I can see it now, the AI in the craft decides against the Miller back at the base and heads to the local bars for a black and tan. A recovery squad is promptly formed to drag the now uncoordinated bots back to work, where they are confined to the base perimeter and reprogrammed to work off of high calorie MREs. Accountants insist on purchasing expired MREs that are unfit for human consumption, but will work fine in the bots. Engineers find that higher than normal failure rates are attributed to the bots being unable to excrete waste products on an appropriate schedule. Unanticipated effects of large unscheduled waste movements while in operation complicate the traffic situation below, devalueing the traffic monitoring bots.
It is a success if they got the data that they were looking for. Even if it failed to put the objects into their assigned orbits, this was a test launch and was designed to test the system. Glitches are bound to appear on the first flight of anything. A successful test does not always mean that the product works. It means that all the test points were touched and data retrieved for all of them.
Did you read the article? He was let go. He was sneaking into the building with the help of the other "still hired engineers". They were found out once when someone tried to fill the office they were squatting in. But they found other offices and continued to sneak in. Also, other sympathetic developers assisted them in developing testing this unoffical software using their own free time.
More than likely this is a small company in asia which is throwing together a bunch of old/cheap components to get a "notebook computer". Chances are that the display is crap, as well as the chassis. The disk drive is probably an old model, which does not really matter for this application (as long as it does not crash). If it is like any of their other "no-name brand" products, paying $500 is too much. Spend the extra $200 or so and get a basic HP or Toshiba. Much better value.
Because this happened 4 years ago when a typical hard drive could only store a dozen movies or so. And a 17 year old is unlikely to be able to afford a large drive (I don't know if he was working or not).
Unfortuneately, for most people to be able to do their work (install programs, move files around, etc) you need to have Admin priviledges. This is a fault of both the applications (for being unable to operate in a restricted envirnoment) and the OS (for not providing adequate methods for doing real work in a restricted environment ).
Until you can install and run the typical program as a user without Admin rights, this problem will persist because the default user will need to have admin rights (to prevent a flurry of tech support calls to Redmond).
Another thing to keep in mind is that household power lines, although carrying a lower voltage, are much closer, and therefore contribute more to the local electric fields.
It's a good thing Edison won out, as to get enough electricity to power anything useful into the air over any real distance would be a huge cancer risk.
Please provide some references for this. I did a paper on this topic about 6 years ago and I could not find one study that provided a link between power and radio frequency radiation and cancer.
What I did find was a lot of people who wanted to blame someone for their ailments. I read several complaints and they all basically read "there were no carcinogens found in the soil/air/water. A percentage of the people all lived close to power lines. Therefore the power lines must be at fault." From what I understand not one of those arguments held up under scrutiny.
I did try it and it was sloooooow. It was a fairly complete IDE, tho.
What is needed are several things in my pie-in-the-sky, IDE.
1) A common interface for help searches. So when I highlight the 'foo' function, I can get a help page on it. Whether that function is in the C library or in some other installed package. The IDE needs to glue these things together.
2) The ability to create a shell application with little interactivity (Project Wizard). I want a GUI app? Press this button. Console app? Press here. Device driver? etc. SharpDevelop does this pretty well.
3) "Visual" code creation. If I am throwing together a simple GUI app, I should be able to lay out a window, menu, etc. using a simple GUI tool. Also, in the visual designer, I should be able to simply doubleclick (or rightclick, or whatever) on the item and have the relevant code section brought up. Again, SharpDevelop does this pretty well, but it will only bring you to the code once it is there - it won't add it (at least from the Properties tab).
If I had the time I'd probably look at contributing to SharpDevelop. But alas, two young kids and a wife in college and I spend most of my time not doing things I'd like to. Also, SharpDevelop is only available for Windows.
Applications should be portable. With Cygwin/X on a Windows machine, most *nix apps should work on any machine (exceptions are tools that directly access the OS or the hardware). Having a sane build procedure is all that is really needed for true cross platform applications.
