You have to watch what you donate and give away to employees. You want to give away equipment that can still be maintained in a usable state. Once, I had to get rid of a bunch of obsolete monitors, and a group of employees were actively requesting them. Instead, I donated them to an Electrical Engineering professor at a local college. He tested each of them before doing anything with them. One caught on fire and caused a serious mess. Big problem! The Electrical Engineering professor was skilled (and ready) for this sort of thing, so it was okay in the end. If I gave these monitors to employees, their houses could have burned down!!!
Lesson: If you give untrained employees or volunteer organizations equipment, make sure it works! Sure you can give the stuff away with a "no guarantees" label. However, your employees are still expecting "safe" equipment that reasonably works. Unless you are confident that you are giving away "good" kit, only send the equipment to trained professionals.
Changing peoples' keyboards can have a significant impact on the bottom-line results for your company. Rearranging keys on someone's keyboard can really slow them down. Also, certain job types and people retrain far more slowly than others. People in highly stressful jobs, managers, and older workers do not pick up on changes quickly. Giving a user their old keyboard back is a zero-cost change, and won't reduce their productivity. Unless you have a pressing reason not to do it, then let them have their old keyboard.
Incidentally, when purchasing new computers, I check the keyboard layout before purchase. Annoying users is a poor career strategy.
Multi-processor programming is becoming a real force / nightmare. Dual-core is only the beginning. At the rate AMD and Intel are moving, we will have Niagara like chips in our home PCs. Already, AMD and Intel have quad-core processors, and are talking about dual quad core (8 core) computers. The average program just can't scale well to 8 cores. Most programs, programming languages, and algorithms don't scale well with increasing the number of cores.
Fortress is proposing a language to automate that scaling. They are discussing language features to deal with multi-CPU systems, where multiple memory banks are present. AMD's multi-CPU system's (Opteron) with HyperTransport each have a separate memory banks for each processor. It makes sense to allocate the half of the array used by CPU #1 in CPU #1's memory bank, and the other half used by CPU #2 in CPU #2's memory bank. Then the threads should be split so first pair of cores on CPU #1 work in the first half of the array, and the second pair of cores on CPU #2 work on the activities related to CPU #2. Currently, all these multi-processor mapping activities happens manually, and it really sucks. It would be wonderful if programming languages supported this activity automatically.
I don't know if Fortress is the answer to the multi-core / multi-CPU problem. I hope something is. The computing world needs a solution.
Hypothetically, then what if Microsoft made a new version of Word or Excel that was so awful nobody wanted to pay for it, and didn't work as well as their previous version that all business people already had.
Actually, Microsoft came close to this at one point. The company I was working for had computers running Word for DOS. They created hundreds of forms written with Word for DOS. The forms were all ISO-9000 quality control forms, so we had to keep them. We wanted to upgrade the computers to Windows. All Word had to do was open the Word for DOS file. It couldn't. We actually had to buy WordPerfect for Windows to open the Word for DOS files.
If you are really in a 2000 degree uncontrolled environment, you will find all your cables melt. The only thing that might last is the Teflon cables, which are really expensive. Keep in mind any spilled molten steel/aluminum/magnesium/whatever will melt through stuff that the computer industry thinks is indestructible. Also, the industrial guys like moving equipment around. Plan on them rearranging anything that looks like it can be remotely moved.
My solution: after experimenting with conduit covered wires, we concluded cables were too expensive. We went all wireless. We used special antennas, and located everything so they got clear signals. (Interference is a bitch, and RF doesn't go around 5000 ton machines.)
Obviously, the only solution to the Analog hole is Digital Eyeballs. Everyone needs to have their eyes replaced with suitably DRM encumbered devices that are uncrackable. Then the high definition TV can be fed directly to your brain, the connection will be secure, and the MPAA will be rich!!!
A few years ago, the running gag at college was to e-mail a harmless sounding link to a teacher. When they opened it, they got one of those porn sites. When they closed it, you got 10 more windows, and so on.
