Yes, but will there always be video output ports? You are, of course, speaking of the analog hole. Let's for the moment just assume that there is reasonably priced hardware that can capture an analog HD stream (which doen't happen to exist yet.) The problem we are seeing is that content providers are working on DRM that will ONLY play the movie when there is NO ANALOG port. Kinda like Vista won't play HD content unless you have a HDCP monitor.
The answer is of course that some Chinese manufacturer will sell a HDCP dongle that will strip HDCP. As for the keys, it will always be possible to bribe an employee at a legit manufacturer and get some keys. In fact, I would bet that someone will start a distributed.net style crack effort in any case.
I agree. I have an application that takes a huge EMC box of storage and 50 and change quad processor servers each with many gigs of ram. Eventually, desktops will be powerful enough to run the application locally, but I don't see that happening for another 10 years based on how fast technology has advanced in the last 10 years. My guess is that the application will evolve in the next 10 years so that it will STILL take a boatload of equipment, and will still not be able to run on a desktop.
The network is the network, and the computer is the computer.
Frankly, the "network is the computer" slogan is just that - total marketing bullshit. People are fully aware that applications run on computers. Nobody I know, from kids to senior citizens, thinks that "the network is the computer." If anything, they can't tell the difference between a remote application and a local application, and get confused thinking that they "downloaded google" to their local machine. The applications look local, and in fact, the UI IS local. So I TOTALLY disagree that to the user it seems as though "the network is the computer." From my observations, it's the opposite. The network is invisible to the end user.
They're not really needed anywhere. Just a small screen, input device, sensor, handheld device, or network terminal is really necessary in most locations.
Except that a terminal is really just a special purpose computer. May as well be a general purpose computer while you are at it. And flash memory is so cheap now that you may as well store stuff on it - very handy for when you don't HAVE net access (airplane, in the car outside of cell range).
Even if you DO have net access, you don't always want to download gigs of data constantly. You want it cached locally.
Yeah, my laptop has ~145ppi (DumdumDell D810 1920x1200). That said, the screen is 8.25" tall, where paper is 11". I still can't comfortably read a PDF without scrolling because of this due to the size of the text. My eyes just are not that good. My desktop monitor is a much lower 85ppi, and that's just not enough dots to read a PDF full height either.
We had a new second telephone company come to town to compete with Verizon / Adelphia. They started pulling their own fiber (this was a year before Verizon announced FIOS. Of course Verizon still hasn't announced FIOS for our town.)
The freakin PINHEADS only offer 1Mb or 2Mb internet - via FIBER. Heck, DSL in this area is as fast or faster.
DSL technology already exists that can offer higher speeds over longer distances than Verizon (and most other ILECs) currently support, but verizon (and other ILECs) just won't deploy it. Instead, they continue to install obsolete technology.
s/consumer internet connection/internet connection with a dynamic IP address/
There. That should fix your comment. Note that IMHO, this should happen ANYWAY. Dynamic addresses sending email direct to port 25 should be blocked. The problem is that there is no accountability with dynamic addresses. Way back when, when the internet was a kinder, more gentle place, this wasn't a problem. It is now, and action needs to be taken even though it causes some hardship for some people. This is why we have the MSA port (587). No, you don't HAVE to use your ISP's mail server, but you may have to rent space with another provider that allows you to send email via their servers. In fact, I setup a number of email servers on dynamic addresses that are configured this way. Works just fine. Personally, I spent the extra few bucks and got a static IP account for home that has an AUP / TOS that allows me to run servers.
Dude - don't know if you are just clueless or trolling, but my blacklists are based on IP addresses (like most blacklists) and also regex patterns on reverse DNS. These are not the only anti-spam measures I have by a longshot, but the combined local and DNSBL based blacklists do reduce (with minimal resource usage) the amount of email that needs additional analysis by 90%.
And yes, there can be collateral damage when an ISP reuses a spammers old IP address. This is why blacklists like spamhaus are better for everyone rather than individual lists. Of course I and most admins have a way to let you out of our blacklists, but you need to request the removal one blacklist at a time, and there may be hundreds of thousands.