When PCs start coming standard with multiple cores in the CPU and a gig or two of RAM, someone will develop Linux for Windows and a nice development environment.
Cygwin/X + gcc + IDE of your choice. Done.
The one thing that MS has as far as development goes is a nice integrated solution with a lot of tools. I have yet to find a free IDE with similiar features.
If the goal of Open Source is to drive a company out of business, then FOSS on Windows is a bad strategy.
If the goal of Open Source is to have as many people use/develop your software as possible, then FOSS on Windows is a good strategy.
Some of us do not have a choice as to what OS we get to use (company computers for example). Why should we be left out?
I use Cygwin/X on my computer, which enables me to take advantage of a good number of FOSS programs. If I were to develop a new program, one of my design goals would be that it run on as many different platforms as possible (including Windows), so that the greatest number of people can participate. Deliberately excluding people does nobody any good.
I understand your point. It is not an ideal setup. But if you take a few precautions, you can avoid the headaches. The measures are not outlandish and are quite within the grasp of Joe Sixpack.
FYI, I don't "bask in the glow". I am regularly checking the Windows systems. It is just that I do not find any problems with viruses and spyware when I do check.
Windows itself isn't too bad if you are behind a firewall. I have several WinXP machines behind a simple NAT firewall and I have never had a problem. Simply keep it patched, Substitute IE/Outlook/Office for Mozilla/OpenOffice.org and you are good to go.
I have been running this for 3 years and every time I run a virus scan, it always comes back negative. It's nice because I don't need to pay the Symantec-McAffee tax every year. People always ask me whay virus program I use and they are very puzzled when I say "I don't need one".
Mozilla/OpenOffice need a retail box to put on the shelves in the antivirus section at BestCircuitDepotUSA, along with a little common sense about Internet useage on the back of the box.
Financial aid is out there. There are thousands of organizations that must give the money away in order to maintain their non-profit status. Applying for them is the hard part. There are companies that will sell you a list, and they will give you your money back if you cannot get any money.
Here is a short list of scholarships and grants that I or my friends have received... 1) $500 per semester because I like to target shoot. 2) $2500 for my roomate because she was a redhead. 3) $1200 per semester for my friend who liked to rebuild Mustangs 4) $1250 per semester for my wife who is in nursing school (plus insurance for herself and two kids for $500 per quarter).
None of these programs had any real requirements other than being enrolled. All we had to do is ask for the money.
Well, how about a sub-section of a larger program?
Every embedded designer needs to do some assembly, usually for configuration registers and such. But if I were developing video processing code, I'd definitely use a C compiler. I wouldn't even think about assembly if it were a modern processor (Pentium 2,3,4, G5 PowerPC, etc). In this case, a compiler with optimizations will outperform hand coded assembly. It may not look pretty, and may actually be a bit larger, but it will run faster. Not only that, but the C code is also portable, so you aren't stuck on CompanyX's processors.
Unless you know the processor inside and out, use assembly only where you absolutely have to. Optimizing compilers will produce code to take advantage of the CPU's features, like multiple pipelines, register renaming, out of order execution, etc. Unless you know how all of this works for your processor, my bet is that the compiler will produce faster code.
Is this true even for small embedded applications?
A current C compiler are much better at taking advantage of a modern processor's pipelines and other optimizations to speed up execution. If you look at the generated assembly it looks ugly as hell and is spattered with nop's, but the code actually runs faster than if you coded the app by hand.
If you are just responding to some external signals and stuffing data through UARTs, then you really don't need a 2GHz microprocessor. But even if you are using an old microcontroller (eg. Motorola 68xxx), I would still recommend C for most of your programming because it is portable. Now you can choose from many different processors and find one that fits your application, instead of being forced to use one that is compatable with your existing assembly code.
It's all about the right tool for the job (and before anyone says G5, I don't know Gx Assembly).
Is not the G5 a PowerPC derivative (which looks a bit like the old Motorola 68xxx assembly).
