They're not only about recouping R&D costs. They are partly about that, though.
The primary idea behind a patent is not to prevent the sharing of innovations but to promote it. Someone who has done big R&D and wants to protect his work has one natural way of doing so: to keep it a complete secret. A patent is supposed to keep anyone from copying his work for a relatively short amount of time in exchange for publication of exactly how to replicate the work.
This all made sense when the economy was mostly mining and manufacturing, the rate of innovation was much slower, patents were granted for industrial processes or entirely new classes of consumer products, and patents actually expired.
Now we have patent extensions, every little tweak of inactive ingredients in a pill gets a new patent, Amazon can patent storing customer information in a database (one-click checkout), Microsoft is threatening to sue people for patents they probably have no rights to have, SCO is suing for patents, trade secrets, and copyrights for things they probably don't even own, and real innovation according to most CEOs of large companies is about how to screw the investors out of a bigger options package.
Try triple-dip. They already collect "royalties" from most of the blank media companies. That's right,they want the cost of the media to be partly theirs, then a tax on it that is all theirs, then to sue you when you buy a song and write it out to of the disks that you've already paid twice to put music onto.
There might be something that says that if you drive 13 or 14 days straight without a day off that you have to have an extended weekend or something. It's more likely a company scheduling issue within Scheider.
My brother-in-law drives 4 to 5.5 days during the week and is home 1.5 to 3 days on/around the weekend.
Most consumers don't buy phones from the manufacturers. They buy phones from the cell company at a discount in exchange for a lock-in contract. The bigger cell companies can, and often do, removed features from stock phone firmware or have the manufacturer do so for them.
So in short, if it's developed, have fun buying your phone at full price and getting your carrier to let you use that fdeature on their network under their contract terms.
It'd be a mess, but they wouldn't be the only ones refusing to move to GPLv3. All the people refusing to move to 3 could still share each other's GPLv2-only code.
Also, any code that still says GPLv2 or later, or whose author expressly puts it under GPLv2 for Novell and GPLv3 for the FSF they could use.
Likewise, BTW, no contributions that are licensed GPLv2-only can be used in GPLv3 projects.
How many people that would be I don't know. It'd still be worse than everyone sticking with the same GPL. If GPLv3 drives a bunch of people to stick with v2, this is the kind of thing we'll all need to worry about.
Good luck finding a cheap knockoff plasma TV for $50. It's not a handbag or a watch. The name on it and the metal of which it's made are not the primary determining factors in its price.
IANATD, but my brother-in-law is and I used to work in a truck-stop once upon a time.
I believe it's either 10 hours then a minimum of 10 hours of not driving, or 8 hours then a minimum of 8 hours not driving. It's the driver's choice (or his company's, I guess) of whether to drive the extra two hours, but the extra two hours of mandatory "rest" time kicks in. That's the way it was years ago when I was a truck-stop cashier.
Of course, this is abused all the time. There are guys who get hundreds of miles ahead of their log books. They're officially in Ohio when they've already dropped off a trailer in Illinois. Some of them who do actually take the time between driving sessions don't actually rest, but at least they're not getting road hypnosis.
Driving a truck is much like anything else. Many drivers follow all the rules, because they'd rather make their money and go home still alive and still licensed, without paying any fines. Some push the limits a little, especially around weekends and holidays when the long-haul guys are just eager to see their families. Some try to screw the system as hard as they can.
Cisco's iPhone trademark dates to 1996, when it originated within Infogear Technology. The iPod dates to 2001. The iMac dates to 1998.
Now, it's true that Cisco wasn't being very protective of the name and may have failed to protect it by legal standards. They may have renamed the Linksys CIT series solely to try to keep the rights to the trademark, and may have failed to do that.
It's clear, though, that in 1996 the iPod and iMac had no influence on the cell phone, IP phone, or even the MP3 player markets. Unless someone within Infogear Technologies was a rival to Einstein and Hawking, there was no way to speculate that the iMac, iPod, then iPhone would come out of Apple over the course of a decade. To run a business on such speculation would likely be suicide anyway.
Maybe Infogear, and now Cisco (or maybe the Cisco of the future) had/has/will have an iTimeMachine, but Occam's Razor suggests coincidence over John Connor coming back to stop Skynet and discussing Apple's product lines with Bob Marshall and/or Sandy Lerner.
So it functions like an electricity-generating windmill with a dehumidifier attached. It either does it without the step in between, or by building both steps built into one device.