I see it a little differently. Some modern ads have become so obnoxios that you just can't sit through them. Some are just annoyingly loud with horrible, harsh music or speech, or quickly flash 47 mini video clips in 15 seconds giving you an instant headache... I suppose the ad agency thought they were doing an attention grabber, but instead they have evoked an instant hatred of whatever product was being advertized. On the web, is there anyone that doesn't have a total hatred of X10?
Use tactful, creative, intelligent ads that are non-intrusive and I won't be blocking / skipping them.
I was watching a football game this sunday and saw an ad for a car that was done up like a drug ad - guy who is normally clostraphobic in small cars finds the new car roomy enough - at the end he goes into a field of wildflowers with a puppy and the model name of the car is shown with the MPG shown in small print much like drugs show the dosage. It was whitty, creative, and not obnoxious at all. I actually backed it up with Tivo and showed it to my wife who also got a kick out of it.
If it wasn't for spamhaus and other blocklist services, it would be up to individual administrators to create their own blacklists (most savvy admins do anyway BTW...) Now I don't know about other admins, but once you are in MY blacklist, you are there FOREVER. If you are in 4,556,865 blacklists, good f-ing luck getting out. Being on ONE list you have a chance.
The other option is a reputation based system where "trusted" submitters send blacklist updates via usenet (GPG signed.) Since there is no single DNS server (or domain) it can't get shut down. I suppose you could also share the main list via bittorrent, freenet, or other such service.
Anyway, I would expect spamhaus to just get a bunch of alternate names registered all over the place in different countries to get out of a TLD that is under US control.
I'm not worried about NK using nukes in missles against us. I AM worried that they will smuggle them into the US and use them here. It's not like it's hard to smuggle stuff into the country - just ask any drug lord...
In enterprise-class servers, they are hot-swap raid. You pull out the old drive and plug in the new drive. Done. The raid controller rebuilds the drive image from parity or mirror.
There are two types of motherboard swaps - exact and non-exact. The problem with the activation garbage is that even EXACT boards can now fail due to different MAC addresses, or other unique bits. With the old copy-protection free versions of windows, exact replacements (which is a common thing to do in enterprise situations) were NEVER a problem.
Keep in mind that in the US, computers are depreciated in 3 years. Anything older than 3 years starts costing you more from a financial viewpoint. The computer ceases to have an asset value, yet you are still spending money on it. This, more than actual machine failure or being too slow, is the main reason to replace your systems.
And just how realistic IS that suggestion in this modern day and age? How realistic will that be 3 years from now? Firewalls, IDS, AV, and other security systems only get you so far, and avoiding any connection at all to the internet is quickly becomming a non-option.
Um, they don't. It's over a multi-year period. Read my previous post closely and carefully.
Of course if you need to reimage your machines every 6 months, then you're obviously doing some other things wrong too.
No. We don't image every PC, we image every unique type. 6 months is about the longest an image is fresh due to various application patches, upgrades, new application rollouts, etc. If you don't keep your images fresh, then you end up doing too much work to get a newly imaged machine current. I have a need to get a machine from blank hard drive to fully running and in an employees hands in 3 hours without spending 3 hours worth of labor doing it. If an image is reasonbly fresh, it takes minimal labor, and minimal machine reboots.
There is also the HUGE issue of hardware diversity, which is generally a very bad thing in larger corporate environments. You try to keep the number of different types of machines very low to keep the costs of imaging and deployment down. It's one thing to vary the CPU speed, memory, and disk, and another to support lots of different models and vendors of equipment. Equipment and software licenses are only a part of the total cost. Depending on diverity, user knowledge, applications, etc., the support and training costs can sometimes be several times more than the other more tangible costs.
An example: if we have 6 configurations supported in a given year (desktop and laptops for general, power user, engineering) and keep equipment for 4 years, you already have 24 images to support! Now have some on 2000 and some on XP: 48 images. Add vista: 72 images. This doesn't even consider the special "one off" boxes. Given that images generally need to be refreshed every 3 to 6 months, it's a huge job. This is why you do all or nothing deployments whenever possible.
Um, in many many cases the company that is providing the "service" is ALSO fully in control of the development process, and usually has the primary developers on staff. QT, MySQL, and Asterisk are prime examples of this. In fact, I would go as far as saying that MOST commercially supported OSS software works this way.