I didn't hear about the gag being actually used to often, but all the IT trained people knew about it. Given this story, you could really kill a person's career with this gag...
Intel's support infrastructure also includes some of the best semiconductor fabrication facilities in the business. Intel has consistently held a significant process advantage at its fabs (fabrication facilities) over the life of the x86 architecture. Essentially, no one else can deliver the volume and performance of chips that Intel can. Even AMD is struggling to compete against Intel (90 nm vs 60 nm).
The process advantage means Intel can get a horrible architecture (x86) to perform acceptably at a decent price/performance point. RISC chips, while faster, require different software. People aren't going to change their software unless a good reason exists. The process advantage of Intel, means that Intel can sell good processors at a reasonable price. Given that, why switch? The x86 is even clobbering Intel's own Itanium (Itanic) architecture in terms of sales.
Other hardware vendors are competitive in market segments that place very high values on particular system metrics. For instance, the ARM processor is very competitive for low power dissipations and 32-bit applications. The 8-bit embedded microcontrollers (PIC, 8051) are really cheap. RISC chips still dominate the high performance computing market.
Wasn't there a problem in Afghanistan where the locals were swiping the USB Memory Keys / Hard Drives and reselling them on the local flee market? Military secrets included...
If you had an open Wireless Access Point, over the course of a couple of months, how many people on average would use it? How many connection attempts would be made? How many hours per attempt?
If you were in a big city, I bet you would record a large number of hits. I have friends that were disappointed when the "free" internet stopped working. You don't even have to be in an apartment building.
It is quite difficult to work in industries were Windows Vista might be used, and not wind up with a pretty mean-spirited anti-Microsoft argument. Typically the train of reasoning goes like this:
1. Power plant uses Windows PC's to monitor "x". 2. If "x" can't be monitored, we shut the power plant down. This is "fail-safe". 3. If enough power plants shut down, then we have to shut down the power grid. Shutting down the power grid affects the entire east-coast. When the power grid is shut-down, we automatically shut down all power plants. This is a fail-safe response. After the power grid is shutdown, it takes a few days to restart things. 4. If we shut down the grid, then several people will die (via indirect sequences of events). At a minimum, many people will be placed in high-risk situations, and large numbers will be inconvenienced.
What would it take to shutdown a network of identical Windows PC's making up a power system? A piece of malware, a rogue anti-virus update, etc. It really wouldn't take all that much to wipe out the power grid for the east coast. A series of inept coincidences could potentially succeed.
As a Professional Engineer, a person who is supposed to be able to advise companies on this stuff, it is extremely difficult to avoid sounding excessively alarmist. I work on industrial applications that are supposed to be fairly high-reliability. It is very difficult to keep Windows PCs isolated from the outside world. If you don't isolate the PC's, then you are vulnerable to Windows service-packs and Windows Anti-Virus software shutting down your production line. How do you even explain the problem to people? Everyone uses a Windows PC, and a Windows PC could never hurt them, right?
What do I recommend? I don't know the answer. Mostly, I try not to think about it too much. With the large amounts of specialized Windows software, it is difficult to think of any easy fixes.
Many industrial and medical applications run on Windows. You forget that Windows NT was advertised as a high-security C3 operating system. Many applications were ported on this advertising. Some of the lock-down permissions in Windows NT were pretty draconian, and worked really well.
With Windows Vista, Microsoft appears to be completely abandoning any pretense of high-reliability.
Many industrial and medical applications have fairly high reliability requirements. Using commodity software and hardware has some cost and reliability advantages. It is easy to source replacement parts, and implement hardware redundancy. Being able to easily obtain replacement hardware is a big advantage if downtime costs are large.
The problem is that Microsoft appears to have abandoned the high-reliability sector. Windows XP has a continuous stream of rolling updates for both XP and the Anti-Virus packages. The result is that your high-reliability application can stop working for no apparent reason. From all indications, Windows Vista will make this worse.
Recently, I have been looking harder and harder at Linux. Linux offers a much more stable platform, and I can customize the installation to make it much more difficult to corrupt. The issue is that such a high software investment has been placed in specialized Windows solutions, that it is difficult to port everything to another operating system overnight.
