Our business (85%) is custom embedded firmware and hardware designs. We make some money from the few products we sell OTS; but even those are by their nature designed to be customized.
Whereever possible, we give changes back and our released code is free. Open source tools allowed us to build our business on a very tight budget. Good Karma.:)
There is still a lot of resistance to releasing source, period, though. We find smaller clients much more receptive to the idea of GPL code inclusion than larger ones who can afford to pay more for propietary development.
You could do a pretty good job with a nanoITX, and a small TFT. The keyboard would be the biggest challenge to do cost-effectively. Then there's the matter of getting it in a case.
I've used a industrial PC with no HD as a primary computer in the field before. It worked well. Looked like a flat slab of metal connected to a screen though, with a half-size keyboard.
Now, if your communications were encrypted end-end with hard encryption with keys you control, this would be a moot point. Coming soon to a VOIP / programmable cell phone near you.
I'm not sure how they released without some sort of GUI building tool. I haven't used Java in awhile, but I remember coming up with Shakespearean-esque curses to describe what I thought of GridLayout and their buddies.
Is there an accepted good builder now? The article seems to imply this is far from a solved problem. (10 years later!)
Stuff everybody uses - linux kernel, utilities, graphics, communications protocols and utilities, web browsers. It is stupid to have 30 of these things. It's better to have one common set everyone improves upon, and this is where open source shines - in software used across industries.
Where it works less well is in how that infrastructure is used an applied. Everybody might contribute to a linux based RTOS, but a the end of the day, it's going to be used someplace, where someone is paid to make something work. That someone doesn't have to open their work.
See the distinction? Pretty clear to me. What is less clear is what we need massive infrastructure companies like MS for, other than to enfore an artificial standard. There are other ways to do that, with their own pros and cons. (See: OpenGL).
And a new nuclear power plant can be built in four or five years.
Scaling up the uranium mines is the main problem in the scenario I am talking about. I would also question the ability to get nuclear approvals, even in a crisis, in 5 years. Coal, at least here, is already being expanded.
Coal fields require little processing and are already capable of scaling in dramatic fashion. Uranium needs a lot of processing.
I could care less if we burned coal 100% for 10 years, if at the end of the 10 years, we had Fusion. Given enough energy and resources, mainly energy, we can repair or engineer around damage to the environment.
And it doesn't take awhile to build new coal power plants? Tell me, what's the better investment for the future?
A new coal boiler can be brought online quickly, as the regulatory requirements - security, environmental, and supply - are much lower. Adding onto an existing plant can be done even faster.
As far as investments for the future go, funding fusion research en masse 20 years ago would seem like a pretty good deal, compared to the little skirmish we have in the middle east now. I'm worried about the mess we're going to be in 10 years from now. Not 25. Not 35. 10.
Barring serious economic recession (always a possibility), nuclear isn't really an option anymore. It takes awhile to get the plants online, and there would have to be a very large number of them built in a very short period of time. As an engineer, that'd be great news.
Unfortunately, coal is about the only buffer fuel left that would take us over that hump that depleting oil supplies will leave. The hump gets worse every single day we wait..
People should have demanded Manhattan Project style investment into nuclear fusion after the last energy crisis. We'll have another chance soon.
We're a small shop. If I can write code once, on a stable platform, and keep it away from Microsoft's API of the week - great. Whenever possible, that means web-based applications. For things where that is not practical, and that is getting smaller, I have had great success with Python as a application environment. Compiling natively provides good speed, and with the toolkits out there it's easy to jump across platforms.
Java offers similar advantages, but I find the GUI code overly complicated for what I want to do.
The answer is different depending on what you want to do, but this is a trend I am noticing more and more.
You are both partially correct. The code has to be made available to those who you ship binaries to, which is easy in our case - we do primarily contract development.
Our own code, is licensed as we deem fit based on the application and client.
We do a lot of embedded linux projects - mainly custom boards, done around some sort of ARM chip with standard connectivity - LCD, ethernet, or wireless options.
My company invested a lot of time in implementing and setting up our own toolchain and utilities, support libraries etc around the linux kernel. What we end up with is a redistributable result with no liscencing whatsoever.
Of course, it requires you be able to work to produce a flash image and toolchain. Once that pain is dealt with..and there is pain, a lot of it on a custom board.. then you're free to do whatever you want.
x1000's, people count pennies, and WinCE is not pennies.
[quote] An editorial team could be drawn from the very same people who have created the products currently in use. A full, usable set could be accomplished in 18 months or less. [/quote]
I would suggest that explaining calculus in a easy to understand manner is not as straightforward as you might think. Ditto engineering statics and dynamics, or undergraduate physics.
