"You're unemployed because you don't have skills that are in demand. Just because you have a skill doesn't mean that someone will give you a job. No-one owes you a job. Why don't you get some training in something that people want?"
This used to be true. But these days it's hard for a software engineer to get a job even if he's got the perfect skillset, simply because his cost of living makes him, by definition, far more expensive than someone offshore. The "problem" is that software is perfectly portable, so unlike, say, milk, there not much reason to prefer a local supplier. This means that while the local dairy only has to compete against the two other local dairies, software engineers have to compete against India, China, etc.
In the long run I try to be philosophical -- demand is driving wages up in India, etc., and wages in the US are dropping, so eventually it'll work itself out. But it's pretty painful in the "short run"...
Actually (I've done some outsourcing) the time zones work out quite nicely. We had the team in india working on the software during their daytime (our night) and then they'd ship us a build to install and test in our daytime, so that we could log bug reports that they'd fix. The result was a 2x accelleration in software development. Required some management tricks to keep working smoothly, but worked like a champ for me.
Historically, before the minimum wage is raised companies make dramatic statements about how the government would force them to fire lots of workers, and then after the minimum wage is raised nothin happens (except that the workers with the crappiest jobs get paid a bit more).
"I think that Bluetooth headsets are a bit silly, because in return for eliminating the cord you have to add a battery and a bunch of electronics to your headset. "
"It's well worth it. Headset cords annoy the hell out of me. I use a Jabra Freespeak headset with my Sony Ericsson T68i phone, and there's no way I'm giving it up. The Treo 600 looks really nice, but I'll wait for a version with Bluetooth. "
I guess I agree that it'd be nice to have bluetooth in the Treo 600, but I'm not going to wait more months for it...
I'm with you -- the instant telephone number portability kicks in I'm switching from my current telco to a GSM carrier so that I can avoid being locked into weird proprietary telephone land again. Any GSM phone works with any GSM account just by sticking in the ID card, so you can swap phones or carriers and, amazingly enough, everything keeds working. The US telco's, whining about how much portability is costing them to implement, just suck. Everyone else on the planet had this the whole time, and they thought we wouldn't figure it out?
I think that Bluetooth headsets are a bit silly, because in return for eliminating the cord you have to add a battery and a bunch of electronics to your headset.
As the owner of a bluetooth laptop, I can tell you that Bluetooth is wonderful because it eliminates all of the random cords that you need to carry around to connect your various devices (e.g. to synch your Palm to your computer, your computer to your phone, send photo's from a camera to your computer, etc.)
The wonderful thing isn't that you can do something you couldn't do before -- it's that there's now a standard way to connect anything to anything else, with no adaptors or weird proprietary cables. Yay!
I bought a couple of Palm SD Bluetooth cards cheap, and I'd be happy to sell 'em for what I paid for 'em (except for the one in my Kyocera Smartphone): $60. If anyone's interested, you've got my email address above.
Personally, I'd like to have seen the performance of BSD's synthetic filesystem. It's a very unusual and cool filesystem that I've had good experiences with, and it'd be nice to see how it stacks up in a more scientific comparison...
Well, rendering text has been "realtime" for 30 years, because the sophistication of what rendering text means has grown in complexity to take advantage of faster CPU's. Sure, on a 1.77 MHz TRS-80 I could edit text (16 lines of 64 characters!), but all text was white on black, monospaced pre-defined characters (not a graphics display, no fonts much less bolding or italics), and word wrapping was a special command that you issued when you were done writing because the computer took several seconds per page to format. On my current Mac (733 MHz G4, fairly slow by today's standards) I'm driving a huge 24-bit deep bitmapped display, where everything displayed is rendered anti-aliased, then mapped onto a display where it's alpha blended. It doesn't allow me to write much faster (OK, saving to a hard drive is a lot faster than saving to cassette tape) but it's doing 400,000 times as much work in order to do realtime spell checking, text wrapping, line layout, proportional spacing, etc.
