There are some famous examples of this working. But only because the parents had time, money, and high standards.
One of the Rockefellers, the son of John D., wrote that when he was a kid, his father gave him an allowance. He was required to keep a proper set of double-entry books on how he spent it, and the books were audited by an accountant. He didn't get the next allowance payment until the books balanced.
Henry Ford II was promised a car for some birthday. On the appointed day, he was taken out to a garage, and there was the car - totally dissembled with all the component parts laid out. A full set of tools was supplied. Eventually, he did get the car assembled and running.
One of my sites, "downside.com", has a MySQL database of every Securities and Exchange Commission filing since 2000. There's a cron job that updates the database from the SEC site every day at 4 AM. This used to run on EZPublishing, until they gave up hosting to focus on "permission e-mail" (really). It's now running on an $14.95 "unlimited" hosting account at HostGator.
It works, but HostGator does have some undocumented restrictions. One was that they kill MySQL requests which run more than a few seconds. So I had to speed up one transaction that could run long (a good idea anyway) and the database upload had to be done a few hundred records at a time. The daily cron job only runs about a minute, and they're OK with that.
Once, HostGator lost a hard drive and lost the database. The cron job can automatically rebuild the database by re-reading the SEC data for each missing day. (This takes care of routine recovery after downtime). But when the cron job ran for hours, rebuilding nine years of missing data, HostGator didn't like it. We had to talk about that one, and they recovered the database from a backup. That took hours of MySQL time, but they did it.
It's a low-traffic site, though. When I did it, nobody else had SEC filings in a free database. Now all the search engines do. I keep it up more as a reminder of the financial mistakes of the dot-com era. (Although I did call the mortgage crisis in 2006 and put that on Downside. This stuff is obvious if you understand the fundamentals.)
We finally get displays which have no flicker whatsoever, and now these people want to put it back in. Do not want.
Hollywood still thinks 24 FPS is good enough. Many home screens are now capable of delivering much better frame rates. The day is close when "direct to disk" will be better than theater quality. Video games may get there first.
It's still a chemical laser. It's quite possible to make chemical lasers powerful enough to be used as weapons, but so far the equipment has been too big to be very useful. The Mobile Tactical High-Energy Laser is able to shoot down artillery shells and small rockets, but the equipment takes up three trailers and costs too much.
The solid state laser people are catching up. The current output record is around 100 KW. This is enough to be marginally useful for anti-aircraft use. Around a megawatt, things start to get militarily interesting.
Cooling is a huge problem for the solid state devices, though. With the chemical lasers, most of the heat is dumped with the spent chemicals. For the solid state devices, the gear has to be cooled, and efficiency is only around 20%.
Something like that was proposed back in the 1960s to "beat the Russians to the moon". The concept was that a rocket capable of a one-way trip was going to be ready before one that could deliver a return vehicle. So the plan was to deliver an astronaut or two to the moon, follow up with supply rockets, and eventually send a return vehicle when the big booster was ready to launch it. But the Saturn V worked, and the big USSR booster blew up on the pad, so this wasn't necessary.
To get information about Mars, we're probably better off delivering more capable robotic vehicles to Mars.
"Agile" happened.
on
Coders At Work
·
· Score: 4, Interesting
Whatever happened to Q/A? Whatever happened to analysts who have a clue?
"Agile programming" happened. No need to figure out the requirements up front, they're going to change anyway. No need to architect the system, we'll just use a "framework" and add features.
This works OK for web sites in PHP, but don't try to do hard real time or database internals or secure software that way. Agile programming will give you a set of loosely coupled features, but for many user-facing applications, that's good enough.
Of course the rate of progress has slowed. The actual rate of progress peaked around 1880.
Edison's lab, in the 1880s, had a goal of a minor invention every three days, and a major breakthrough every two weeks. That was with forty people.
Look at 1880-1910. That's when it all happened. Railroads were everywhere and locomotives became huge and powerful. Telegraphs were deployed everywhere the tracks went. The steel industry went from making railroad rails and pig iron to a huge industry making steel for everything.