The fact that it is a unconventional windmill is a separate feature, although what I like to call "wingstacks" or "windwheels" have been proposed (and perhaps even built before although I can't recall seeing one for real), and vertical-axis windmills are gaining popularity.
A "wingstack" is basically a bunch of lightweight wings -- think airfoils or big kites -- in a loop on a cable. They work like a waterwheel. In a water wheel, the water provides enough power to move the wheel (and sometimes part of the water) around the axis enough that the next spoke gets into the path of the water. In a wingstack, the wings on one side adjust angle of attack enough to get lift and go up, while the ones on the other side produce downforce and go down. Since they're connected and there's a pivot point at the bottom, they move around a circuit and could generate electricity. They're actually, as I recall, instead intended to keep something like an antenna aloft at a second pivot on the upper end. It was Wired, Popular Mechanics, Scientific American, or somewhere like those that I saw these.
So, is it patentable? Is obviousness contrary to patents in Australia?
At some point we can't make the cores any faster. This has to do with roadblocks both in the current state of technology (meaning it can't be done now) and in the limits of physics (meaning that at some point it would take a major breakthrough in raw scientific understanding to make any progress at all).
If you need ten very different things in a particular order and the way each one of them is constructed depends on the result of the one before it, then it's possible that you'd be best served by one processor that's ten times as fast, but that's getting progressively harder and more expensive to do.
If you need one hundred very similar small things right now, you'd be best served by something which can deliver high volumes of the stuff. This is the kind of thing the Cell is very good at.
If you need ten different sorts of large things that may have things about them them overlap in time because the state of one doesn't necessarily depend on the state of the others, you'd be best served by ten processors each working on a partial solution the whole time, and the final piece being tied together at the end. This is the sort of thing multi-core general processors are getting better at.
One issue many people in this thread are addressing is the difference between the second and third options. Another is how to best achieve the third one, since it is not always easy to figure out which parts of which tasks are going to be independent enough to be calculated separately nor easy to tell the computer the difference once the programmer knows.
Multi-core is the only sure way to get much additional performance for the near future. Breakthroughs are wonderful and maybe even worth a wager. They're definitely not the norm, though, and therefore cannot be counted upon as a strategy. Optical computing or quantum computing may end up being widespread in a few years, and taking stock in someone working on them may pay off big. People who work in the field, though, have to hedge that bet.
Once upon a time, there were mille coins in the U.S., and my parents remember them. The individual states often minted them, which would likely not be allowed under our increasingly powerful central government of today.
We do not have anything smaller than a penny actually minted any more, specifically because each of the smaller coins experienced this same situation of costing more than its own value. Many things are still priced in half-penny or tenth of a penny denominations, especially things sold in bulk. The final price is just rounded to the nearest penny. (Or sometimes bumped up to the next penny in favor of the vendor for any fraction).
If the penny goes away, the same thing will probably still be done, only we'll be rounding to nickels or dimes.
The fat teenagers surely get some of their heft from their fat parents. The eating and exercise habits of a parent are often passed on to children. These teenagers would probably be thin if they had thin parents in the suburbs and still be fat if they had fat parents in the city.
Sure, the by-car-only lifestyle doesn't help. Lots of developments even lack sidewalks, and zoning often puts any type of grocery, arcade, bar, or any other type of business miles away from where people live. It's more about how homes are situated near businesses and schools or far away from them than whether you're in the city or the suburbs. Unfortunately, many people specifically want their homes as far away from businesses as possible. The traffic, the buildings being empty at night, the looks of commercial property, and more are given as reasons.
These trends almost certainly combine to allow children to be fat. It's not one or the other.
The people employing the programmers think the programmer should use the least resources necessary for the task assigned. Since people buy software as long as their current system doesn't make it painfully slow to use, the programmer's time is more judiciously spent fixing bugs or writing another bloated product than optimizing something that runs fast enough.
In the days of older, slower systems with less memory, programmers were more frugal with computer resources because there was a need. Then, much of the world was doing single-user, single-tasking things on ridiculously fast systems. Now, in these days of multi-tasking (most people still only use desktops one user at a time) it'd be nice to see some apps that are a bit more miserly since even a few megs per application of waste can add up.
It's still a matter of programmer time vs. benefit. A company could go broke trying to wring more application features out of less memory and fewer processor cycles. If their apps can be a little more efficient than the next company without their programmers having to take much more time to do it, that's likely an acceptable trade in the marketplace. Still, it' the feature checklist that people shop for.
