Have you ever actually tried FlightGear? It's painful. Far worse than Flight Simulator. Maybe half the cockpits actually work. Sometimes the artificial horizon ball goes floating around the cockpit. (Somebody did their transforms in the wrong order.) And there's a wierd "turbulence" effect when you cross from one scenery region to another, because the scenery loading concurrency was botched.
Today's launch was the last of the Atlas IIAS line. There were earlier models, and there will be models yet to come. There was an Atlas IIA, an Atlas II, and, of course, the Atlas A, the first US ICBM.
Meet the Atlas Family, all 15 of them. First flight of a small prototype was in 1947. The first real Atlas flew in 1957. Alan Shepard flew into space on an Atlas D.
It's a big pressurized stainless steel can with engines. Still a good design after half a century.
There are many people doing amateur robotics. There's plenty of interest in the DARPA Grand Challenge. We have a team, hardware, a shop, funding, a working robot vehicle, and a big chunk of the software working. We cover all your expenses. Free munchies and drinks. Occasional TV coverage. It's the coolest robotics project in town.
We've had some good people. It was easier last year, because there were good programmers left over from the dot-com collapse taking temp jobs. Things have picked up, and we lost those guys to places like Sun and Apple. Now they feel they have to work 60+ hours a week to keep their jobs.
Since then, it's been harder to find people with a clue.
So we have to keep looking. If you're bored out of your mind doing something like high-performance ad servers, are in Silicon Valley, and have free time, get in touch. We have fun problems to work on. And there's money at the end of this.
That seems to be the case. C++ does seem to be on the way out. (I blame this on the C++ committee, which is focused on the wrong problems, but that's another issue.)
If you need hard real time, and the problem is complicated enough that you need elaborate data structures, C++ is still the language of choice.
There's a Java working group for hard real-time Java, but that's not a technology ready for prime time yet.
The best people I've found have been game programmers. They're used to geometric computation and tight timing limits.
Then again, we've been waiting 4 years for Enron to settle...
That's been grinding along. Ben Glisan (former Enron treasurer) is in prison.
(Inmate #20293-179, Bastrop Federal Correctional Institution).
He gave up the Fastows. Lea Fastow is in prison.
(Inmate # 20290-179, Houston Federal Detention Center). Andrew Fastow (former Enron CFO) pled guilty and is going to jail soon. He gave up Skilling and Lay (former Enron CEOs). They've been indicted and are out on bail.
It's just like taking down an organized crime operation. Which Enron was. One step at a time, until the guy at the top goes down.
I run a DARPA Grand Challenge team, Team Overbot. We're in Silicon Valley and looking for volunteers.
We have a robot vehicle that runs, and need programmers. You get a share of the $2,000,000 prize if we win. Many people express interest.
Then I ask them to send me 1000 lines of C++ they're proud of. Doesn't matter what it does; I just want to see how they code. Many of them look scared. "Is C OK?" "I'm not really that good at C++". "Can I use Python?".
When someone sends us code, I read it and send comments back. I'm looking for robustness. ("We have received your code sample. Your first buffer overflow is on line 52. Thank you for your interest in Team Overbot.") I'm looking for some basic knowledge of C++. I'm looking for a reasonable level of comments.
I think the number of good programmers out there is declining. There are hordes of sysadmins and low-level coders, more than ever, but most of them aren't that good.
Apple's "solution" is to show fake pictures with no wires connected, not even a power cord. You could probably go after them with a false advertising complaint, since they show the screen lit with no power cord. The thing does have 10 jacks on the back, after all.
With all the Bluetooth stuff, could you cruise by in a van with a directional antenna and take over the machine remotely?
Bluetooth devices really should require physical contact to "bond", and do a key exchange, like the better cordless phones. Did Apple do that? Probably not.
Get a big tower case and a 21" CRT. Stencil your
name in huge letters on everything. Back up offsite to a vaulting system far, far away. Then don't worry about it too much.
12 CPUs on one board, but no shared memory. If they had a 12-way multiprocessor for $10K, that might be worth something, but right now it's an overpriced cluster.
Is there a market for this? If there is, it shouldn't be hard to get the price down by a factor of 2 or 3, at which point it starts to look reasonable.