Electric motors went from toy-sized to powering locomotives and streetcars. Telephones and electric lights were deployed. The first generating station started up in 1882. By 1910, huge turbine-powered steam plants were going into service, and most big cities had power. 1885 saw the first gasoline car that worked well (Benz), the Diesel was invented in 1892, and the first large-scale production of cars started in 1902 (Olds). The first airplane flew in 1903. Radio went from experimental to transatlantic. Even Hollerith's first computing equipment was up and running.
Steam, steel, and electricity - that's when it all came together.
No other period, before or since, had as much change in the way the world worked.
Electronic components are harder to get hold of, and information about electronics is harder to get hold of.
Actually, that's totally wrong, but there is a problem. As someone else mentioned, there's Digi-Key, which has most of the electronic parts in existence, takes orders on line, says online what it has in stock, has data sheets for all the parts on line, and ships within hours. As recently as the 1990s, many electronics distributors wouldn't even take credit cards. Having printed circuit boards made now means designing on line (with free software, even), sending a file to a board house, and waiting a few days for the board to show up. You can even get free simulation programs (try LTspice) to try analog circuits before you build them.
The problems for hobbyists and kids aren't on the parts side. They're on the engineering side. In the 1950s, building an audio amplifier or a radio was a reasonable project. Something that turned lights on when it got dark impressed people. Now, who would bother? Nobody would be impressed. Nobody would use the thing. So why do it?
Building anything comparable to even low-end consumer electronics requires engineering skills way beyond the hobbyist level. That's the big problem. Understanding basic electronics isn't enough. You need a good knowledge of electronics (at the Art of Electronics level), and then programming skills, possibly down to the FPGA level. It's quite possible to get all these skills, but it's a lot of information to absorb.
The other problem is that surface-mount part assembly requires special tools, magnifiers, microscopes, and the precision of a watchmaker. Kids have trouble working with that level of precision. Many newer parts are surface-mount only. Yes, you can solder surface mount parts by putting a computer controlled temperature controller on a toaster oven, but even setting that up costs a few hundred dollars.
Most electronics hobbyists today are pros who build stuff in their spare time.
This was perhaps the first domain name registered under ".com", not the first domain name.
What you're seeing here is the beginning of DNS replacing HOSTS.TXT. Before DNS, every site had to FTP over a new copy of the HOSTS.TXT file from SRI-NIC to update the name to IP address translation. There were thousands of names in HOSTS.TXT before the transition, and they all predate this one. Many were grandfathered into ".com". I had domain names in HOSTS.TXT from
1982 or so.
The original idea was to have a much more hierarchical system. Big organizations would have one (1) domain, like "FORD", with other domains under that. So the global name file was expected to be small.
When something like this happened in the auto industry with the Ford Pinto, there were lawsuits, recalls, scandals, and the demise of the brand. The Ford Pinto would occasionally catch fire if hit from the rear by the "external force" of another vehicle. The gas tank could be pushed into the differential, causing a leak and explosion. There were only 27 such incidents, out of millions of Pintos built.
The situation is very similar. The iPhone has a lot of energy stored in a fragile container, and damage to that container can release the energy and cause a fire or explosion. Such devices must be engineered to fail in a safe way when damaged, just as cars are. (Cars very seldom blow up in collisions, despite what one sees in movies).
The computer and phone industries aren't used to being held to the safety standards of the auto industry. Legally, though, they have the same responsibilities. Apple is now finding that out.
"Imagine the outcry if the courts were to legalize patents on English prose. Suddenly, you could get a "literary patent" on novels employing a particular kind of plot twist..."
Copyright on literary concepts is strong enough to survive conversion from book to film, even when nothing remains of the original dialogue. It's strong enough to cover original sequels. Read Harry Potter and the Unauthorized Sequel. The concept of "scenes a faire" covers the concept of literary "prior art" and prevents re-copyrighting the obvious. This is generally considered workable, although it took some litigation in the 1980s before the law settled down as regarding video game "look and feel".