My experience is every parallel design fails to achieve its promised potential. Software programmers have a hard time making use of multiple core and multiple processor systems. Almost all programmers have learned software development on single-core/single-processor machines. The result is almost all programs run well on single-core/single-processor computers. Dual-core development isn't too difficult, because many programs have certain natural parallelisms that make it easy to keep two cores busy. If you look at most parallel processing curves, times get really tough when you have more the 4 cores. Not many programs are easy to parallelize at the 4+ core level. The Cell on the PS3 has 7 cores.
A general lack of availability of multi-processor/multi-core developers, and the high-difficulty level of multi-thread software development, will mean that the PS3 development runs late and over-budget. This is a big problem for someone thinking of developing software for the PS3.
There is no way that Microsoft could co-opt Linux to replace it's brain-damaged O/S development efforts. Linux and Windows have simply developed too far in incompatible directions. Further, the day that Microsoft even mentioned "maybe we should switch to Linux", everyone would immediately stop purchasing from Microsoft. Linux is free.
A more likely scenario is Microsoft going the way of the old IBM mainframes. Initially, IBM really dominated all computing of any consequence. The IBM vision was not a low-cost vision. As such, the old DEC PDP and VAX minicomputers were developed and sold at much lower prices. Eventually, the IBM PC came along. The PC was cheaper than the mini-computer, and had a new totally different and incompatible operating system. The PC eventually became the dominant form of computer.
Today, we see Linux in a huge number of embedded applications (like the PS3 or the TiVO). These are less expensive computers than a PC. Eventually, we may see that these embedded machines are present in greater numbers than the PCs running Windows. When this happens, this will be the end of Windows as the dominant operating system. The PC marketplace will have fundamentally changed at that point. This shift may not be that far away. Google is already running Linux, and the ranks of embedded devices are continually increasing. Linux devices may already outnumber Windows PCs. We just haven't noticed it yet, because we are so PC focused.
Computers are almost 100% efficient as space heaters. Almost every watt consumed gets converted to heat.
The energy in the light radiated from the monitor or from the LEDs in the computer case is very small compared to the energy consumed by the computer. Computers do no useful physical work. The result is that almost all energy consumed by a computer is converted to heat.
Vista may be the last wide-scale operating system deployment ever. It requires so much work to regression test all of the in-house and external applications, that some companies may not bother. Instead, you can just use Virtualization, and run all the old applications on the old operating systems in a compatibility mode.
With virtualization, the application only needs to get tested on one operating system. No more O/S upgrades for all those specialized applications.
Welcome to Microsoft. Microsoft can't stop itself from creating programming dead ends. Below is history by product. Can anyone name a few products or API's that I missed?
Windows: Windows 2.0 to Windows 3.0: triggered rewrites. Windows 3.1 to Windows/95: triggered rewrites. Even Windows/NT 3.1 to Windows/95 even triggered rewrites.
Basic: QuickBasic 4.x to VisualBasic 1.0: triggered rewrites VisualBasic 3.0 (16-bit) to VisualBasic 4.0 (32-bit): triggered rewrites. VisualBasic 5/6/7 to VisualBasic.NET: triggered rewrites.
C++: Microsoft C 6.0 to Visual C++ 1.0: triggered rewrites. Visual C++ 1.x to Visual C++ 2.0 (32-bit): triggered rewrites. Visual C++ 2005: marked almost the entire Standard C library as deprecated.
I bumped into an electrical circuit that was dead easy. It was one of the simplest things that I built in a long time. It got patented.
Essentially, anyone skilled in the field knew that the technique didn't work for the application intended. As such, no one bothered patenting (or even documenting) it. Along came a professor one day. He used a limited sample, and concluded the technique worked great. If you expanded the sample size, the technique failed: in every test run we did.
The big problem for me was this: The professor essentially created a blocking patent. Any device used for the application would need to have much in common with his implementation. In fact, it would be difficult to create a new device that didn't use some of the same key ideas of his circuit. Interestingly, all the commercial products in the field were more complicated than his, and all used similar technologies.