Speed, memory requirements, DLL hell, and even stability are likely going to be blamed on the OS vendor more than the application vendor anyway. Sad but true.The OS should be fast, responsive, and stable on its own. It should keep an errant application from bringing down the system. In fact, killing the application that's causing problems can be the key to the OS keeping the rtest of the system stable. There's no reason to expect MS, Novell, Apple, Red Hat, or Sun to make third-party applications fast and reliable. Many people still do.
The study of crystalline structures in general is very important. LCD? Liquid Crystal Display. Timings in your PC? Crystals. E-paper? Likely the solutions will involve crystals. Self-assembly of electronics in the future? Crystals and the way they form are a big part of that, since crystals are one of the ways that things self-assemble. There are lots of examples besides these, I'm sure.
The study of snowflakes specifically has uses for weather forecasting. Ever wonder why the guy on TV says 3-5 inches of snow and it turns out to be 7? It' not necessarily because there's more water in the snow than anticipated. The shape and size of the flakes makes a sizable difference in the volume of the snow. People's lives and safety can depend on good weather forecasting these days. The budgets of road crews always does, since it costs money to field trucks overnight in anticipation of snow whether it actually falls or not.
Snow is also an excellent indicator of pollution patterns since it straps lots of particulate matter in the flakes. What's in the air actually effects the shape and size of snowflakes, too.
Besides weather forecasting and pollution tracking, snowflakes, being so plentiful, can serve as an excellent example of crystal formation in general, since it is in large part a generally applicable principle.
No, not hardware abstraction. That's what DX10 already does.
Virtualization allows a whole different OS to run inside the current one as an application. Sometimes the host OS is a stripped-down one especially designed for hosting other OSes. Other times it's a general-purpose OS like Windows.
Emulation at this level doesn't necessarily mean fully emulating the hardware. The hardware for Windows and for other x86 OSes is the same. You can emulate the IO devices for the drivers on the guest OS, while letting it use the CPU natively. You end up, if the emulation is done thinly enough, with a second driver-to-hardware call for each output operation as your main overhead. Since we're talking about sending longer and longer programs to the GPU to let it process those itself, you'll actually see the overhead go down as shader programs get larger.
Using something like Xen, Bochs, VirtualPC, Parallels, or VMware, you just maintain compatibility with your virtualization or emulation environment. That system's developers and maintainers worry about compatibility with the host OS.
It's not outrageous to suggest that big-box retail game studios could support an OS under a virtualized or close-to-hardware emulated environment. One framework that supports most PC hardware, OpenGL, and a good number of IO devices well and that uses a cheap virtualization/emulation layer might even become a de facto standard eventually if the idea proves promising in practice. DOSBox does great things for DOS games under XP. I know people who play older Windows games under Wine. It's not too far fetched that in the reasonably near future you could boot a Linux, ucLinux, FreeBSD, NetBSD, VxWorks, QNX, or even OS X instance in a window on XP or Vista and play a game there with pretty good performance.
If I was a shareholder of a publicly traded utility company (I am not, but I own shares of a fund that owns shares of some), I'd be very ready to disagree with you (and I am).
Cooperative utility organizations are supposed to be nonprofit. The publicly traded companies expect to make money, and the shareholders expect them to do so. The city in which I lived before my last move had a municipal electric system. They overbuilt capacity and started selling electric to neighboring communities, too. Even they turned a bit of an operational surplus, which instead of being taken as profit went into increased maintenance, a couple of extra positions, and a shift of some of the money into other city funds so things could be done on utility money rather than taxes.
Sorry to respond to myself, but I did a little research.
Actually, it looks like Darl doesn't like to sell at all. Larry Goldfarb and BayStar Capital Group, however, had been selling thousands of shares every few days in late 2004.
What could be really telling is who sold the most between March/April of '03 and February/March of '04, during which time the stock went from around $3 to around $17 and started to slide back down to around $8.
Darl and Kevin probably made all their money in a pump & dump, when SCO's stock skyrocketed upon suing IBM and getting investments from that capital investment group that was acting as a front for Microsoft. SCOX stock right now is $1.06, but it hasn't always been. Anyone know for sure if and when Darl sold a bunch of shares?
Slashdot Mirror
They're not only about recouping R&D costs. They are partly about that, though.
The primary idea behind a patent is not to prevent the sharing of innovations but to promote it. Someone who has done big R&D and wants to protect his work has one natural way of doing so: to keep it a complete secret. A patent is supposed to keep anyone from copying his work for a relatively short amount of time in exchange for publication of exactly how to replicate the work.