"Small businesses and nonprofit organizations far removed from the traditional software industry have IT departments producing potentially infringing software. The Brookings Institution's Ben Klemens has" documented that this is not a theoretical problem"
Following the "documented" link leads to a set of PowerPoint slides by someone listed as "Senior Statistician, Mood and Affective Disorders, NIMH". (Where does the Cato Institute find these people?) He's grumbling about infringement lawsuits directed against the Green Bay Packers, Caterpillar, Kraft Foods, J. Crew, Linens and Things, McDonalds, Dole Food, and Oprah Winfrey. All occupy dominant positions in their industry. (Technically, the Green Bay Packers are a "small business", with only 189 employees, but the business is valued at $911 million.) None is a nonprofit.
Massive's agreement with Blizzard positions the firm as the sole advertising provider for Battle.net, the online Blizzard-only gaming service due for a significant relaunch upon the release of next year's StarCraft II.
I wonder what will be advertised in StarCraft.
Maybe interstitial ads during loading.
One of my back-burner ideas is to write a video player that inserts ads whenever the stream isn't keeping up. Recognize breaks in the video, and buffer ahead until you have enough video locally to play to the next break. During idle periods, download ads. Whenever the player doesn't have the main stream buffered out to the next break, play ads until the stream catches up. "And now, a word from our sponsor".
Not sure about that. Do diesel locomotives have wiring like this?
Yes. That's the power electronics for a 150 ton Diesel-electric mining truck. Modern locomotives have similar gear, but I can't find a picture quickly.
A real question is whether basic research mines a depleting resource. The issue is not whether more can be discovered. The question is whether basic research is still cost-effective.
The problem is that we've found most of the easy hits in science and technology. Edison's lab, circa 1880, had a goal of one minor invention every three days and one major invention every two months. This was with a rather modest staff, about 40 people. In no area of science is that level of output possible today. Everything that easy has already been done.
One guy built the first IC in two months at Texas Instruments, without much help. Several early microprocessors were designed by teams of about 5 people. It took 3,000 people, in grey cubicles in Santa Clara, to design the Pentium Pro/II/III architecture, the first superscalar IC. (Getting a working design out of 3000 people for something as tightly integrated as a fast superscalar CPU for a complex architecture was an incredible management achievement.)
Xerox PARC, in its heyday, had a nearly open field in which to work. PARC was funded on the concept that if they did things which cost too much now, they'd be affordable later and Xerox would own the technology. So they built things like the Dover (a $100,000 laser printer) and the Alto (a $30,000 single-user computer), and they built enough of them to be useful experimentally.
So they were able to make major progress with about 40 people in the computer science section. (PARC seemed bigger, but a big part of PARC was copier technology development. Everybody forgets those people, but they filled much of the building.) Now it takes resources like that to develop a midrange cell phone.
There's a curve here, it gets steeper, and not in a good way.
The easy stuff is discovered/invented first, then the harder stuff. Each new bit of progress comes at a higher price. It's like mining lower and lower grade ore as the good stuff runs out.
Those are places I know about and visited at one time or another, all in semiconductors and electronics. I can't speak for biotech, which seems not to be as far up the curve yet. On the other hand, consider aviation and rocketry, where the easy part ended in the 1960s. Look at progress in aviation between 1903 and 1956 (Wright Flyer to Boeing 707), then 1956 to 2009 (Boeing 707 to Boeing 787). Or rocketry between 1929 and 1969 (Goddard rocket to Apollo) and 1969 to 2009 (Apollo to Shuttle).
Researchers have been sounding this alarm for years, if not decades. But what makes this significant is hearing it from the likes of BusinessWeek. If the Wall Street Journal ever catches on, we might be close to some real change.
That's a good point. The Wall Street Journal's approach would probably be a bigger R&D tax credit.
So now we see the far-reaching disaster that occurs when we enforce these stupid software patents
It could be worse. When Kodak lost a patent case against Polaroid in 1985, they were given 30 days to stop making film for instant cameras, exit the business, and buy back all retailer inventory. Which they did. Then they lost a class action suit and had to buy back the cameras already in consumer hands.
Virtual trade shows? Let's wait until they're actually successful before tossing them in the pot?
Who goes to physical trade shows any more? COMDEX is gone. E3 is half-dead. Things are slow in the convention business.
Back in the mid-1980s, Autodesk had booths at about fifty trade shows a year. Nobody does that any more.