That said, I do agree with most of the pricing concerns in the FA. RedHat is insane. I looked into a significant number of "server" licenses for one of my clients, and Windows 2003 server would have been 1/3rd the cost over 3 years. The problem is that RedHat somehow thinks that the support costs of N servers is cost-of-one*N where we all know that this isn't true (quantity discounts in the 100 unit range are minimal.) This must be why most hosting companies use FC or CentOS. I can see paying that support cost for a company that has one or two servers, but the math just doesn't work with 100. They really need a license model option where you pay for X hours of support or Y number of incidents (or at LEAST a serious price break for large quantities.)
Sure, there are RFCs and other standards to ensure that if an e-mail isn't delivered, someone is notified, but those are hardly written in stone.
No, the RFC's aren't written in stone, but they DO exist, and they exist for a reason. No email software HAS to adhear to the written standards, but when they don't, your chance of having problems grows dramatically.
Many of the anti-spam systems out there deliberatly violate the standards in order to be more effective. Sometimes the violation is minor and everything still works well. Other times the anti-spam solutions are so poorly designed that they have constant problems with a high rate of legitimate mail lost (verizon is a great example of this.) This is why I pay more for a static IP and run my own home email server for family and friends (which I've done for about 10 years now.) If I have a problem, I: 1) know about it, 2) can fix it NOW.
First, the premise of your question is wrong. Perhaps you need to rephrase it like: "Why are labor costs higher in industrialized nations?" I would suggest if you REALLY want to get an answer that you take some time to read up on (china and US) history and economics. It's not a topic that can be effectivly answered in a short post as you need the background to understand it. I'll give you a few clues to start you down your path: Standard of living, Cost of living, economic model. Look at the level different segments of the economy make up the GDP, and then look at the employment in those segments.
You start out saying that the reasons that outsourcing is big is because the US doesn't compete well, innovate, and work hard. I refute saying that it's really labor costs. You claim that I'm wrong that it's really about labor costs, while really agreeing with my statments and throwing in other bizzare "reasons" that I refute. Then you finally admit that it's all labor costs but with the attitude that I'm still wrong.
Whatever.
The facts are that Labor is really one of the ONLY savings, and it's so huge that it offsets all the other expenses that are LARGER such as the cost of lower quality (which is huge,) the costs of shipping, costs of delays due to shipping, managing manufacturing on the other side of the planet (which frequently requires that someone MOVE there and LOTS of flights,) retraining workers all the time, bribes that need to be paid to the corrupt government, etc. The only way that you can really grasp the issues is if you were involved (as I have been) with projects that were moved overseas. It's not all roses - in fact it is a royal pain in the ass that brings it's own challenges that need to be dealt with continuously.
OK, setting aside the fact that your post is borderline incoherent, you said No then brought up labor related costs as the BIGGEST factor. Legal issues related to labor is a labor cost. Benefit related cost such as healthcare and unemployment is a labor cost.
If you somehow think that using a manufacturer in China means no contracts or legal issues, then you are out of your tree. In fact companies that are not totally clueless will stipulate in the contract that their home state or whatever has jurisdiction over all disputes and not some random province in China.
That said, you have ignored EVERY SINGLE company press release and quotes from executives who univerally claim that labor costs IS their reason for moving operations overseas. NONE cite regulations or legal costs. But whatever.
GM is losing billions of dollars due to sheer stupidity.
The biggie is that they REFUSE to develop reusable modular parts. Instead, every freakin model of car has unique parts that change almost every year for no damn reason. This results in significatly higher engineering costs because you are designing the same damn part 30 times instead of one, significantly higher production costs as you need 30 times more tooling, lower quality since your costs are higher and you are only producing 100,000 parts instead of 3,000,000 (lose the economies of scale but still have to compete in price), etc.
GM also refuses to really get serious about fuel economy. It's not that they can't produce economical engines, they just don't WANT to because they have a bizzare idea of what consumers desire.
GM, if it WANTED to, could do a 180 in 5 years and start blowing the competition away if they made those two simple changes.
Let's be VERY clear here: the ONLY reason globalization and outsourcing is expanding is because LABOR costs in India and China are significantly lower. If we only paid our people $1,200 / year like they do in China, then we would blow the chineese away. Product quality of outsourced manufacturing is also typically lower mainly due to the huge turnover rate in chineese companies - they have increasing competition for qualified workers and pay bottom dollar for them.