Besides, coding for modern microprocessors in assembly is a waste of time and effort. A current optimizing C compiler will beat the pants of your hand assembled code for all but the most trivial programs.
My property taxes are going up roughly 27-33% a year for the past four years,
Damn, where do you live?
I'm in Central Florida (which is booming, property-wise) and my property taxes have been relatively constant (3-4% increase). IIRC, my accountant told me that the government can only adjust your property value by a certian percentage per year. This is to prevent people from being taxed out of their homes. However if you receive an equity loan or refinance, the government is free to use that value as the assessed value. This is one of the negatives of equity loans/refinancing.
As far as medical bills go, I have true insurance that kicks in for the big stuff (like being hit by a bus). I take care of all the little stuff, and I pay less than if I payed an HMO. A family plan would have cost me something like $700 per month, but I only spent about $2000 (insurance premiums + doctors and medicine) this year. Of course this only works for healthy people without chronic issues. About the only time that HMOs pay for themselves is if you plan on having children, because the cost to get a child to 2 years, when their costs go down significantly, is probably around $20,000.
My guess is that while it may be less than ideal coverage, it is... 1) Better than what they had before 2) Cost effective.
More probes may provide better results, but I think right now they are looking for the big picture. If they find something interesting from the data, they can then populate interesting areas with more probes.
I wonder what the environmentalists think about more of the probe wreckages being spread all over the oceans.
Remember that NASA received a complete change in direction from the rather useless ISS to a return to the moon (equally useless?). It will take them some time to complete the change as steering a government organization is like maneuvering a loaded oil tanker.
If you ask me, NASA should provide funds to organizations like the XPrize and let man's natural motivations (greed, glory lust, etc) provide the drive to get to the moon. NASA could also facilitate things by making things available (wind tunnels, computer modelling time, etc) to the public at a much reduced cost. This would allow individuals and small companies to test their ideas more fully and attract private funding if their ideas have merit.
From the article... There is also, inevitably, some crossover with the more banal lists of things to do before you die, even if the scientists' equivalent of visiting Everest is much more interesting. The Earth's rotation causes a 20-kilometre bulge at the equator, making Chimborazo volcano in Ecuador the highest mountain above sea level.
Wouldn't the sea bulge as well (if not more)? Perhaps they should say the point on Earth furthest from the center of mass.
Government also has no competitive pressures to innovate or operate efficiently. The best solution (in my opinion) is for local government to own the access lines and have private industry provide the service. The last mile lines should terminate at a government building where space is leased out to the service providers for their equipment. Much like the city owns the roads and business uses the roads to get to your house. This is the only solution that provides a relatively level, competetive field for large and small service providers while minimizing the role of government.
CSI is Scooby Doo for adults. I hate the fact that every single room has mood lighting and every line has to be dramatic. How do they see anything with the lights off?
It started out pretty good. The sets were nice, the hallways looked like a typical government building and they would have those impromptu meetings in the breakroom. It had a much better "workplace" feel to it. Now they work in their decorated offices that are _huge_ and filled with specimens instead of the normal, two guys to an office with white walls and flourescent lights (maybe a fake plant for some greenery).
They are trying to make every moment dramatic with lighting and script. Adding David Caruso to the cast is evidence of this. That guy does not have an off switch. I know nobody who acts like that - even the primadonnas in the lab laugh and spit food and behave like a human being most of the time. I don't watch CSI-Miami for that reason.
I think they should also show it more like how they typically work - with multiple cases going on. The character might have one thats in court, one or two in the lab waiting on results, and a new one that they are getting assigned.
The drama (and plot) should come from the interaction of the characters, not the science. The science should just be an interesting side show. When they started putting the science as the lead character, the show lost its appeal. If I want science, I'll watch Nova. I do not trust Hollywood with scientific accuracy.
Slashdot Mirror
Customer: What is a tiny-fly-eating-aircraft doing in my soup?
Waiter: I believe he is doing the backstroke.
yuk, yuk.