(Speaking of stretching things, Autodesk in the 1980s had one of the first lightweight foldable space-frame booths, which fitted into a suitcase with wheels. The two women who did all the trade shows would take the frame out, unfold it, grab opposite ends, and pull, creating an instant booth. They were sometimes glared at by the union labor who usually put up booths and took hours to do it.)
A much more effective concept is solar roofs. Rather than putting panels on top of roofs, the panels are the roof. This has many advantages. Rather than paying for a roof and solar panels, plus the headaches of attaching panels to a roof, you only pay for one surface.
Mounting roof panels to rafters is easier than mounting panels to existing roofs. The wiring is on the inside, where it's in a dry space. The panels behave better in high winds, since winds can't get under them. And you can mix solar panels and plain roof panels, using solar panels only on the surfaces pitched to get the most sun.
I see the Apple PR people are out in force, modding this down. It usually takes about 45 minutes to 2 hours for a post critical of Apple to be modded down, which may reflect the Google crawl and update delay. It would be fun if Slashdot graphed this. The usual trajectory is that something critical of Apple gets modded up in the first few minutes, then the Apple promoters get involved.
Keep the device away from you. If you are using iPod for music or iPhone to talk to someone or surfing, use the head sets, ear phones and Bluetooth devices.
Be aware of changes in the device. If the iPhone is getting hotter, if its start to make noises, raise smoke or shake un normally, it's a sign something is about to happen.
Use safety devices for your iPhone and iPod. Any kind of cover, plastic cover can save you when the iPhone is about to explode when the plastic cover will absorb most of the pieces which are about to blast away.
Listen to your iPhone and get to know when the iPhone is about is explode in you hand on in the pocket. When you feel this is the time, the best thing to do is to throw it away and let the iPhone explode far from you. Later on, please collect all the pieces.
The sad thing about the Computer Museum is that almost nothing there works. The Difference Engine replica is about it, and that's entirely mechanical. Some people tried to restore an IBM 1620 back in 1999, but they never got it working.
It's almost the last computer museum, too. The ones in Boston, San Diego, and Germany went bust. There's one still open in Bozeman, Montana. There are a few others which are just stuff in storage. That's about it.
People on phones are bad enough, but I've noticed that a significant fraction of young kids, mostly boys, seem to have zero situational awareness outside of a wedge about 90 degrees wide in front of them.
I notice this because I have horses. Parents are always bringing kids to horse barns. Usually, the girls are interested in the horses, and the boys would rather be playing a video game. Some have portable games with them. In a busy, working stable, there are people leading and riding horses all over the place. Occasionally there's a loose horse (not a big deal; they'll head for their friends or food). Some awareness of large, moving, three quarter ton animals is needed to survive, or at least not to tie up traffic. I've had kids not notice when a horse came up behind them, clicking steel shoes on cement. I've seen horses, being careful of the kids, trying to get them to move out of the way. Some kids don't notice a horse breathing down their neck, literally. (To a horse, breathing down your neck is a polite hello. A nudge with the nose from behind is a demand for attention comparable to yelling at someone. People who still don't get it will usually be shouldered aside if the horse really wants to get through.)
How will these kids ever survive a bad neighborhood or heavy traffic? Will they need a heads-up display with a tail warning system, like fighter aircraft?
This is embarrassing. Look at the thing. It looks like a land speed record vehicle. It's turbine powered. They took it to the Bonneville Salt Flats, where reaching 200 MPH in a straight line is no big deal.
And they went 140MPH. Most production sports cars can do that. Some dragsters now exceed 300MPH for a quarter mile. If you don't have to corner, going fast is easy.
The current land speed record for wheel-driven vehicles is 451 MPH. (The record for thrust-driven vehicles is over Mach 1, but those are really low-flying aircraft.) The record for electrics is 257 MPH. There was an unsuccessful British attempt to break 300 MPH with an electric car in 2005; the power train works but the vehicle was unstable in a crosswind.
357 MPH has been reached with a TGV train. (Maglevs do slightly better, with the record there being 361 MPH.)
So 140 MPH on the Bonneville Salt Flats just isn't very impressive.
Open source software has traditionally had overly aggressive command completion. Developers tend to be keyboard-oriented but not strong typists, while most end-users are now mouse-oriented.