The FACT is that US productivity continues to rise due to the FACT that people are working harder, longer, and developing new business systems and technologies. Check the Bureau of Labor Statistics web site for the data.
As an FYI, there is a big difference between Debian, which has an open release process from the very start, and Windows, which is closed to most until right before the release. The release processes are also very different in many other ways that are too numerous to get into here which you obviously don't understand given your comment, but one of those is that Debian is not on a profit and marketing driven release schedule.
Lastly, IMHO your PIII comment is nothing but pure flamebait, but I would also mention that I won't be the #1 first day adopter of hardware either. I like to see a few months of user reports before I buy to make sure there aren't any critical major flaws. Price also tends to come down after initial release. A good example is core2 motherboards. Many in the first generation have/had serious issues.
But let's stop the BS and get back to the meat of the issue. Vista, while it DOES have new features, has NO compelling features that businesses will say "I can't live without that." Given the history of Windows releases, it's insane to be day-1 adopters. It's one thing to buy licenses at a discount on day one, another to actually deploy (although I would be that MS ties the discount to actual deployment.) Given the potential catastrophic problems that first releases of windows has historically had, it wouldn't be worth the lost productivity / revenue which could easily be 100 times the cost savings of the discount in ONE problem alone.
Slashdot Mirror
Yes, but will there always be video output ports? You are, of course, speaking of the analog hole. Let's for the moment just assume that there is reasonably priced hardware that can capture an analog HD stream (which doen't happen to exist yet.) The problem we are seeing is that content providers are working on DRM that will ONLY play the movie when there is NO ANALOG port. Kinda like Vista won't play HD content unless you have a HDCP monitor.
The answer is of course that some Chinese manufacturer will sell a HDCP dongle that will strip HDCP. As for the keys, it will always be possible to bribe an employee at a legit manufacturer and get some keys. In fact, I would bet that someone will start a distributed.net style crack effort in any case.
I agree. I have an application that takes a huge EMC box of storage and 50 and change quad processor servers each with many gigs of ram. Eventually, desktops will be powerful enough to run the application locally, but I don't see that happening for another 10 years based on how fast technology has advanced in the last 10 years. My guess is that the application will evolve in the next 10 years so that it will STILL take a boatload of equipment, and will still not be able to run on a desktop.
The network is the network, and the computer is the computer.
Frankly, the "network is the computer" slogan is just that - total marketing bullshit. People are fully aware that applications run on computers. Nobody I know, from kids to senior citizens, thinks that "the network is the computer." If anything, they can't tell the difference between a remote application and a local application, and get confused thinking that they "downloaded google" to their local machine. The applications look local, and in fact, the UI IS local. So I TOTALLY disagree that to the user it seems as though "the network is the computer." From my observations, it's the opposite. The network is invisible to the end user.
They're not really needed anywhere. Just a small screen, input device, sensor, handheld device, or network terminal is really necessary in most locations.
Except that a terminal is really just a special purpose computer. May as well be a general purpose computer while you are at it. And flash memory is so cheap now that you may as well store stuff on it - very handy for when you don't HAVE net access (airplane, in the car outside of cell range).
Even if you DO have net access, you don't always want to download gigs of data constantly. You want it cached locally.
Yeah, my laptop has ~145ppi (DumdumDell D810 1920x1200). That said, the screen is 8.25" tall, where paper is 11". I still can't comfortably read a PDF without scrolling because of this due to the size of the text. My eyes just are not that good. My desktop monitor is a much lower 85ppi, and that's just not enough dots to read a PDF full height either.
We had a new second telephone company come to town to compete with Verizon / Adelphia. They started pulling their own fiber (this was a year before Verizon announced FIOS. Of course Verizon still hasn't announced FIOS for our town.)
The freakin PINHEADS only offer 1Mb or 2Mb internet - via FIBER. Heck, DSL in this area is as fast or faster.