The patent office issued a patent that can be used to annoy (and perhaps make money from) everyone else in that field. The professor succeeded in doing two things: a) wasting a whole of research and industry money, and b) creating a legal obstacle for anyone else creating a better solution.
I am a firm believer in making the obviousness test tougher.
> Every line written is a ticking patent timebomb.
Actually, it is worse than this. The ticking time bomb includes every line of code that runs on any computer your software touches. Almost all software projects depend on outside packages to run. These include libraries, operating systems, programming environments, and even other software packages. To force a change in your project, all someone needs to do is force a change in any related piece of software. Worse, the change in the other piece of software might have nothing to do with your package. The patent infringement could simply be triggered by how or where the unrelated piece of software is used.
For instance, if I write a web application using ActiveX on Internet Explorer, then some company (Eolas) can sue Microsoft (not me) and force Microsoft to change its code, and now I need to change my application that uses ActiveX.
You can also get in trouble for patents if you are a Canadian company (RIM), with your server in Canada (no s/w patents here). If your customers are in the U.S., and the software on your server violates a U.S. patent held by a U.S. company (NTP), then you can still have to pay.
Finally, if you write a piece of software that implements a patent-free algorithm (like principal components analysis), and your customer implements it into a business or industrial process (polymer processing), then your customer can be sued for using your software.
Software patents are a complete cancer. You can't design any kind of complex system without potentially violating someones patent on something.
In Canada, the government collects a small tax on the sale of each recordable CD. In concept, this money is distributed amongst the artists. When I last looked at the issue a while ago, the government still hadn't figured out how to divide up the money. As such, it was sitting in a government account lost in the "general revenue" pit.
Essentially, no fair way of distributing the money really exists.
Slashdot Mirror
You have to watch what you donate and give away to employees. You want to give away equipment that can still be maintained in a usable state. Once, I had to get rid of a bunch of obsolete monitors, and a group of employees were actively requesting them. Instead, I donated them to an Electrical Engineering professor at a local college. He tested each of them before doing anything with them. One caught on fire and caused a serious mess. Big problem! The Electrical Engineering professor was skilled (and ready) for this sort of thing, so it was okay in the end. If I gave these monitors to employees, their houses could have burned down!!!
Lesson: If you give untrained employees or volunteer organizations equipment, make sure it works! Sure you can give the stuff away with a "no guarantees" label. However, your employees are still expecting "safe" equipment that reasonably works. Unless you are confident that you are giving away "good" kit, only send the equipment to trained professionals.
Changing peoples' keyboards can have a significant impact on the bottom-line results for your company. Rearranging keys on someone's keyboard can really slow them down. Also, certain job types and people retrain far more slowly than others. People in highly stressful jobs, managers, and older workers do not pick up on changes quickly. Giving a user their old keyboard back is a zero-cost change, and won't reduce their productivity. Unless you have a pressing reason not to do it, then let them have their old keyboard.
Incidentally, when purchasing new computers, I check the keyboard layout before purchase. Annoying users is a poor career strategy.
Multi-processor programming is becoming a real force / nightmare. Dual-core is only the beginning. At the rate AMD and Intel are moving, we will have Niagara like chips in our home PCs. Already, AMD and Intel have quad-core processors, and are talking about dual quad core (8 core) computers. The average program just can't scale well to 8 cores. Most programs, programming languages, and algorithms don't scale well with increasing the number of cores.
Fortress is proposing a language to automate that scaling. They are discussing language features to deal with multi-CPU systems, where multiple memory banks are present. AMD's multi-CPU system's (Opteron) with HyperTransport each have a separate memory banks for each processor. It makes sense to allocate the half of the array used by CPU #1 in CPU #1's memory bank, and the other half used by CPU #2 in CPU #2's memory bank. Then the threads should be split so first pair of cores on CPU #1 work in the first half of the array, and the second pair of cores on CPU #2 work on the activities related to CPU #2. Currently, all these multi-processor mapping activities happens manually, and it really sucks. It would be wonderful if programming languages supported this activity automatically.