A classic example is Open Office's word completion. It assumes that the user is looking at the screen and interacting with the program from keystroke to keystroke as they type.
If Firefox completely removed command completion and just kept the feature that feeds non-URLs into Google, most users would probably be happier.
Slashdot Mirror
There are some famous examples of this working. But only because the parents had time, money, and high standards.
One of the Rockefellers, the son of John D., wrote that when he was a kid, his father gave him an allowance. He was required to keep a proper set of double-entry books on how he spent it, and the books were audited by an accountant. He didn't get the next allowance payment until the books balanced.
Henry Ford II was promised a car for some birthday. On the appointed day, he was taken out to a garage, and there was the car - totally dissembled with all the component parts laid out. A full set of tools was supplied. Eventually, he did get the car assembled and running.
If you have the resources, it can work.
One of my sites, "downside.com", has a MySQL database of every Securities and Exchange Commission filing since 2000. There's a cron job that updates the database from the SEC site every day at 4 AM. This used to run on EZPublishing, until they gave up hosting to focus on "permission e-mail" (really). It's now running on an $14.95 "unlimited" hosting account at HostGator.
It works, but HostGator does have some undocumented restrictions. One was that they kill MySQL requests which run more than a few seconds. So I had to speed up one transaction that could run long (a good idea anyway) and the database upload had to be done a few hundred records at a time. The daily cron job only runs about a minute, and they're OK with that.
Once, HostGator lost a hard drive and lost the database. The cron job can automatically rebuild the database by re-reading the SEC data for each missing day. (This takes care of routine recovery after downtime). But when the cron job ran for hours, rebuilding nine years of missing data, HostGator didn't like it. We had to talk about that one, and they recovered the database from a backup. That took hours of MySQL time, but they did it.
It's a low-traffic site, though. When I did it, nobody else had SEC filings in a free database. Now all the search engines do. I keep it up more as a reminder of the financial mistakes of the dot-com era. (Although I did call the mortgage crisis in 2006 and put that on Downside. This stuff is obvious if you understand the fundamentals.)
We finally get displays which have no flicker whatsoever, and now these people want to put it back in. Do not want.
Hollywood still thinks 24 FPS is good enough. Many home screens are now capable of delivering much better frame rates. The day is close when "direct to disk" will be better than theater quality. Video games may get there first.
It's still a chemical laser. It's quite possible to make chemical lasers powerful enough to be used as weapons, but so far the equipment has been too big to be very useful. The Mobile Tactical High-Energy Laser is able to shoot down artillery shells and small rockets, but the equipment takes up three trailers and costs too much.
The solid state laser people are catching up. The current output record is around 100 KW. This is enough to be marginally useful for anti-aircraft use. Around a megawatt, things start to get militarily interesting.
Cooling is a huge problem for the solid state devices, though. With the chemical lasers, most of the heat is dumped with the spent chemicals. For the solid state devices, the gear has to be cooled, and efficiency is only around 20%.
Something like that was proposed back in the 1960s to "beat the Russians to the moon". The concept was that a rocket capable of a one-way trip was going to be ready before one that could deliver a return vehicle. So the plan was to deliver an astronaut or two to the moon, follow up with supply rockets, and eventually send a return vehicle when the big booster was ready to launch it. But the Saturn V worked, and the big USSR booster blew up on the pad, so this wasn't necessary.
To get information about Mars, we're probably better off delivering more capable robotic vehicles to Mars.
Whatever happened to Q/A? Whatever happened to analysts who have a clue?
"Agile programming" happened. No need to figure out the requirements up front, they're going to change anyway. No need to architect the system, we'll just use a "framework" and add features.
This works OK for web sites in PHP, but don't try to do hard real time or database internals or secure software that way. Agile programming will give you a set of loosely coupled features, but for many user-facing applications, that's good enough.
Of course the rate of progress has slowed. The actual rate of progress peaked around 1880. Edison's lab, in the 1880s, had a goal of a minor invention every three days, and a major breakthrough every two weeks. That was with forty people.