DSL technology already exists that can offer higher speeds over longer distances than Verizon (and most other ILECs) currently support, but verizon (and other ILECs) just won't deploy it. Instead, they continue to install obsolete technology.
s/consumer internet connection/internet connection with a dynamic IP address/
There. That should fix your comment. Note that IMHO, this should happen ANYWAY. Dynamic addresses sending email direct to port 25 should be blocked. The problem is that there is no accountability with dynamic addresses. Way back when, when the internet was a kinder, more gentle place, this wasn't a problem. It is now, and action needs to be taken even though it causes some hardship for some people. This is why we have the MSA port (587). No, you don't HAVE to use your ISP's mail server, but you may have to rent space with another provider that allows you to send email via their servers. In fact, I setup a number of email servers on dynamic addresses that are configured this way. Works just fine. Personally, I spent the extra few bucks and got a static IP account for home that has an AUP / TOS that allows me to run servers.
Dude - don't know if you are just clueless or trolling, but my blacklists are based on IP addresses (like most blacklists) and also regex patterns on reverse DNS. These are not the only anti-spam measures I have by a longshot, but the combined local and DNSBL based blacklists do reduce (with minimal resource usage) the amount of email that needs additional analysis by 90%.
And yes, there can be collateral damage when an ISP reuses a spammers old IP address. This is why blacklists like spamhaus are better for everyone rather than individual lists. Of course I and most admins have a way to let you out of our blacklists, but you need to request the removal one blacklist at a time, and there may be hundreds of thousands.
I see it a little differently. Some modern ads have become so obnoxios that you just can't sit through them. Some are just annoyingly loud with horrible, harsh music or speech, or quickly flash 47 mini video clips in 15 seconds giving you an instant headache... I suppose the ad agency thought they were doing an attention grabber, but instead they have evoked an instant hatred of whatever product was being advertized. On the web, is there anyone that doesn't have a total hatred of X10?
Use tactful, creative, intelligent ads that are non-intrusive and I won't be blocking / skipping them.
I was watching a football game this sunday and saw an ad for a car that was done up like a drug ad - guy who is normally clostraphobic in small cars finds the new car roomy enough - at the end he goes into a field of wildflowers with a puppy and the model name of the car is shown with the MPG shown in small print much like drugs show the dosage. It was whitty, creative, and not obnoxious at all. I actually backed it up with Tivo and showed it to my wife who also got a kick out of it.
Here's the deal though.
If it wasn't for spamhaus and other blocklist services, it would be up to individual administrators to create their own blacklists (most savvy admins do anyway BTW...) Now I don't know about other admins, but once you are in MY blacklist, you are there FOREVER. If you are in 4,556,865 blacklists, good f-ing luck getting out. Being on ONE list you have a chance.
The other option is a reputation based system where "trusted" submitters send blacklist updates via usenet (GPG signed.) Since there is no single DNS server (or domain) it can't get shut down. I suppose you could also share the main list via bittorrent, freenet, or other such service.
Anyway, I would expect spamhaus to just get a bunch of alternate names registered all over the place in different countries to get out of a TLD that is under US control.
I'm not worried about NK using nukes in missles against us. I AM worried that they will smuggle them into the US and use them here. It's not like it's hard to smuggle stuff into the country - just ask any drug lord...
In enterprise-class servers, they are hot-swap raid. You pull out the old drive and plug in the new drive. Done. The raid controller rebuilds the drive image from parity or mirror.
There are two types of motherboard swaps - exact and non-exact. The problem with the activation garbage is that even EXACT boards can now fail due to different MAC addresses, or other unique bits. With the old copy-protection free versions of windows, exact replacements (which is a common thing to do in enterprise situations) were NEVER a problem.
Keep in mind that in the US, computers are depreciated in 3 years. Anything older than 3 years starts costing you more from a financial viewpoint. The computer ceases to have an asset value, yet you are still spending money on it. This, more than actual machine failure or being too slow, is the main reason to replace your systems.
And just how realistic IS that suggestion in this modern day and age? How realistic will that be 3 years from now? Firewalls, IDS, AV, and other security systems only get you so far, and avoiding any connection at all to the internet is quickly becomming a non-option.
Why are you buying models that change that often?
Um, they don't. It's over a multi-year period. Read my previous post closely and carefully.
Of course if you need to reimage your machines every 6 months, then you're obviously doing some other things wrong too.
No. We don't image every PC, we image every unique type. 6 months is about the longest an image is fresh due to various application patches, upgrades, new application rollouts, etc. If you don't keep your images fresh, then you end up doing too much work to get a newly imaged machine current. I have a need to get a machine from blank hard drive to fully running and in an employees hands in 3 hours without spending 3 hours worth of labor doing it. If an image is reasonbly fresh, it takes minimal labor, and minimal machine reboots.