It's amusing that they used Transmeta CPUs. Transmeta's most useful feature is that their CPUs use very little power when doing nothing. That may say something about the CPU utilization these guys get in their clusters.
Slashdot's so-called "editors" should take his submissions and make them into stories that reference the primary sources, ommitting any references to his blog.
I still remember seeing the original IMP at Case Tech, one of the first sixteen nodes. It was also the first to be removed. Case lost their R&D contract, which was to develop something like a VHDL compiler, decades too early. So DARPA took their IMP back.
But nobody really cared at Case, because the emphasis there was on "high-capacity, fast-turnaround batch computing". They got really good at batch job processing. It was so cost-effective that Case stayed with it years after other schools went interactive.
We'll need this to support the distribution of pirated movies over "file-sharing" networks with inefficient protocols.
"file-sharing" systems pumping around MP3 files are already using orders of magnitude more bandwidth than they should. The RIAA only generates a few gigabytes of new content per week, expressed as MP3 files. If it just went out on a netnews binary group, the bandwidth consumption would be trivial. No file would traverse any link more than once. No frantic inter-node polling.
The consumer electronics industry could just buy out the music industry and throw all the content into the public domain. The entire music industry isn't that big; it's about the size of Compaq when HP acquired it. Content could be viewed as a loss leader for the hardware.
Powering a car suspension with a linear motor is impressive. That's quite a power to weight ratio for a linear motor. Usually, you're lucky to find one that can lift 10x its own weight. And they apparently have the servo drivers to power them sized for automotive applications.
If they put those actuators into production, they'll have applications in robotics.
And yes, great hackers don't like languages that don't trust them, for obvious reasons.
Great programmers, however, do.
I know several people who were the primary authors of the core of major, widely used applications that work. They're all paranoid about bugs. They don't trust their own code. They all hate program crashes, and have no problem with putting in more up-front work to eliminate even a faint possibility of a crash later.
Incidentally, the programmers I'm talking about range from "well off" to "very wealthy". There may be a correlation here.
A few years ago, I received an e-mail that began "I am going to kill you tonight". It wasn't intended for me. I own the.com domain that corresponds to a religious school in ".co.uk", so I get misaddressed mail now and then. It was addressed to another student at the school.
This was less than a year after Columbine. So I felt I should call someone. I got through to some minor official at the school. It turned out to be a 12 year old, They weren't too worried. In the US, something like this probably would bring out a SWAT team.
The big problem is that making small jet engines is still too expensive. Most light planes are still powered by reciprocating engines. There's been talk of small jets for general aviation for decades, but nobody seems to be able to bring it off.
It's not that you can't build a small jet engine.
It's that the price doesn't decline much with size. Engines sized for small aircraft aren't much cheaper than those built for business jets.
Slashdot Mirror
Have you ever actually tried FlightGear? It's painful. Far worse than Flight Simulator. Maybe half the cockpits actually work. Sometimes the artificial horizon ball goes floating around the cockpit. (Somebody did their transforms in the wrong order.) And there's a wierd "turbulence" effect when you cross from one scenery region to another, because the scenery loading concurrency was botched.
Today's launch was the last of the Atlas IIAS line. There were earlier models, and there will be models yet to come. There was an Atlas IIA, an Atlas II, and, of course, the Atlas A, the first US ICBM.
Meet the Atlas Family, all 15 of them. First flight of a small prototype was in 1947. The first real Atlas flew in 1957. Alan Shepard flew into space on an Atlas D.
It's a big pressurized stainless steel can with engines. Still a good design after half a century.
Thinking like that led to problems like "Windows XP - Surviving the First Day".
We've had some good people. It was easier last year, because there were good programmers left over from the dot-com collapse taking temp jobs. Things have picked up, and we lost those guys to places like Sun and Apple. Now they feel they have to work 60+ hours a week to keep their jobs.
Since then, it's been harder to find people with a clue. So we have to keep looking. If you're bored out of your mind doing something like high-performance ad servers, are in Silicon Valley, and have free time, get in touch. We have fun problems to work on. And there's money at the end of this.
If you need hard real time, and the problem is complicated enough that you need elaborate data structures, C++ is still the language of choice. There's a Java working group for hard real-time Java, but that's not a technology ready for prime time yet.