Look at 1880-1910. That's when it all happened. Railroads were everywhere and locomotives became huge and powerful. Telegraphs were deployed everywhere the tracks went. The steel industry went from making railroad rails and pig iron to a huge industry making steel for everything. Electric motors went from toy-sized to powering locomotives and streetcars. Telephones and electric lights were deployed. The first generating station started up in 1882. By 1910, huge turbine-powered steam plants were going into service, and most big cities had power. 1885 saw the first gasoline car that worked well (Benz), the Diesel was invented in 1892, and the first large-scale production of cars started in 1902 (Olds). The first airplane flew in 1903. Radio went from experimental to transatlantic. Even Hollerith's first computing equipment was up and running.
Steam, steel, and electricity - that's when it all came together. No other period, before or since, had as much change in the way the world worked.
We've already had the Singularity.
Electronic components are harder to get hold of, and information about electronics is harder to get hold of.
Actually, that's totally wrong, but there is a problem. As someone else mentioned, there's Digi-Key, which has most of the electronic parts in existence, takes orders on line, says online what it has in stock, has data sheets for all the parts on line, and ships within hours. As recently as the 1990s, many electronics distributors wouldn't even take credit cards. Having printed circuit boards made now means designing on line (with free software, even), sending a file to a board house, and waiting a few days for the board to show up. You can even get free simulation programs (try LTspice) to try analog circuits before you build them.
The problems for hobbyists and kids aren't on the parts side. They're on the engineering side. In the 1950s, building an audio amplifier or a radio was a reasonable project. Something that turned lights on when it got dark impressed people. Now, who would bother? Nobody would be impressed. Nobody would use the thing. So why do it?
Building anything comparable to even low-end consumer electronics requires engineering skills way beyond the hobbyist level. That's the big problem. Understanding basic electronics isn't enough. You need a good knowledge of electronics (at the Art of Electronics level), and then programming skills, possibly down to the FPGA level. It's quite possible to get all these skills, but it's a lot of information to absorb.
The other problem is that surface-mount part assembly requires special tools, magnifiers, microscopes, and the precision of a watchmaker. Kids have trouble working with that level of precision. Many newer parts are surface-mount only. Yes, you can solder surface mount parts by putting a computer controlled temperature controller on a toaster oven, but even setting that up costs a few hundred dollars.
Most electronics hobbyists today are pros who build stuff in their spare time.
This was perhaps the first domain name registered under ".com", not the first domain name.
What you're seeing here is the beginning of DNS replacing HOSTS.TXT. Before DNS, every site had to FTP over a new copy of the HOSTS.TXT file from SRI-NIC to update the name to IP address translation. There were thousands of names in HOSTS.TXT before the transition, and they all predate this one. Many were grandfathered into ".com". I had domain names in HOSTS.TXT from 1982 or so.
The original idea was to have a much more hierarchical system. Big organizations would have one (1) domain, like "FORD", with other domains under that. So the global name file was expected to be small.
When something like this happened in the auto industry with the Ford Pinto, there were lawsuits, recalls, scandals, and the demise of the brand. The Ford Pinto would occasionally catch fire if hit from the rear by the "external force" of another vehicle. The gas tank could be pushed into the differential, causing a leak and explosion. There were only 27 such incidents, out of millions of Pintos built.
The situation is very similar. The iPhone has a lot of energy stored in a fragile container, and damage to that container can release the energy and cause a fire or explosion. Such devices must be engineered to fail in a safe way when damaged, just as cars are. (Cars very seldom blow up in collisions, despite what one sees in movies).
The computer and phone industries aren't used to being held to the safety standards of the auto industry. Legally, though, they have the same responsibilities. Apple is now finding that out.
"Imagine the outcry if the courts were to legalize patents on English prose. Suddenly, you could get a "literary patent" on novels employing a particular kind of plot twist..."
Copyright on literary concepts is strong enough to survive conversion from book to film, even when nothing remains of the original dialogue. It's strong enough to cover original sequels. Read Harry Potter and the Unauthorized Sequel. The concept of "scenes a faire" covers the concept of literary "prior art" and prevents re-copyrighting the obvious. This is generally considered workable, although it took some litigation in the 1980s before the law settled down as regarding video game "look and feel".