And to bring it back to the question of whether applications spend any time executing OS calls, _all_ of that work is being done automagically by the operating system. In fact, I can build my own (simple) word processor using Cocoa in minutes, by snapping together prebuilt framework components that understand fonts, layout, etc., so there's almost no application level code at all.
Well, I guess the question of whether the HTML rendering engine is "in the OS" is a philosophical one -- one could argue that only the lowest levels (kernel, filesystem, VM, drivers, etc.) are really the OS. Or you could say that all of the Cocoa and Carbon frameworks and below (including QuickTime, the HTML rendering engine, etc.) is in the OS, which I think is what any Mac developer would say, since that's all stuff provided by Apple that applications can use. Or you could say that "everything in the MacOS X box" is in the OS, which would include applications.
That being said, when people say "MacOS X 10.3 is faster than 10.2" they're talking about the performance of the total system, not attempting to differentiate between the performance of the filesystem, drivers, GUI, frameworks, applications, etc., so the distinction you're trying to draw is not terribly relevant.
So while I guess you might be right in saying that applications spend fairly little time executing time in the OS if you define the OS as only the kernel, I think I'll stick with my definition of OS: the total software system as provided by Apple. And pretty consistently for nearly 20 years now (I've been developing Mac app's a long time -- I used to use the Lisa Pascal cross-compiler) well written Mac app's typically spend over 90% of their time in OS calls. So when Apple tunes their code, it really does have a strong impact on overall application performance.
OK, well, I've been running various 10.3 beta's for a while now, and it _is_ that much faster. Modern applications (i.e. with a GUI, over a decent app framework) spend the vast majority of their time inside system calls (excluding compute-intensive stuff like Photoshop, of course). Think about it -- did it take more CPU to generate a string to display, or to render every character anti-aliased and alpha-blended onto the display? How much time does iMovie spend decoding and encoding video, or reading and writing files? Apple appears to have really tuned the display code...
Aside from that, there are huge chunks of code above the OS that Apple also tuned -- Mail.app, Safari, iTunes, iMovie, etc. So when Safari runs 3x faster than before (subjective number), I can't really tell without a lot of work whether it's the Safari application layer, the HTML engine ("in the OS"), the text layout engine, the display drivers, etc. But I can tell you that after using 10.3, I wouldn't voluntarily run 10.2 again.
Hilights problems with compulsory licensing
on
Why Only Music?
·
· Score: 3, Insightful
Taking the compulsory licensing model and applying it to other media hilights the flaws in the model, for music as well as other media.
One problem is in how money moves. Instead of dealing with the relatively straightforward counting of albums (or games, or DVD's) sold, you'd have to somehow measure files moving around the network. Since there's no way to accurately measure how files move around a p2p network(especially with encrypted networks like FreeNet), the division of money would be the subject of ongoing (legal) fighting. If people weren't fighting over the numbers, they'd be fighting over the formula used to split the money.
Then there's pricing. In a network where everything is the same price (free), how do you differentiate between low-value products cranked out for almost nothing and the huge, high-value productions? There are plenty of niches where the number of customers is small but people are willing to pay more (e.g. massive RPG's, enterprise software).
Another problem is that the total dollars for the music (etc.) business would be fixed, rather than fluctuating up and down based on people's desire for music (etc.). This means that there's less reward for making great music (i.e. the classics that grow the business). Instead, the winning strategy would be to crank out "OK" music as cheaply as possible to collect the guaranteed income.
Imagine a new game company introducing a great new game, Deer Hunter. That game created a new genre of game, selling through an entirely new distribution channel (cheap, simple games selling through grocery stores, etc.), which grew the game industry dramatically. In the compulsory licensing model, before launching the game they'd have to make sure that their new distribution model was counted properly, and that the proper accounting would be done. And even if they were a hit that doubled the size of the game business (which is what actually happened), instead of getting 100% more income into the industry, the new game would suck 50% of the fixed revenue from other companies, so they make 1/2 what they would have, and other companies whose "sales" didn't actually go down would lose 50% of their revenue.