Get the rights to the "best of breed" textbooks; I know there are clear favorites in Engineering and Mathematics. From there, use them as the base in wiki format to extend them. A good set of undergraduate texts would do lots of good for the developing world and poor students everywhere. Buying books is EXPENSIVE, and in most engineering related disiplines, a real waste, since the base mathematics has not changed in many years.
They buy YouTube, and with a little tightening of the noose, they're removed as a threat and they've been made an example of for anyone else who thinks to follow - for example, Microsoft.
Google can then move into this market at will. I'm all for draconian copyright enforcement, because it will lead to widespread civil disobedience and ultimately, a changing of the laws in what the public deems it's interest. It needs to get a little worse still, but the seeds are already there.
Speaking in evolutionary terms, doesn't the survival rate of a species increase with rapid, frequent, and earlier reproduction? In a massive die-off, it will be the offspring of the poor who are better suited for survival. They have a more diverse gene pool, and they have sheer volume on thier side.
Except the rich people control where the tanks, missiles, and bombs go.
With their aquisition of ATI, I am much more worried about chipset instability. Anyone else remember the bad old days with the horrible via chipsets and mystery conflicts with nvidia hardware?
Then the finger pointing starts, and we're stuck in the middle. I'm upgrading for the first time in 3 years, hopefully I can wait all this mess out. It'll be an AMD chip though. If I had to pick, I'd go with whatever platform Nvidia supports in the future. Their commitment to driver quality deserves to be rewarded and won my loyalty - and interestingly enough, I have never purchased another ATI product after their little opengl driver fiasco.
The issue is can it be done at a net energy profit.
I'd be more optimistic if we were looking for ways to turn hudson's bay into a hydroelectric site to run the pumps.
That graph is one of the reasons I've held off on having kids. You better hope those guys at Steorn have something. Or sometime, between now and ~2030, someone comes up with something similar. Or makes fusion work. Or we're all toast.
Slashdot Mirror
Anyone remember print.google.com?
I'm waiting for google to buy a huge book publisher next, put the whole catalog up, and sell ads..
I'd pay a few bucks a month (maybe more than a few) for a HD feed of the planet in near-realtime.
Any rich geeks want to put up a spy sat? I'm sure I'm not alone.
Our business (85%) is custom embedded firmware and hardware designs. We make some money from the few products we sell OTS; but even those are by their nature designed to be customized.
:)
Whereever possible, we give changes back and our released code is free. Open source tools allowed us to build our business on a very tight budget. Good Karma.
There is still a lot of resistance to releasing source, period, though. We find smaller clients much more receptive to the idea of GPL code inclusion than larger ones who can afford to pay more for propietary development.
You could do a pretty good job with a nanoITX, and a small TFT. The keyboard would be the biggest challenge to do cost-effectively. Then there's the matter of getting it in a case.
I've used a industrial PC with no HD as a primary computer in the field before. It worked well. Looked like a flat slab of metal connected to a screen though, with a half-size keyboard.
You already have one of these: Optoelectronics Scout
Now, if your communications were encrypted end-end with hard encryption with keys you control, this would be a moot point. Coming soon to a VOIP / programmable cell phone near you.
How seamless windows compatibility worked out for OS2.
I'm not sure how they released without some sort of GUI building tool. I haven't used Java in awhile, but I remember coming up with Shakespearean-esque curses to describe what I thought of GridLayout and their buddies.
Is there an accepted good builder now? The article seems to imply this is far from a solved problem. (10 years later!)
Obsolete them.
:)
If you had an affordable, reliable means to detect mines easily and quickly, they would be removed from warfare very quickly.
Of course, they'd be replaced with equally as bad or worse.
Stuff everybody uses - linux kernel, utilities, graphics, communications protocols and utilities, web browsers. It is stupid to have 30 of these things. It's better to have one common set everyone improves upon, and this is where open source shines - in software used across industries.
Where it works less well is in how that infrastructure is used an applied. Everybody might contribute to a linux based RTOS, but a the end of the day, it's going to be used someplace, where someone is paid to make something work. That someone doesn't have to open their work.
See the distinction? Pretty clear to me. What is less clear is what we need massive infrastructure companies like MS for, other than to enfore an artificial standard. There are other ways to do that, with their own pros and cons. (See: OpenGL).
And a new nuclear power plant can be built in four or five years.
Scaling up the uranium mines is the main problem in the scenario I am talking about. I would also question the ability to get nuclear approvals, even in a crisis, in 5 years. Coal, at least here, is already being expanded.
Coal fields require little processing and are already capable of scaling in dramatic fashion. Uranium needs a lot of processing.
I could care less if we burned coal 100% for 10 years, if at the end of the 10 years, we had Fusion. Given enough energy and resources, mainly energy, we can repair or engineer around damage to the environment.