I can see it now, the AI in the craft decides against the Miller back at the base and heads to the local bars for a black and tan. A recovery squad is promptly formed to drag the now uncoordinated bots back to work, where they are confined to the base perimeter and reprogrammed to work off of high calorie MREs. Accountants insist on purchasing expired MREs that are unfit for human consumption, but will work fine in the bots. Engineers find that higher than normal failure rates are attributed to the bots being unable to excrete waste products on an appropriate schedule. Unanticipated effects of large unscheduled waste movements while in operation complicate the traffic situation below, devalueing the traffic monitoring bots.
It is a success if they got the data that they were looking for. Even if it failed to put the objects into their assigned orbits, this was a test launch and was designed to test the system. Glitches are bound to appear on the first flight of anything. A successful test does not always mean that the product works. It means that all the test points were touched and data retrieved for all of them.
Did you read the article? He was let go. He was sneaking into the building with the help of the other "still hired engineers". They were found out once when someone tried to fill the office they were squatting in. But they found other offices and continued to sneak in. Also, other sympathetic developers assisted them in developing testing this unoffical software using their own free time.
Maybe he mentioned something about convex curves. She broke up with him after that.
More than likely this is a small company in asia which is throwing together a bunch of old/cheap components to get a "notebook computer". Chances are that the display is crap, as well as the chassis. The disk drive is probably an old model, which does not really matter for this application (as long as it does not crash). If it is like any of their other "no-name brand" products, paying $500 is too much. Spend the extra $200 or so and get a basic HP or Toshiba. Much better value.
Because this happened 4 years ago when a typical hard drive could only store a dozen movies or so. And a 17 year old is unlikely to be able to afford a large drive (I don't know if he was working or not).
Running untrusted code as an Administrator.
Unfortuneately, for most people to be able to do their work (install programs, move files around, etc) you need to have Admin priviledges. This is a fault of both the applications (for being unable to operate in a restricted envirnoment) and the OS (for not providing adequate methods for doing real work in a restricted environment ).
Until you can install and run the typical program as a user without Admin rights, this problem will persist because the default user will need to have admin rights (to prevent a flurry of tech support calls to Redmond).
Another thing to keep in mind is that household power lines, although carrying a lower voltage, are much closer, and therefore contribute more to the local electric fields.
It's a good thing Edison won out, as to get enough electricity to power anything useful into the air over any real distance would be a huge cancer risk.
Please provide some references for this. I did a paper on this topic about 6 years ago and I could not find one study that provided a link between power and radio frequency radiation and cancer.
What I did find was a lot of people who wanted to blame someone for their ailments. I read several complaints and they all basically read "there were no carcinogens found in the soil/air/water. A percentage of the people all lived close to power lines. Therefore the power lines must be at fault." From what I understand not one of those arguments held up under scrutiny.
I did try it and it was sloooooow. It was a fairly complete IDE, tho.
What is needed are several things in my pie-in-the-sky, IDE.
1) A common interface for help searches. So when I highlight the 'foo' function, I can get a help page on it. Whether that function is in the C library or in some other installed package. The IDE needs to glue these things together.
2) The ability to create a shell application with little interactivity (Project Wizard). I want a GUI app? Press this button. Console app? Press here. Device driver? etc. SharpDevelop does this pretty well.
3) "Visual" code creation. If I am throwing together a simple GUI app, I should be able to lay out a window, menu, etc. using a simple GUI tool. Also, in the visual designer, I should be able to simply doubleclick (or rightclick, or whatever) on the item and have the relevant code section brought up. Again, SharpDevelop does this pretty well, but it will only bring you to the code once it is there - it won't add it (at least from the Properties tab).
If I had the time I'd probably look at contributing to SharpDevelop. But alas, two young kids and a wife in college and I spend most of my time not doing things I'd like to. Also, SharpDevelop is only available for Windows.
Applications should be portable. With Cygwin/X on a Windows machine, most *nix apps should work on any machine (exceptions are tools that directly access the OS or the hardware). Having a sane build procedure is all that is really needed for true cross platform applications.