WMP can play MPEG4 encoded files, if they're packaged as AVI's.
I find it annoying that Divx keeps promoting its brand as if it were something different from MPEG4. If they're incompatible with the MPEG4 spec for some reason, then they're just broken...
I don't doubt that the other protocols _could_ be vulnerable to this sort of spoofing. My point, though, was that since they actually did the analysis on a protocol that is used by only a fairly small percentage of file sharers, it probably can't form the basis of a legal defense of people getting subpoenad by the RIAA.
I wouldn't call Kazaa-Lite reverse engineering; all they did is take the Kazaa binaries and strip out the adware and spyware, leaving the core of the application untouched.
It's cool to hear from someone at SGI -- information is good!
It's very cool that SGI can conclusively compare the Linux codebase and the System V codebase as licensed by SCO. If a major company (far larger/more relevant than SCO!) with no agenda (not being sued, right?) all of the relevant information can say "nothing to see here, move along" that pretty much wipes out whatever credibility SCO might have with business people.
Heh. I read the link you posted, and my reaction was "Cool, Disney isn't as spineless as I thought!"
I just can't take seriously any list of complaints that starts out with "FACT: In "The Little Mermaid," a scene depicts a priest becoming noticeably aroused while presiding over a wedding. [Editor's Note: This could be the priest's knees. Please check it out for yourself]. Also, a castle spire on the jacket art for the video resembles a phallus." and then complaining that "FACT: Disney has extended company health benefits to live-in partners of homosexual employees (the policy does not cover unmarried heterosexual couples who live together) - The Orlando Sentinel, 10/7/95; USA Today, 10/19/95; Daily Variety, 10/9/95" is just crazy...
I guess that Beauty is in the eye of the beholder.
"As a result of that exhaustive investigation, SGI has discovered a few additional code segments (similar in nature to the segments referred to above and trivial in amount) that may arguably be related to UNIX code. We are in the process of removing and replacing these segments."
Isn't SCO costing SGI an awful lot of time and effort (i.e. money) to scour code in order to find the "iunfringing code" that supposedly SCO knows about and won't release? I'd think that SGI could have a case to sue SCO for refusing to identify the "infringing" code based on SGI having to spend money attempting to address SCO's claims without SCO helping? I'm pretty sure that the law requires that if you believe someone is infringing on your copyrights that you identify the infringements so that they can stop. Am I missing something, or is SCO just building a case against itself by refusing to allow people to remove the infringing code?
The local game stores that I visit all tell me that the GameCube is "flying out the door" -- Toys-R-Us in midtown NY had huge stacks of GC's, and the whole time I was there you could see people walk over, see the price, and pick up units -- probably sold ten in the hour I was there.
There is the exception of one GameStop where the manager appears to be an X-Box fan, and has so effectively hidden the GameCube displays (literally hidden in the back of the store, behind all of the used software and books, with no sign) that you'd have to search the entire store to find it. Amazingly, at that store, when I asked about the effect of the price drop on GC sales, they sounded annoyed, as if people had been bugging them about it all day, and they'd be happer if the GC went away so that they could give more space to the X-Box. I really have to wonder about that manager -- they're so offensive towards buyers of anything but the X-box there, I'd have to think that they're losing sales. Heck, their PS2 section is hard to find, and the PS2 is outselling the X-Box at what, 2:1?
That's an awful lot of writing in order to point out that if a buyer has a product shipped to a third location, the "sale" takes place at that third location. You're right -- I didn't see any need to go into a relatively obscure case where the buyer doesn't have the product shipped to themselves, but you're right. The sale takes place where the good is delivered, not where the seller or buyer are located. So nobody collects sales tax where the buyer or seller are, because the sale didn't take place there.