"Small businesses and nonprofit organizations far removed from the traditional software industry have IT departments producing potentially infringing software. The Brookings Institution's Ben Klemens has" documented that this is not a theoretical problem"
Following the "documented" link leads to a set of PowerPoint slides by someone listed as "Senior Statistician, Mood and Affective Disorders, NIMH". (Where does the Cato Institute find these people?) He's grumbling about infringement lawsuits directed against the Green Bay Packers, Caterpillar, Kraft Foods, J. Crew, Linens and Things, McDonalds, Dole Food, and Oprah Winfrey. All occupy dominant positions in their industry. (Technically, the Green Bay Packers are a "small business", with only 189 employees, but the business is valued at $911 million.) None is a nonprofit.
Massive's agreement with Blizzard positions the firm as the sole advertising provider for Battle.net, the online Blizzard-only gaming service due for a significant relaunch upon the release of next year's StarCraft II.
I wonder what will be advertised in StarCraft. Maybe interstitial ads during loading.
One of my back-burner ideas is to write a video player that inserts ads whenever the stream isn't keeping up. Recognize breaks in the video, and buffer ahead until you have enough video locally to play to the next break. During idle periods, download ads. Whenever the player doesn't have the main stream buffered out to the next break, play ads until the stream catches up. "And now, a word from our sponsor".
I'm tempted to put this into VideoLan as a joke, and have it run 1950s drive in movie ads.
Not sure about that. Do diesel locomotives have wiring like this?
Yes. That's the power electronics for a 150 ton Diesel-electric mining truck. Modern locomotives have similar gear, but I can't find a picture quickly.
A real question is whether basic research mines a depleting resource. The issue is not whether more can be discovered. The question is whether basic research is still cost-effective.
The problem is that we've found most of the easy hits in science and technology. Edison's lab, circa 1880, had a goal of one minor invention every three days and one major invention every two months. This was with a rather modest staff, about 40 people. In no area of science is that level of output possible today. Everything that easy has already been done.
One guy built the first IC in two months at Texas Instruments, without much help. Several early microprocessors were designed by teams of about 5 people. It took 3,000 people, in grey cubicles in Santa Clara, to design the Pentium Pro/II/III architecture, the first superscalar IC. (Getting a working design out of 3000 people for something as tightly integrated as a fast superscalar CPU for a complex architecture was an incredible management achievement.)
Xerox PARC, in its heyday, had a nearly open field in which to work. PARC was funded on the concept that if they did things which cost too much now, they'd be affordable later and Xerox would own the technology. So they built things like the Dover (a $100,000 laser printer) and the Alto (a $30,000 single-user computer), and they built enough of them to be useful experimentally. So they were able to make major progress with about 40 people in the computer science section. (PARC seemed bigger, but a big part of PARC was copier technology development. Everybody forgets those people, but they filled much of the building.) Now it takes resources like that to develop a midrange cell phone.
There's a curve here, it gets steeper, and not in a good way. The easy stuff is discovered/invented first, then the harder stuff. Each new bit of progress comes at a higher price. It's like mining lower and lower grade ore as the good stuff runs out.
Those are places I know about and visited at one time or another, all in semiconductors and electronics. I can't speak for biotech, which seems not to be as far up the curve yet. On the other hand, consider aviation and rocketry, where the easy part ended in the 1960s. Look at progress in aviation between 1903 and 1956 (Wright Flyer to Boeing 707), then 1956 to 2009 (Boeing 707 to Boeing 787). Or rocketry between 1929 and 1969 (Goddard rocket to Apollo) and 1969 to 2009 (Apollo to Shuttle).
Researchers have been sounding this alarm for years, if not decades. But what makes this significant is hearing it from the likes of BusinessWeek. If the Wall Street Journal ever catches on, we might be close to some real change.
That's a good point. The Wall Street Journal's approach would probably be a bigger R&D tax credit.
So now we see the far-reaching disaster that occurs when we enforce these stupid software patents
It could be worse. When Kodak lost a patent case against Polaroid in 1985, they were given 30 days to stop making film for instant cameras, exit the business, and buy back all retailer inventory. Which they did. Then they lost a class action suit and had to buy back the cameras already in consumer hands.
Virtual trade shows? Let's wait until they're actually successful before tossing them in the pot?