I don't know if Fortress is the answer to the multi-core / multi-CPU problem. I hope something is. The computing world needs a solution.
Actually, Microsoft came close to this at one point. The company I was working for had computers running Word for DOS. They created hundreds of forms written with Word for DOS. The forms were all ISO-9000 quality control forms, so we had to keep them. We wanted to upgrade the computers to Windows. All Word had to do was open the Word for DOS file. It couldn't. We actually had to buy WordPerfect for Windows to open the Word for DOS files.
It's Microsoft software. We will buy it!!!!!!
If you are really in a 2000 degree uncontrolled environment, you will find all your cables melt. The only thing that might last is the Teflon cables, which are really expensive. Keep in mind any spilled molten steel/aluminum/magnesium/whatever will melt through stuff that the computer industry thinks is indestructible. Also, the industrial guys like moving equipment around. Plan on them rearranging anything that looks like it can be remotely moved.
My solution: after experimenting with conduit covered wires, we concluded cables were too expensive. We went all wireless. We used special antennas, and located everything so they got clear signals. (Interference is a bitch, and RF doesn't go around 5000 ton machines.)
Obviously, the only solution to the Analog hole is Digital Eyeballs. Everyone needs to have their eyes replaced with suitably DRM encumbered devices that are uncrackable. Then the high definition TV can be fed directly to your brain, the connection will be secure, and the MPAA will be rich!!!
A few years ago, the running gag at college was to e-mail a harmless sounding link to a teacher. When they opened it, they got one of those porn sites. When they closed it, you got 10 more windows, and so on.
...
I didn't hear about the gag being actually used to often, but all the IT trained people knew about it. Given this story, you could really kill a person's career with this gag
Intel's support infrastructure also includes some of the best semiconductor fabrication facilities in the business. Intel has consistently held a significant process advantage at its fabs (fabrication facilities) over the life of the x86 architecture. Essentially, no one else can deliver the volume and performance of chips that Intel can. Even AMD is struggling to compete against Intel (90 nm vs 60 nm).
The process advantage means Intel can get a horrible architecture (x86) to perform acceptably at a decent price/performance point. RISC chips, while faster, require different software. People aren't going to change their software unless a good reason exists. The process advantage of Intel, means that Intel can sell good processors at a reasonable price. Given that, why switch? The x86 is even clobbering Intel's own Itanium (Itanic) architecture in terms of sales.
Other hardware vendors are competitive in market segments that place very high values on particular system metrics. For instance, the ARM processor is very competitive for low power dissipations and 32-bit applications. The 8-bit embedded microcontrollers (PIC, 8051) are really cheap. RISC chips still dominate the high performance computing market.
http://it.slashdot.org/article.pl?sid=06/04/13/041 7200
Wasn't there a problem in Afghanistan where the locals were swiping the USB Memory Keys / Hard Drives and reselling them on the local flee market? Military secrets included ...
Because you can't stick the password on the monitor. The annoying IT security types complain when that happens.
I just tell my friends my password, so they can remember it for me. I also keep a copy of it in the bottom of my desk drawer, underneath my pens.
If you had an open Wireless Access Point, over the course of a couple of months, how many people on average would use it? How many connection attempts would be made? How many hours per attempt?
If you were in a big city, I bet you would record a large number of hits. I have friends that were disappointed when the "free" internet stopped working. You don't even have to be in an apartment building.
It is quite difficult to work in industries were Windows Vista might be used, and not wind up with a pretty mean-spirited anti-Microsoft argument. Typically the train of reasoning goes like this:
1. Power plant uses Windows PC's to monitor "x".
2. If "x" can't be monitored, we shut the power plant down. This is "fail-safe".
3. If enough power plants shut down, then we have to shut down the power grid. Shutting down the power grid affects the entire east-coast. When the power grid is shut-down, we automatically shut down all power plants. This is a fail-safe response. After the power grid is shutdown, it takes a few days to restart things.