When PCs start coming standard with multiple cores in the CPU and a gig or two of RAM, someone will develop Linux for Windows and a nice development environment.
Cygwin/X + gcc + IDE of your choice. Done.
The one thing that MS has as far as development goes is a nice integrated solution with a lot of tools. I have yet to find a free IDE with similiar features.
If the goal of Open Source is to drive a company out of business, then FOSS on Windows is a bad strategy.
If the goal of Open Source is to have as many people use/develop your software as possible, then FOSS on Windows is a good strategy.
Some of us do not have a choice as to what OS we get to use (company computers for example). Why should we be left out?
I use Cygwin/X on my computer, which enables me to take advantage of a good number of FOSS programs. If I were to develop a new program, one of my design goals would be that it run on as many different platforms as possible (including Windows), so that the greatest number of people can participate. Deliberately excluding people does nobody any good.
I understand your point. It is not an ideal setup. But if you take a few precautions, you can avoid the headaches. The measures are not outlandish and are quite within the grasp of Joe Sixpack.
FYI, I don't "bask in the glow". I am regularly checking the Windows systems. It is just that I do not find any problems with viruses and spyware when I do check.
Windows itself isn't too bad if you are behind a firewall. I have several WinXP machines behind a simple NAT firewall and I have never had a problem. Simply keep it patched, Substitute IE/Outlook/Office for Mozilla/OpenOffice.org and you are good to go.
I have been running this for 3 years and every time I run a virus scan, it always comes back negative. It's nice because I don't need to pay the Symantec-McAffee tax every year. People always ask me whay virus program I use and they are very puzzled when I say "I don't need one".
Mozilla/OpenOffice need a retail box to put on the shelves in the antivirus section at BestCircuitDepotUSA, along with a little common sense about Internet useage on the back of the box.
Financial aid is out there. There are thousands of organizations that must give the money away in order to maintain their non-profit status. Applying for them is the hard part. There are companies that will sell you a list, and they will give you your money back if you cannot get any money.
Here is a short list of scholarships and grants that I or my friends have received...
1) $500 per semester because I like to target shoot.
2) $2500 for my roomate because she was a redhead.
3) $1200 per semester for my friend who liked to rebuild Mustangs
4) $1250 per semester for my wife who is in nursing school (plus insurance for herself and two kids for $500 per quarter).
None of these programs had any real requirements other than being enrolled. All we had to do is ask for the money.
Well, how about a sub-section of a larger program?
Every embedded designer needs to do some assembly, usually for configuration registers and such. But if I were developing video processing code, I'd definitely use a C compiler. I wouldn't even think about assembly if it were a modern processor (Pentium 2,3,4, G5 PowerPC, etc). In this case, a compiler with optimizations will outperform hand coded assembly. It may not look pretty, and may actually be a bit larger, but it will run faster. Not only that, but the C code is also portable, so you aren't stuck on CompanyX's processors.
Unless you know the processor inside and out, use assembly only where you absolutely have to. Optimizing compilers will produce code to take advantage of the CPU's features, like multiple pipelines, register renaming, out of order execution, etc. Unless you know how all of this works for your processor, my bet is that the compiler will produce faster code.
Is this true even for small embedded applications?
A current C compiler are much better at taking advantage of a modern processor's pipelines and other optimizations to speed up execution. If you look at the generated assembly it looks ugly as hell and is spattered with nop's, but the code actually runs faster than if you coded the app by hand.
If you are just responding to some external signals and stuffing data through UARTs, then you really don't need a 2GHz microprocessor. But even if you are using an old microcontroller (eg. Motorola 68xxx), I would still recommend C for most of your programming because it is portable. Now you can choose from many different processors and find one that fits your application, instead of being forced to use one that is compatable with your existing assembly code.
It's all about the right tool for the job (and before anyone says G5, I don't know Gx Assembly).