This all made sense when the economy was mostly mining and manufacturing, the rate of innovation was much slower, patents were granted for industrial processes or entirely new classes of consumer products, and patents actually expired.
Now we have patent extensions, every little tweak of inactive ingredients in a pill gets a new patent, Amazon can patent storing customer information in a database (one-click checkout), Microsoft is threatening to sue people for patents they probably have no rights to have, SCO is suing for patents, trade secrets, and copyrights for things they probably don't even own, and real innovation according to most CEOs of large companies is about how to screw the investors out of a bigger options package.
Try triple-dip. They already collect "royalties" from most of the blank media companies. That's right,they want the cost of the media to be partly theirs, then a tax on it that is all theirs, then to sue you when you buy a song and write it out to of the disks that you've already paid twice to put music onto.
There might be something that says that if you drive 13 or 14 days straight without a day off that you have to have an extended weekend or something. It's more likely a company scheduling issue within Scheider.
My brother-in-law drives 4 to 5.5 days during the week and is home 1.5 to 3 days on/around the weekend.
Most consumers don't buy phones from the manufacturers. They buy phones from the cell company at a discount in exchange for a lock-in contract. The bigger cell companies can, and often do, removed features from stock phone firmware or have the manufacturer do so for them.
So in short, if it's developed, have fun buying your phone at full price and getting your carrier to let you use that fdeature on their network under their contract terms.
It'd be a mess, but they wouldn't be the only ones refusing to move to GPLv3. All the people refusing to move to 3 could still share each other's GPLv2-only code.
Also, any code that still says GPLv2 or later, or whose author expressly puts it under GPLv2 for Novell and GPLv3 for the FSF they could use.
Likewise, BTW, no contributions that are licensed GPLv2-only can be used in GPLv3 projects.
How many people that would be I don't know. It'd still be worse than everyone sticking with the same GPL. If GPLv3 drives a bunch of people to stick with v2, this is the kind of thing we'll all need to worry about.
Good luck finding a cheap knockoff plasma TV for $50. It's not a handbag or a watch. The name on it and the metal of which it's made are not the primary determining factors in its price.
IANATD, but my brother-in-law is and I used to work in a truck-stop once upon a time.
I believe it's either 10 hours then a minimum of 10 hours of not driving, or 8 hours then a minimum of 8 hours not driving. It's the driver's choice (or his company's, I guess) of whether to drive the extra two hours, but the extra two hours of mandatory "rest" time kicks in. That's the way it was years ago when I was a truck-stop cashier.
Of course, this is abused all the time. There are guys who get hundreds of miles ahead of their log books. They're officially in Ohio when they've already dropped off a trailer in Illinois. Some of them who do actually take the time between driving sessions don't actually rest, but at least they're not getting road hypnosis.
Driving a truck is much like anything else. Many drivers follow all the rules, because they'd rather make their money and go home still alive and still licensed, without paying any fines. Some push the limits a little, especially around weekends and holidays when the long-haul guys are just eager to see their families. Some try to screw the system as hard as they can.
...except when it isn't.
How many people in California can the Feds arrest and prosecute for medical marijuana on their own now that the state refuses to do it for them?
How many people in Maine will have to get Federally-approved state driver IDs if the state refuses to issue them?
Besides, every state must have a constitution of its own, which must be approved by Congress at the time of induction into the Union.
When was the last time you voted an FBI agent or a Federal prosecutor into power?
Cisco's iPhone trademark dates to 1996, when it originated within Infogear Technology. The iPod dates to 2001. The iMac dates to 1998.
Now, it's true that Cisco wasn't being very protective of the name and may have failed to protect it by legal standards. They may have renamed the Linksys CIT series solely to try to keep the rights to the trademark, and may have failed to do that.
It's clear, though, that in 1996 the iPod and iMac had no influence on the cell phone, IP phone, or even the MP3 player markets. Unless someone within Infogear Technologies was a rival to Einstein and Hawking, there was no way to speculate that the iMac, iPod, then iPhone would come out of Apple over the course of a decade. To run a business on such speculation would likely be suicide anyway.
Maybe Infogear, and now Cisco (or maybe the Cisco of the future) had/has/will have an iTimeMachine, but Occam's Razor suggests coincidence over John Connor coming back to stop Skynet and discussing Apple's product lines with Bob Marshall and/or Sandy Lerner.
So it functions like an electricity-generating windmill with a dehumidifier attached. It either does it without the step in between, or by building both steps built into one device.
The fact that it is a unconventional windmill is a separate feature, although what I like to call "wingstacks" or "windwheels" have been proposed (and perhaps even built before although I can't recall seeing one for real), and vertical-axis windmills are gaining popularity.