4. If we shut down the grid, then several people will die (via indirect sequences of events). At a minimum, many people will be placed in high-risk situations, and large numbers will be inconvenienced.
What would it take to shutdown a network of identical Windows PC's making up a power system? A piece of malware, a rogue anti-virus update, etc. It really wouldn't take all that much to wipe out the power grid for the east coast. A series of inept coincidences could potentially succeed.
As a Professional Engineer, a person who is supposed to be able to advise companies on this stuff, it is extremely difficult to avoid sounding excessively alarmist. I work on industrial applications that are supposed to be fairly high-reliability. It is very difficult to keep Windows PCs isolated from the outside world. If you don't isolate the PC's, then you are vulnerable to Windows service-packs and Windows Anti-Virus software shutting down your production line. How do you even explain the problem to people? Everyone uses a Windows PC, and a Windows PC could never hurt them, right?
What do I recommend? I don't know the answer. Mostly, I try not to think about it too much. With the large amounts of specialized Windows software, it is difficult to think of any easy fixes.
Many industrial and medical applications run on Windows. You forget that Windows NT was advertised as a high-security C3 operating system. Many applications were ported on this advertising. Some of the lock-down permissions in Windows NT were pretty draconian, and worked really well.
With Windows Vista, Microsoft appears to be completely abandoning any pretense of high-reliability.
Many industrial and medical applications have fairly high reliability requirements. Using commodity software and hardware has some cost and reliability advantages. It is easy to source replacement parts, and implement hardware redundancy. Being able to easily obtain replacement hardware is a big advantage if downtime costs are large.
The problem is that Microsoft appears to have abandoned the high-reliability sector. Windows XP has a continuous stream of rolling updates for both XP and the Anti-Virus packages. The result is that your high-reliability application can stop working for no apparent reason. From all indications, Windows Vista will make this worse.
Recently, I have been looking harder and harder at Linux. Linux offers a much more stable platform, and I can customize the installation to make it much more difficult to corrupt. The issue is that such a high software investment has been placed in specialized Windows solutions, that it is difficult to port everything to another operating system overnight.
Thank you for the tip. Merry Christmas!
My experience is every parallel design fails to achieve its promised potential. Software programmers have a hard time making use of multiple core and multiple processor systems. Almost all programmers have learned software development on single-core/single-processor machines. The result is almost all programs run well on single-core/single-processor computers. Dual-core development isn't too difficult, because many programs have certain natural parallelisms that make it easy to keep two cores busy. If you look at most parallel processing curves, times get really tough when you have more the 4 cores. Not many programs are easy to parallelize at the 4+ core level. The Cell on the PS3 has 7 cores.
A general lack of availability of multi-processor/multi-core developers, and the high-difficulty level of multi-thread software development, will mean that the PS3 development runs late and over-budget. This is a big problem for someone thinking of developing software for the PS3.
There is no way that Microsoft could co-opt Linux to replace it's brain-damaged O/S development efforts. Linux and Windows have simply developed too far in incompatible directions. Further, the day that Microsoft even mentioned "maybe we should switch to Linux", everyone would immediately stop purchasing from Microsoft. Linux is free.
A more likely scenario is Microsoft going the way of the old IBM mainframes. Initially, IBM really dominated all computing of any consequence. The IBM vision was not a low-cost vision. As such, the old DEC PDP and VAX minicomputers were developed and sold at much lower prices. Eventually, the IBM PC came along. The PC was cheaper than the mini-computer, and had a new totally different and incompatible operating system. The PC eventually became the dominant form of computer.
Today, we see Linux in a huge number of embedded applications (like the PS3 or the TiVO). These are less expensive computers than a PC. Eventually, we may see that these embedded machines are present in greater numbers than the PCs running Windows. When this happens, this will be the end of Windows as the dominant operating system. The PC marketplace will have fundamentally changed at that point. This shift may not be that far away. Google is already running Linux, and the ranks of embedded devices are continually increasing. Linux devices may already outnumber Windows PCs. We just haven't noticed it yet, because we are so PC focused.