And it doesn't take awhile to build new coal power plants? Tell me, what's the better investment for the future?
A new coal boiler can be brought online quickly, as the regulatory requirements - security, environmental, and supply - are much lower. Adding onto an existing plant can be done even faster.
As far as investments for the future go, funding fusion research en masse 20 years ago would seem like a pretty good deal, compared to the little skirmish we have in the middle east now. I'm worried about the mess we're going to be in 10 years from now. Not 25. Not 35. 10.
Barring serious economic recession (always a possibility), nuclear isn't really an option anymore. It takes awhile to get the plants online, and there would have to be a very large number of them built in a very short period of time. As an engineer, that'd be great news.
..
Unfortunately, coal is about the only buffer fuel left that would take us over that hump that depleting oil supplies will leave. The hump gets worse every single day we wait
People should have demanded Manhattan Project style investment into nuclear fusion after the last energy crisis. We'll have another chance soon.
Of me kicking the dust off my Java compiler.
Great move guys.
We're a small shop. If I can write code once, on a stable platform, and keep it away from Microsoft's API of the week - great. Whenever possible, that means web-based applications. For things where that is not practical, and that is getting smaller, I have had great success with Python as a application environment. Compiling natively provides good speed, and with the toolkits out there it's easy to jump across platforms.
Java offers similar advantages, but I find the GUI code overly complicated for what I want to do.
The answer is different depending on what you want to do, but this is a trend I am noticing more and more.
[quote]
The real fix is that we start living sustainably.
[/quote]
Very few people who say this actually understand what the implications of that statement are.
Anyone know how these compare? I want to purchase something to replace my G4 Powerbook soon, and am being seduced by the smaller size on the Macbook..
You are both partially correct. The code has to be made available to those who you ship binaries to, which is easy in our case - we do primarily contract development.
Our own code, is licensed as we deem fit based on the application and client.
We do a lot of embedded linux projects - mainly custom boards, done around some sort of ARM chip with standard connectivity - LCD, ethernet, or wireless options.
My company invested a lot of time in implementing and setting up our own toolchain and utilities, support libraries etc around the linux kernel. What we end up with is a redistributable result with no liscencing whatsoever.
Of course, it requires you be able to work to produce a flash image and toolchain. Once that pain is dealt with..and there is pain, a lot of it on a custom board.. then you're free to do whatever you want.
x1000's, people count pennies, and WinCE is not pennies.
[quote]
An editorial team could be drawn from the very same people who have created the products currently in use. A full, usable set could be accomplished in 18 months or less.
[/quote]
I would suggest that explaining calculus in a easy to understand manner is not as straightforward as you might think. Ditto engineering statics and dynamics, or undergraduate physics.
Only terrorists would want that. :rolleyes:
Get the rights to the "best of breed" textbooks; I know there are clear favorites in Engineering and Mathematics. From there, use them as the base in wiki format to extend them. A good set of undergraduate texts would do lots of good for the developing world and poor students everywhere. Buying books is EXPENSIVE, and in most engineering related disiplines, a real waste, since the base mathematics has not changed in many years.
They buy YouTube, and with a little tightening of the noose, they're removed as a threat and they've been made an example of for anyone else who thinks to follow - for example, Microsoft.
Google can then move into this market at will. I'm all for draconian copyright enforcement, because it will lead to widespread civil disobedience and ultimately, a changing of the laws in what the public deems it's interest. It needs to get a little worse still, but the seeds are already there.
Speaking in evolutionary terms, doesn't the survival rate of a species increase with rapid, frequent, and earlier reproduction?
In a massive die-off, it will be the offspring of the poor who are better suited for survival. They have a more diverse gene pool, and they have sheer volume on thier side.
Except the rich people control where the tanks, missiles, and bombs go.
With their aquisition of ATI, I am much more worried about chipset instability. Anyone else remember the bad old days with the horrible via chipsets and mystery conflicts with nvidia hardware?
Then the finger pointing starts, and we're stuck in the middle. I'm upgrading for the first time in 3 years, hopefully I can wait all this mess out. It'll be an AMD chip though. If I had to pick, I'd go with whatever platform Nvidia supports in the future. Their commitment to driver quality deserves to be rewarded and won my loyalty - and interestingly enough, I have never purchased another ATI product after their little opengl driver fiasco.
Why doesn't AMD have a chipset, anyway?
The issue is not can you make money doing it.
The issue is can it be done at a net energy profit.
I'd be more optimistic if we were looking for ways to turn hudson's bay into a hydroelectric site to run the pumps.
That graph is one of the reasons I've held off on having kids. You better hope those guys at Steorn have something. Or sometime, between now and ~2030, someone comes up with something similar. Or makes fusion work. Or we're all toast.