A "wingstack" is basically a bunch of lightweight wings -- think airfoils or big kites -- in a loop on a cable. They work like a waterwheel. In a water wheel, the water provides enough power to move the wheel (and sometimes part of the water) around the axis enough that the next spoke gets into the path of the water. In a wingstack, the wings on one side adjust angle of attack enough to get lift and go up, while the ones on the other side produce downforce and go down. Since they're connected and there's a pivot point at the bottom, they move around a circuit and could generate electricity. They're actually, as I recall, instead intended to keep something like an antenna aloft at a second pivot on the upper end. It was Wired, Popular Mechanics, Scientific American, or somewhere like those that I saw these.
So, is it patentable? Is obviousness contrary to patents in Australia?
At some point we can't make the cores any faster. This has to do with roadblocks both in the current state of technology (meaning it can't be done now) and in the limits of physics (meaning that at some point it would take a major breakthrough in raw scientific understanding to make any progress at all).
If you need ten very different things in a particular order and the way each one of them is constructed depends on the result of the one before it, then it's possible that you'd be best served by one processor that's ten times as fast, but that's getting progressively harder and more expensive to do.
If you need one hundred very similar small things right now, you'd be best served by something which can deliver high volumes of the stuff. This is the kind of thing the Cell is very good at.
If you need ten different sorts of large things that may have things about them them overlap in time because the state of one doesn't necessarily depend on the state of the others, you'd be best served by ten processors each working on a partial solution the whole time, and the final piece being tied together at the end. This is the sort of thing multi-core general processors are getting better at.
One issue many people in this thread are addressing is the difference between the second and third options. Another is how to best achieve the third one, since it is not always easy to figure out which parts of which tasks are going to be independent enough to be calculated separately nor easy to tell the computer the difference once the programmer knows.
Multi-core is the only sure way to get much additional performance for the near future. Breakthroughs are wonderful and maybe even worth a wager. They're definitely not the norm, though, and therefore cannot be counted upon as a strategy. Optical computing or quantum computing may end up being widespread in a few years, and taking stock in someone working on them may pay off big. People who work in the field, though, have to hedge that bet.
Once upon a time, there were mille coins in the U.S., and my parents remember them. The individual states often minted them, which would likely not be allowed under our increasingly powerful central government of today.
We do not have anything smaller than a penny actually minted any more, specifically because each of the smaller coins experienced this same situation of costing more than its own value. Many things are still priced in half-penny or tenth of a penny denominations, especially things sold in bulk. The final price is just rounded to the nearest penny. (Or sometimes bumped up to the next penny in favor of the vendor for any fraction).
If the penny goes away, the same thing will probably still be done, only we'll be rounding to nickels or dimes.
X-rays are much safer for your laptop than the giant magnet they use to detect metal on you.
Especially on old Intel Pentiums with division bugs.
The fat teenagers surely get some of their heft from their fat parents. The eating and exercise habits of a parent are often passed on to children. These teenagers would probably be thin if they had thin parents in the suburbs and still be fat if they had fat parents in the city.
Sure, the by-car-only lifestyle doesn't help. Lots of developments even lack sidewalks, and zoning often puts any type of grocery, arcade, bar, or any other type of business miles away from where people live. It's more about how homes are situated near businesses and schools or far away from them than whether you're in the city or the suburbs. Unfortunately, many people specifically want their homes as far away from businesses as possible. The traffic, the buildings being empty at night, the looks of commercial property, and more are given as reasons.
These trends almost certainly combine to allow children to be fat. It's not one or the other.
The people employing the programmers think the programmer should use the least resources necessary for the task assigned. Since people buy software as long as their current system doesn't make it painfully slow to use, the programmer's time is more judiciously spent fixing bugs or writing another bloated product than optimizing something that runs fast enough.
In the days of older, slower systems with less memory, programmers were more frugal with computer resources because there was a need. Then, much of the world was doing single-user, single-tasking things on ridiculously fast systems. Now, in these days of multi-tasking (most people still only use desktops one user at a time) it'd be nice to see some apps that are a bit more miserly since even a few megs per application of waste can add up.
It's still a matter of programmer time vs. benefit. A company could go broke trying to wring more application features out of less memory and fewer processor cycles. If their apps can be a little more efficient than the next company without their programmers having to take much more time to do it, that's likely an acceptable trade in the marketplace. Still, it' the feature checklist that people shop for.