Is not the G5 a PowerPC derivative (which looks a bit like the old Motorola 68xxx assembly).
Besides, coding for modern microprocessors in assembly is a waste of time and effort. A current optimizing C compiler will beat the pants of your hand assembled code for all but the most trivial programs.
My property taxes are going up roughly 27-33% a year for the past four years,
Damn, where do you live?
I'm in Central Florida (which is booming, property-wise) and my property taxes have been relatively constant (3-4% increase). IIRC, my accountant told me that the government can only adjust your property value by a certian percentage per year. This is to prevent people from being taxed out of their homes. However if you receive an equity loan or refinance, the government is free to use that value as the assessed value. This is one of the negatives of equity loans/refinancing.
As far as medical bills go, I have true insurance that kicks in for the big stuff (like being hit by a bus). I take care of all the little stuff, and I pay less than if I payed an HMO. A family plan would have cost me something like $700 per month, but I only spent about $2000 (insurance premiums + doctors and medicine) this year. Of course this only works for healthy people without chronic issues. About the only time that HMOs pay for themselves is if you plan on having children, because the cost to get a child to 2 years, when their costs go down significantly, is probably around $20,000.
My guess is that while it may be less than ideal coverage, it is...
1) Better than what they had before
2) Cost effective.
More probes may provide better results, but I think right now they are looking for the big picture. If they find something interesting from the data, they can then populate interesting areas with more probes.
I wonder what the environmentalists think about more of the probe wreckages being spread all over the oceans.
Remember that NASA received a complete change in direction from the rather useless ISS to a return to the moon (equally useless?). It will take them some time to complete the change as steering a government organization is like maneuvering a loaded oil tanker.
/end "if I were king" speech.
If you ask me, NASA should provide funds to organizations like the XPrize and let man's natural motivations (greed, glory lust, etc) provide the drive to get to the moon. NASA could also facilitate things by making things available (wind tunnels, computer modelling time, etc) to the public at a much reduced cost. This would allow individuals and small companies to test their ideas more fully and attract private funding if their ideas have merit.
From the article...
There is also, inevitably, some crossover with the more banal lists of things to do before you die, even if the scientists' equivalent of visiting Everest is much more interesting. The Earth's rotation causes a 20-kilometre bulge at the equator, making Chimborazo volcano in Ecuador the highest mountain above sea level.
Wouldn't the sea bulge as well (if not more)? Perhaps they should say the point on Earth furthest from the center of mass.
Government also has no competitive pressures to innovate or operate efficiently. The best solution (in my opinion) is for local government to own the access lines and have private industry provide the service. The last mile lines should terminate at a government building where space is leased out to the service providers for their equipment. Much like the city owns the roads and business uses the roads to get to your house. This is the only solution that provides a relatively level, competetive field for large and small service providers while minimizing the role of government.
CSI is Scooby Doo for adults. I hate the fact that every single room has mood lighting and every line has to be dramatic. How do they see anything with the lights off?
It started out pretty good. The sets were nice, the hallways looked like a typical government building and they would have those impromptu meetings in the breakroom. It had a much better "workplace" feel to it. Now they work in their decorated offices that are _huge_ and filled with specimens instead of the normal, two guys to an office with white walls and flourescent lights (maybe a fake plant for some greenery).
They are trying to make every moment dramatic with lighting and script. Adding David Caruso to the cast is evidence of this. That guy does not have an off switch. I know nobody who acts like that - even the primadonnas in the lab laugh and spit food and behave like a human being most of the time. I don't watch CSI-Miami for that reason.
I think they should also show it more like how they typically work - with multiple cases going on. The character might have one thats in court, one or two in the lab waiting on results, and a new one that they are getting assigned.
The drama (and plot) should come from the interaction of the characters, not the science. The science should just be an interesting side show. When they started putting the science as the lead character, the show lost its appeal. If I want science, I'll watch Nova. I do not trust Hollywood with scientific accuracy.
Anyway, enough CSI bashing. CSI is on - Gotta go!