Legally, however, states can tax all sales that take place in the state, which is the case we're talking about. The reason that states can't tax goods shipped into the state by companies that don't have a physical presence in the state is that courts ruled in 1992 that the administrative burden placed on companies to comply was too high. If the states (and cities) can find a way to simplify the tax situation to the point where a court would rule that the burden was reasonable, my understanding (IANAL, but I have asked lawyers about just this issue) is that the courts would probably allow states to collect sales tax on products sold to buyers in the state.
Many states currently attempt to collect a "use" tax on sales into the state by companies that aren't currently required to collect sales tax. Practically speaking this is unenforceable, but it establishes the legal precident that they have the right to collect the tax. The only real debate is whether they can stop squabbling long enough to simplify the tax codes for mail order sales sufficiently that they won't get thrown out of court the next time they actually try to collect taxes on sales from companies that don't have offices in their state.
If we want to discuss how states got into budget problems, we'd have to discuss things like unfunded federal mandates that shift costs from the federal budget down to the states, and how many states can't run a deficit in an economic downturn while the federal government can, how it's nearly impossible to kill a bureacracy, and other things that are way off topic.
While it's interesting that apparently Gnutella can in theory be spoofed, I can't believe that this could form much of a legal defense since the spoofs are specific to Gnutella, so this has nothing to do with the vast majority of p2p usage.
"Under the state's proposed "Freeware Initiative," there would be no exceptions to the rule permitting only open source/Linux software."
While I haven't found the actual text of the law, the summaries state that closed source software would be allowed if there were no open source alternative. So it strongly promotes open source software, but it does not mandate it.
"rather the decision to exclude all other competitors in the bidding process"
I have to point out that requiring the government to purchase only open source software does not exclude _any_ vendors from the process. It may cause some vendors to decide that they don't want to satisfy the government's requirements. But if MS were to produce products that were open source, they could bid for a project along with IBM, HP and all of the other companies that can bid on delivering open source systems.
Slashdot Mirror
"You're unemployed because you don't have skills that are in demand. Just because you have a skill doesn't mean that someone will give you a job. No-one owes you a job. Why don't you get some training in something that people want?"
This used to be true. But these days it's hard for a software engineer to get a job even if he's got the perfect skillset, simply because his cost of living makes him, by definition, far more expensive than someone offshore. The "problem" is that software is perfectly portable, so unlike, say, milk, there not much reason to prefer a local supplier. This means that while the local dairy only has to compete against the two other local dairies, software engineers have to compete against India, China, etc.
In the long run I try to be philosophical -- demand is driving wages up in India, etc., and wages in the US are dropping, so eventually it'll work itself out. But it's pretty painful in the "short run"...
Actually (I've done some outsourcing) the time zones work out quite nicely. We had the team in india working on the software during their daytime (our night) and then they'd ship us a build to install and test in our daytime, so that we could log bug reports that they'd fix. The result was a 2x accelleration in software development. Required some management tricks to keep working smoothly, but worked like a champ for me.
Historically, before the minimum wage is raised companies make dramatic statements about how the government would force them to fire lots of workers, and then after the minimum wage is raised nothin happens (except that the workers with the crappiest jobs get paid a bit more).
"I think that Bluetooth headsets are a bit silly, because in return for eliminating the cord you have to add a battery and a bunch of electronics to your headset. "
"It's well worth it. Headset cords annoy the hell out of me. I use a Jabra Freespeak headset with my Sony Ericsson T68i phone, and there's no way I'm giving it up. The Treo 600 looks really nice, but I'll wait for a version with Bluetooth. "
I guess I agree that it'd be nice to have bluetooth in the Treo 600, but I'm not going to wait more months for it...
I'm with you -- the instant telephone number portability kicks in I'm switching from my current telco to a GSM carrier so that I can avoid being locked into weird proprietary telephone land again. Any GSM phone works with any GSM account just by sticking in the ID card, so you can swap phones or carriers and, amazingly enough, everything keeds working. The US telco's, whining about how much portability is costing them to implement, just suck. Everyone else on the planet had this the whole time, and they thought we wouldn't figure it out?
