead. Outside bio, though, things are slow.)