Who goes to physical trade shows any more? COMDEX is gone. E3 is half-dead. Things are slow in the convention business.
Back in the mid-1980s, Autodesk had booths at about fifty trade shows a year. Nobody does that any more.
(Speaking of stretching things, Autodesk in the 1980s had one of the first lightweight foldable space-frame booths, which fitted into a suitcase with wheels. The two women who did all the trade shows would take the frame out, unfold it, grab opposite ends, and pull, creating an instant booth. They were sometimes glared at by the union labor who usually put up booths and took hours to do it.)
A much more effective concept is solar roofs. Rather than putting panels on top of roofs, the panels are the roof. This has many advantages. Rather than paying for a roof and solar panels, plus the headaches of attaching panels to a roof, you only pay for one surface. Mounting roof panels to rafters is easier than mounting panels to existing roofs. The wiring is on the inside, where it's in a dry space. The panels behave better in high winds, since winds can't get under them. And you can mix solar panels and plain roof panels, using solar panels only on the surfaces pitched to get the most sun.
Roads are a much tougher environment than roofs.
I see the Apple PR people are out in force, modding this down. It usually takes about 45 minutes to 2 hours for a post critical of Apple to be modded down, which may reflect the Google crawl and update delay. It would be fun if Slashdot graphed this. The usual trajectory is that something critical of Apple gets modded up in the first few minutes, then the Apple promoters get involved.
Get to know when your iPhone is about to explode.
The sad thing about the Computer Museum is that almost nothing there works. The Difference Engine replica is about it, and that's entirely mechanical. Some people tried to restore an IBM 1620 back in 1999, but they never got it working.
It's almost the last computer museum, too. The ones in Boston, San Diego, and Germany went bust. There's one still open in Bozeman, Montana. There are a few others which are just stuff in storage. That's about it.
The history of this field disappears very fast.
People on phones are bad enough, but I've noticed that a significant fraction of young kids, mostly boys, seem to have zero situational awareness outside of a wedge about 90 degrees wide in front of them.
I notice this because I have horses. Parents are always bringing kids to horse barns. Usually, the girls are interested in the horses, and the boys would rather be playing a video game. Some have portable games with them. In a busy, working stable, there are people leading and riding horses all over the place. Occasionally there's a loose horse (not a big deal; they'll head for their friends or food). Some awareness of large, moving, three quarter ton animals is needed to survive, or at least not to tie up traffic. I've had kids not notice when a horse came up behind them, clicking steel shoes on cement. I've seen horses, being careful of the kids, trying to get them to move out of the way. Some kids don't notice a horse breathing down their neck, literally. (To a horse, breathing down your neck is a polite hello. A nudge with the nose from behind is a demand for attention comparable to yelling at someone. People who still don't get it will usually be shouldered aside if the horse really wants to get through.)
How will these kids ever survive a bad neighborhood or heavy traffic? Will they need a heads-up display with a tail warning system, like fighter aircraft?
China had 40 "death vans" built by Junguan Group to standardize executions. Now we know why.
This is embarrassing. Look at the thing. It looks like a land speed record vehicle. It's turbine powered. They took it to the Bonneville Salt Flats, where reaching 200 MPH in a straight line is no big deal. And they went 140MPH. Most production sports cars can do that. Some dragsters now exceed 300MPH for a quarter mile. If you don't have to corner, going fast is easy.
The current land speed record for wheel-driven vehicles is 451 MPH. (The record for thrust-driven vehicles is over Mach 1, but those are really low-flying aircraft.) The record for electrics is 257 MPH. There was an unsuccessful British attempt to break 300 MPH with an electric car in 2005; the power train works but the vehicle was unstable in a crosswind. 357 MPH has been reached with a TGV train. (Maglevs do slightly better, with the record there being 361 MPH.)
So 140 MPH on the Bonneville Salt Flats just isn't very impressive.
Open source software has traditionally had overly aggressive command completion. Developers tend to be keyboard-oriented but not strong typists, while most end-users are now mouse-oriented.
A classic example is Open Office's word completion. It assumes that the user is looking at the screen and interacting with the program from keystroke to keystroke as they type.
If Firefox completely removed command completion and just kept the feature that feeds non-URLs into Google, most users would probably be happier.