I think that Bluetooth headsets are a bit silly, because in return for eliminating the cord you have to add a battery and a bunch of electronics to your headset.
As the owner of a bluetooth laptop, I can tell you that Bluetooth is wonderful because it eliminates all of the random cords that you need to carry around to connect your various devices (e.g. to synch your Palm to your computer, your computer to your phone, send photo's from a camera to your computer, etc.)
The wonderful thing isn't that you can do something you couldn't do before -- it's that there's now a standard way to connect anything to anything else, with no adaptors or weird proprietary cables. Yay!
"I'd imagine this is the last actual handspring release, before they fully merge with Palm.
:-)
Is this analagous to when Apple bought out Power Computing?"
More like when Apple bought NeXT, as far as I can see.
I bought a couple of Palm SD Bluetooth cards cheap, and I'd be happy to sell 'em for what I paid for 'em (except for the one in my Kyocera Smartphone): $60. If anyone's interested, you've got my email address above.
Personally, I'd like to have seen the performance of BSD's synthetic filesystem. It's a very unusual and cool filesystem that I've had good experiences with, and it'd be nice to see how it stacks up in a more scientific comparison...
Well, rendering text has been "realtime" for 30 years, because the sophistication of what rendering text means has grown in complexity to take advantage of faster CPU's. Sure, on a 1.77 MHz TRS-80 I could edit text (16 lines of 64 characters!), but all text was white on black, monospaced pre-defined characters (not a graphics display, no fonts much less bolding or italics), and word wrapping was a special command that you issued when you were done writing because the computer took several seconds per page to format. On my current Mac (733 MHz G4, fairly slow by today's standards) I'm driving a huge 24-bit deep bitmapped display, where everything displayed is rendered anti-aliased, then mapped onto a display where it's alpha blended. It doesn't allow me to write much faster (OK, saving to a hard drive is a lot faster than saving to cassette tape) but it's doing 400,000 times as much work in order to do realtime spell checking, text wrapping, line layout, proportional spacing, etc.
And to bring it back to the question of whether applications spend any time executing OS calls, _all_ of that work is being done automagically by the operating system. In fact, I can build my own (simple) word processor using Cocoa in minutes, by snapping together prebuilt framework components that understand fonts, layout, etc., so there's almost no application level code at all.
Well, I guess the question of whether the HTML rendering engine is "in the OS" is a philosophical one -- one could argue that only the lowest levels (kernel, filesystem, VM, drivers, etc.) are really the OS. Or you could say that all of the Cocoa and Carbon frameworks and below (including QuickTime, the HTML rendering engine, etc.) is in the OS, which I think is what any Mac developer would say, since that's all stuff provided by Apple that applications can use. Or you could say that "everything in the MacOS X box" is in the OS, which would include applications.
That being said, when people say "MacOS X 10.3 is faster than 10.2" they're talking about the performance of the total system, not attempting to differentiate between the performance of the filesystem, drivers, GUI, frameworks, applications, etc., so the distinction you're trying to draw is not terribly relevant.
So while I guess you might be right in saying that applications spend fairly little time executing time in the OS if you define the OS as only the kernel, I think I'll stick with my definition of OS: the total software system as provided by Apple. And pretty consistently for nearly 20 years now (I've been developing Mac app's a long time -- I used to use the Lisa Pascal cross-compiler) well written Mac app's typically spend over 90% of their time in OS calls. So when Apple tunes their code, it really does have a strong impact on overall application performance.
OK, well, I've been running various 10.3 beta's for a while now, and it _is_ that much faster. Modern applications (i.e. with a GUI, over a decent app framework) spend the vast majority of their time inside system calls (excluding compute-intensive stuff like Photoshop, of course). Think about it -- did it take more CPU to generate a string to display, or to render every character anti-aliased and alpha-blended onto the display? How much time does iMovie spend decoding and encoding video, or reading and writing files? Apple appears to have really tuned the display code...