The best people I've found have been game programmers. They're used to geometric computation and tight timing limits.
That's been grinding along. Ben Glisan (former Enron treasurer) is in prison. (Inmate #20293-179, Bastrop Federal Correctional Institution). He gave up the Fastows. Lea Fastow is in prison. (Inmate # 20290-179, Houston Federal Detention Center). Andrew Fastow (former Enron CFO) pled guilty and is going to jail soon. He gave up Skilling and Lay (former Enron CEOs). They've been indicted and are out on bail.
It's just like taking down an organized crime operation. Which Enron was. One step at a time, until the guy at the top goes down.
Then I ask them to send me 1000 lines of C++ they're proud of. Doesn't matter what it does; I just want to see how they code. Many of them look scared. "Is C OK?" "I'm not really that good at C++". "Can I use Python?".
When someone sends us code, I read it and send comments back. I'm looking for robustness. ("We have received your code sample. Your first buffer overflow is on line 52. Thank you for your interest in Team Overbot.") I'm looking for some basic knowledge of C++. I'm looking for a reasonable level of comments.
I think the number of good programmers out there is declining. There are hordes of sysadmins and low-level coders, more than ever, but most of them aren't that good.
With all the Bluetooth stuff, could you cruise by in a van with a directional antenna and take over the machine remotely?
Bluetooth devices really should require physical contact to "bond", and do a key exchange, like the better cordless phones. Did Apple do that? Probably not.
Get a big tower case and a 21" CRT. Stencil your name in huge letters on everything. Back up offsite to a vaulting system far, far away. Then don't worry about it too much.
This Slashdot page is being served with a Microsoft ad boasting about their security. Really.
No way should a soldier's superior officer have any way to tell how that soldier voted. No way. That's fundamentally un-American.
Is there a market for this? If there is, it shouldn't be hard to get the price down by a factor of 2 or 3, at which point it starts to look reasonable.
It's amusing that they used Transmeta CPUs. Transmeta's most useful feature is that their CPUs use very little power when doing nothing. That may say something about the CPU utilization these guys get in their clusters.
Slashdot's so-called "editors" should take his submissions and make them into stories that reference the primary sources, ommitting any references to his blog.
But nobody really cared at Case, because the emphasis there was on "high-capacity, fast-turnaround batch computing". They got really good at batch job processing. It was so cost-effective that Case stayed with it years after other schools went interactive.
"file-sharing" systems pumping around MP3 files are already using orders of magnitude more bandwidth than they should. The RIAA only generates a few gigabytes of new content per week, expressed as MP3 files. If it just went out on a netnews binary group, the bandwidth consumption would be trivial. No file would traverse any link more than once. No frantic inter-node polling.
The consumer electronics industry could just buy out the music industry and throw all the content into the public domain. The entire music industry isn't that big; it's about the size of Compaq when HP acquired it. Content could be viewed as a loss leader for the hardware.
Apple seems to be headed in that direction.
"Rip, Mix, Burn" is going to come back to haunt him.
Never send rotating machinery to do an IC's job.
Wireline ring voltage is 88V 20Hz, so all you get is a loud bass buzz.
If they put those actuators into production, they'll have applications in robotics.
Cute, but providing support is so expensive, even if outsourced to Outer Nowhere, that it wouldn't pay.
Great programmers, however, do.
I know several people who were the primary authors of the core of major, widely used applications that work. They're all paranoid about bugs. They don't trust their own code. They all hate program crashes, and have no problem with putting in more up-front work to eliminate even a faint possibility of a crash later.
Incidentally, the programmers I'm talking about range from "well off" to "very wealthy". There may be a correlation here.
This was less than a year after Columbine. So I felt I should call someone. I got through to some minor official at the school. It turned out to be a 12 year old, They weren't too worried. In the US, something like this probably would bring out a SWAT team.
You're right. I used a Symbolics with that keyboard once. Most of those buttons didn't do anything.
Of course, they had to put in "themes", but at least it doesn't download them itself.
It's not that you can't build a small jet engine. It's that the price doesn't decline much with size. Engines sized for small aircraft aren't much cheaper than those built for business jets.
There was an effort at NASA to fix this problem, but it failed and was cancelled in 2002.