Speed, memory requirements, DLL hell, and even stability are likely going to be blamed on the OS vendor more than the application vendor anyway. Sad but true.The OS should be fast, responsive, and stable on its own. It should keep an errant application from bringing down the system. In fact, killing the application that's causing problems can be the key to the OS keeping the rtest of the system stable. There's no reason to expect MS, Novell, Apple, Red Hat, or Sun to make third-party applications fast and reliable. Many people still do.
Apparently, too many potential contributors.
/. /.
The steps are like this:
1. build web hosting presence that will survive
2. announce yourself on
3. ???
4. profit!
It seems someone put #2 ahead of #1.
The study of crystalline structures in general is very important. LCD? Liquid Crystal Display. Timings in your PC? Crystals. E-paper? Likely the solutions will involve crystals. Self-assembly of electronics in the future? Crystals and the way they form are a big part of that, since crystals are one of the ways that things self-assemble. There are lots of examples besides these, I'm sure.
The study of snowflakes specifically has uses for weather forecasting. Ever wonder why the guy on TV says 3-5 inches of snow and it turns out to be 7? It' not necessarily because there's more water in the snow than anticipated. The shape and size of the flakes makes a sizable difference in the volume of the snow. People's lives and safety can depend on good weather forecasting these days. The budgets of road crews always does, since it costs money to field trucks overnight in anticipation of snow whether it actually falls or not.
Snow is also an excellent indicator of pollution patterns since it straps lots of particulate matter in the flakes. What's in the air actually effects the shape and size of snowflakes, too.
Besides weather forecasting and pollution tracking, snowflakes, being so plentiful, can serve as an excellent example of crystal formation in general, since it is in large part a generally applicable principle.
No, not hardware abstraction. That's what DX10 already does.
Virtualization allows a whole different OS to run inside the current one as an application. Sometimes the host OS is a stripped-down one especially designed for hosting other OSes. Other times it's a general-purpose OS like Windows.
Emulation at this level doesn't necessarily mean fully emulating the hardware. The hardware for Windows and for other x86 OSes is the same. You can emulate the IO devices for the drivers on the guest OS, while letting it use the CPU natively. You end up, if the emulation is done thinly enough, with a second driver-to-hardware call for each output operation as your main overhead. Since we're talking about sending longer and longer programs to the GPU to let it process those itself, you'll actually see the overhead go down as shader programs get larger.
Using something like Xen, Bochs, VirtualPC, Parallels, or VMware, you just maintain compatibility with your virtualization or emulation environment. That system's developers and maintainers worry about compatibility with the host OS.
It's not outrageous to suggest that big-box retail game studios could support an OS under a virtualized or close-to-hardware emulated environment. One framework that supports most PC hardware, OpenGL, and a good number of IO devices well and that uses a cheap virtualization/emulation layer might even become a de facto standard eventually if the idea proves promising in practice. DOSBox does great things for DOS games under XP. I know people who play older Windows games under Wine. It's not too far fetched that in the reasonably near future you could boot a Linux, ucLinux, FreeBSD, NetBSD, VxWorks, QNX, or even OS X instance in a window on XP or Vista and play a game there with pretty good performance.
I don't have mod points today, but that's one of the most "Insightful" leaps of thought I've ever read on /.
Kudos to you.
If I was a shareholder of a publicly traded utility company (I am not, but I own shares of a fund that owns shares of some), I'd be very ready to disagree with you (and I am).
Cooperative utility organizations are supposed to be nonprofit. The publicly traded companies expect to make money, and the shareholders expect them to do so. The city in which I lived before my last move had a municipal electric system. They overbuilt capacity and started selling electric to neighboring communities, too. Even they turned a bit of an operational surplus, which instead of being taken as profit went into increased maintenance, a couple of extra positions, and a shift of some of the money into other city funds so things could be done on utility money rather than taxes.
So no, utilities are not as a rule non-profit.
Sorry to respond to myself, but I did a little research.
Actually, it looks like Darl doesn't like to sell at all. Larry Goldfarb and BayStar Capital Group, however, had been selling thousands of shares every few days in late 2004.
What could be really telling is who sold the most between March/April of '03 and February/March of '04, during which time the stock went from around $3 to around $17 and started to slide back down to around $8.
Darl and Kevin probably made all their money in a pump & dump, when SCO's stock skyrocketed upon suing IBM and getting investments from that capital investment group that was acting as a front for Microsoft. SCOX stock right now is $1.06, but it hasn't always been. Anyone know for sure if and when Darl sold a bunch of shares?