Aside from that, there are huge chunks of code above the OS that Apple also tuned -- Mail.app, Safari, iTunes, iMovie, etc. So when Safari runs 3x faster than before (subjective number), I can't really tell without a lot of work whether it's the Safari application layer, the HTML engine ("in the OS"), the text layout engine, the display drivers, etc. But I can tell you that after using 10.3, I wouldn't voluntarily run 10.2 again.
Taking the compulsory licensing model and applying it to other media hilights the flaws in the model, for music as well as other media.
One problem is in how money moves. Instead of dealing with the relatively straightforward counting of albums (or games, or DVD's) sold, you'd have to somehow measure files moving around the network. Since there's no way to accurately measure how files move around a p2p network(especially with encrypted networks like FreeNet), the division of money would be the subject of ongoing (legal) fighting. If people weren't fighting over the numbers, they'd be fighting over the formula used to split the money.
Then there's pricing. In a network where everything is the same price (free), how do you differentiate between low-value products cranked out for almost nothing and the huge, high-value productions? There are plenty of niches where the number of customers is small but people are willing to pay more (e.g. massive RPG's, enterprise software).
Another problem is that the total dollars for the music (etc.) business would be fixed, rather than fluctuating up and down based on people's desire for music (etc.). This means that there's less reward for making great music (i.e. the classics that grow the business). Instead, the winning strategy would be to crank out "OK" music as cheaply as possible to collect the guaranteed income.
Imagine a new game company introducing a great new game, Deer Hunter. That game created a new genre of game, selling through an entirely new distribution channel (cheap, simple games selling through grocery stores, etc.), which grew the game industry dramatically. In the compulsory licensing model, before launching the game they'd have to make sure that their new distribution model was counted properly, and that the proper accounting would be done. And even if they were a hit that doubled the size of the game business (which is what actually happened), instead of getting 100% more income into the industry, the new game would suck 50% of the fixed revenue from other companies, so they make 1/2 what they would have, and other companies whose "sales" didn't actually go down would lose 50% of their revenue.
WMP can play MPEG4 encoded files, if they're packaged as AVI's.
I find it annoying that Divx keeps promoting its brand as if it were something different from MPEG4. If they're incompatible with the MPEG4 spec for some reason, then they're just broken...
I don't doubt that the other protocols _could_ be vulnerable to this sort of spoofing. My point, though, was that since they actually did the analysis on a protocol that is used by only a fairly small percentage of file sharers, it probably can't form the basis of a legal defense of people getting subpoenad by the RIAA.
I wouldn't call Kazaa-Lite reverse engineering; all they did is take the Kazaa binaries and strip out the adware and spyware, leaving the core of the application untouched.
It's cool to hear from someone at SGI -- information is good!
It's very cool that SGI can conclusively compare the Linux codebase and the System V codebase as licensed by SCO. If a major company (far larger/more relevant than SCO!) with no agenda (not being sued, right?) all of the relevant information can say "nothing to see here, move along" that pretty much wipes out whatever credibility SCO might have with business people.
Heh. I read the link you posted, and my reaction was "Cool, Disney isn't as spineless as I thought!"
I just can't take seriously any list of complaints that starts out with "FACT: In "The Little Mermaid," a scene depicts a priest becoming noticeably aroused while presiding over a wedding. [Editor's Note: This could be the priest's knees. Please check it out for yourself]. Also, a castle spire on the jacket art for the video resembles a phallus." and then complaining that "FACT: Disney has extended company health benefits to live-in partners of homosexual employees (the policy does not cover unmarried heterosexual couples who live together) - The Orlando Sentinel, 10/7/95; USA Today, 10/19/95; Daily Variety, 10/9/95" is just crazy...
I guess that Beauty is in the eye of the beholder.
Imagine the pr0n that Disney could produce! All those mousketeers grow up... and then there are the animated characters! Dopey Does Dallas!
"As a result of that exhaustive investigation, SGI has discovered a few additional code segments (similar in nature to the segments referred to above and trivial in amount) that may arguably be related to UNIX code. We are in the process of removing and replacing these segments."