Computers are almost 100% efficient as space heaters. Almost every watt consumed gets converted to heat.
The energy in the light radiated from the monitor or from the LEDs in the computer case is very small compared to the energy consumed by the computer. Computers do no useful physical work. The result is that almost all energy consumed by a computer is converted to heat.
Vista may be the last wide-scale operating system deployment ever. It requires so much work to regression test all of the in-house and external applications, that some companies may not bother. Instead, you can just use Virtualization, and run all the old applications on the old operating systems in a compatibility mode.
With virtualization, the application only needs to get tested on one operating system. No more O/S upgrades for all those specialized applications.
Welcome to Microsoft. Microsoft can't stop itself from creating programming dead ends. Below is history by product. Can anyone name a few products or API's that I missed?
.NET: triggered rewrites.
...
Windows:
Windows 2.0 to Windows 3.0: triggered rewrites.
Windows 3.1 to Windows/95: triggered rewrites.
Even Windows/NT 3.1 to Windows/95 even triggered rewrites.
Basic:
QuickBasic 4.x to VisualBasic 1.0: triggered rewrites
VisualBasic 3.0 (16-bit) to VisualBasic 4.0 (32-bit): triggered rewrites.
VisualBasic 5/6/7 to VisualBasic
C++:
Microsoft C 6.0 to Visual C++ 1.0: triggered rewrites.
Visual C++ 1.x to Visual C++ 2.0 (32-bit): triggered rewrites.
Visual C++ 2005: marked almost the entire Standard C library as deprecated.
Canceled Languages:
VisualJ++ (Java), QuickPascal, Fortran,
I bumped into an electrical circuit that was dead easy. It was one of the simplest things that I built in a long time. It got patented.
Essentially, anyone skilled in the field knew that the technique didn't work for the application intended. As such, no one bothered patenting (or even documenting) it. Along came a professor one day. He used a limited sample, and concluded the technique worked great. If you expanded the sample size, the technique failed: in every test run we did.
The big problem for me was this: The professor essentially created a blocking patent. Any device used for the application would need to have much in common with his implementation. In fact, it would be difficult to create a new device that didn't use some of the same key ideas of his circuit. Interestingly, all the commercial products in the field were more complicated than his, and all used similar technologies.
The patent office issued a patent that can be used to annoy (and perhaps make money from) everyone else in that field. The professor succeeded in doing two things: a) wasting a whole of research and industry money, and b) creating a legal obstacle for anyone else creating a better solution.
I am a firm believer in making the obviousness test tougher.
Either that, or engineering couldn't figure out if they would rather complain about the promotion, or try to win the promotion!
> Every line written is a ticking patent timebomb.
Actually, it is worse than this. The ticking time bomb includes every line of code that runs on any computer your software touches. Almost all software projects depend on outside packages to run. These include libraries, operating systems, programming environments, and even other software packages. To force a change in your project, all someone needs to do is force a change in any related piece of software. Worse, the change in the other piece of software might have nothing to do with your package. The patent infringement could simply be triggered by how or where the unrelated piece of software is used.
For instance, if I write a web application using ActiveX on Internet Explorer, then some company (Eolas) can sue Microsoft (not me) and force Microsoft to change its code, and now I need to change my application that uses ActiveX.
You can also get in trouble for patents if you are a Canadian company (RIM), with your server in Canada (no s/w patents here). If your customers are in the U.S., and the software on your server violates a U.S. patent held by a U.S. company (NTP), then you can still have to pay.
Finally, if you write a piece of software that implements a patent-free algorithm (like principal components analysis), and your customer implements it into a business or industrial process (polymer processing), then your customer can be sued for using your software.
Software patents are a complete cancer. You can't design any kind of complex system without potentially violating someones patent on something.
In Canada, the government collects a small tax on the sale of each recordable CD. In concept, this money is distributed amongst the artists. When I last looked at the issue a while ago, the government still hadn't figured out how to divide up the money. As such, it was sitting in a government account lost in the "general revenue" pit.
Essentially, no fair way of distributing the money really exists.