Isn't SCO costing SGI an awful lot of time and effort (i.e. money) to scour code in order to find the "iunfringing code" that supposedly SCO knows about and won't release? I'd think that SGI could have a case to sue SCO for refusing to identify the "infringing" code based on SGI having to spend money attempting to address SCO's claims without SCO helping? I'm pretty sure that the law requires that if you believe someone is infringing on your copyrights that you identify the infringements so that they can stop. Am I missing something, or is SCO just building a case against itself by refusing to allow people to remove the infringing code?
The local game stores that I visit all tell me that the GameCube is "flying out the door" -- Toys-R-Us in midtown NY had huge stacks of GC's, and the whole time I was there you could see people walk over, see the price, and pick up units -- probably sold ten in the hour I was there.
There is the exception of one GameStop where the manager appears to be an X-Box fan, and has so effectively hidden the GameCube displays (literally hidden in the back of the store, behind all of the used software and books, with no sign) that you'd have to search the entire store to find it. Amazingly, at that store, when I asked about the effect of the price drop on GC sales, they sounded annoyed, as if people had been bugging them about it all day, and they'd be happer if the GC went away so that they could give more space to the X-Box. I really have to wonder about that manager -- they're so offensive towards buyers of anything but the X-box there, I'd have to think that they're losing sales. Heck, their PS2 section is hard to find, and the PS2 is outselling the X-Box at what, 2:1?
Anywa
The next generation X-box? How about small and cheap to manufacture?
Forgive me, but I have no idea what point you're trying to make. Could you restate?
That's an awful lot of writing in order to point out that if a buyer has a product shipped to a third location, the "sale" takes place at that third location. You're right -- I didn't see any need to go into a relatively obscure case where the buyer doesn't have the product shipped to themselves, but you're right. The sale takes place where the good is delivered, not where the seller or buyer are located. So nobody collects sales tax where the buyer or seller are, because the sale didn't take place there.
Legally, however, states can tax all sales that take place in the state, which is the case we're talking about. The reason that states can't tax goods shipped into the state by companies that don't have a physical presence in the state is that courts ruled in 1992 that the administrative burden placed on companies to comply was too high. If the states (and cities) can find a way to simplify the tax situation to the point where a court would rule that the burden was reasonable, my understanding (IANAL, but I have asked lawyers about just this issue) is that the courts would probably allow states to collect sales tax on products sold to buyers in the state.
Many states currently attempt to collect a "use" tax on sales into the state by companies that aren't currently required to collect sales tax. Practically speaking this is unenforceable, but it establishes the legal precident that they have the right to collect the tax. The only real debate is whether they can stop squabbling long enough to simplify the tax codes for mail order sales sufficiently that they won't get thrown out of court the next time they actually try to collect taxes on sales from companies that don't have offices in their state.
If we want to discuss how states got into budget problems, we'd have to discuss things like unfunded federal mandates that shift costs from the federal budget down to the states, and how many states can't run a deficit in an economic downturn while the federal government can, how it's nearly impossible to kill a bureacracy, and other things that are way off topic.
- LP
While it's interesting that apparently Gnutella can in theory be spoofed, I can't believe that this could form much of a legal defense since the spoofs are specific to Gnutella, so this has nothing to do with the vast majority of p2p usage.
"Under the state's proposed "Freeware Initiative," there would be no exceptions to the rule permitting only open source/Linux software."
While I haven't found the actual text of the law, the summaries state that closed source software would be allowed if there were no open source alternative. So it strongly promotes open source software, but it does not mandate it.
"rather the decision to exclude all other competitors in the bidding process"
I have to point out that requiring the government to purchase only open source software does not exclude _any_ vendors from the process. It may cause some vendors to decide that they don't want to satisfy the government's requirements. But if MS were to produce products that were open source, they could bid for a project along with IBM, HP and all of the other companies that can bid on delivering open source systems.