There is also the HUGE issue of hardware diversity, which is generally a very bad thing in larger corporate environments. You try to keep the number of different types of machines very low to keep the costs of imaging and deployment down. It's one thing to vary the CPU speed, memory, and disk, and another to support lots of different models and vendors of equipment. Equipment and software licenses are only a part of the total cost. Depending on diverity, user knowledge, applications, etc., the support and training costs can sometimes be several times more than the other more tangible costs.
An example: if we have 6 configurations supported in a given year (desktop and laptops for general, power user, engineering) and keep equipment for 4 years, you already have 24 images to support! Now have some on 2000 and some on XP: 48 images. Add vista: 72 images. This doesn't even consider the special "one off" boxes. Given that images generally need to be refreshed every 3 to 6 months, it's a huge job. This is why you do all or nothing deployments whenever possible.
If anything, the numbers in the FA are low.
Um, in many many cases the company that is providing the "service" is ALSO fully in control of the development process, and usually has the primary developers on staff. QT, MySQL, and Asterisk are prime examples of this. In fact, I would go as far as saying that MOST commercially supported OSS software works this way.
That said, I do agree with most of the pricing concerns in the FA. RedHat is insane. I looked into a significant number of "server" licenses for one of my clients, and Windows 2003 server would have been 1/3rd the cost over 3 years. The problem is that RedHat somehow thinks that the support costs of N servers is cost-of-one*N where we all know that this isn't true (quantity discounts in the 100 unit range are minimal.) This must be why most hosting companies use FC or CentOS. I can see paying that support cost for a company that has one or two servers, but the math just doesn't work with 100. They really need a license model option where you pay for X hours of support or Y number of incidents (or at LEAST a serious price break for large quantities.)
Sure, there are RFCs and other standards to ensure that if an e-mail isn't delivered, someone is notified, but those are hardly written in stone.
No, the RFC's aren't written in stone, but they DO exist, and they exist for a reason. No email software HAS to adhear to the written standards, but when they don't, your chance of having problems grows dramatically.
Many of the anti-spam systems out there deliberatly violate the standards in order to be more effective. Sometimes the violation is minor and everything still works well. Other times the anti-spam solutions are so poorly designed that they have constant problems with a high rate of legitimate mail lost (verizon is a great example of this.) This is why I pay more for a static IP and run my own home email server for family and friends (which I've done for about 10 years now.) If I have a problem, I: 1) know about it, 2) can fix it NOW.
First, the premise of your question is wrong. Perhaps you need to rephrase it like: "Why are labor costs higher in industrialized nations?" I would suggest if you REALLY want to get an answer that you take some time to read up on (china and US) history and economics. It's not a topic that can be effectivly answered in a short post as you need the background to understand it. I'll give you a few clues to start you down your path: Standard of living, Cost of living, economic model. Look at the level different segments of the economy make up the GDP, and then look at the employment in those segments.
You start out saying that the reasons that outsourcing is big is because the US doesn't compete well, innovate, and work hard. I refute saying that it's really labor costs. You claim that I'm wrong that it's really about labor costs, while really agreeing with my statments and throwing in other bizzare "reasons" that I refute. Then you finally admit that it's all labor costs but with the attitude that I'm still wrong.
Whatever.
The facts are that Labor is really one of the ONLY savings, and it's so huge that it offsets all the other expenses that are LARGER such as the cost of lower quality (which is huge,) the costs of shipping, costs of delays due to shipping, managing manufacturing on the other side of the planet (which frequently requires that someone MOVE there and LOTS of flights,) retraining workers all the time, bribes that need to be paid to the corrupt government, etc. The only way that you can really grasp the issues is if you were involved (as I have been) with projects that were moved overseas. It's not all roses - in fact it is a royal pain in the ass that brings it's own challenges that need to be dealt with continuously.
OK, setting aside the fact that your post is borderline incoherent, you said No then brought up labor related costs as the BIGGEST factor. Legal issues related to labor is a labor cost. Benefit related cost such as healthcare and unemployment is a labor cost.
If you somehow think that using a manufacturer in China means no contracts or legal issues, then you are out of your tree. In fact companies that are not totally clueless will stipulate in the contract that their home state or whatever has jurisdiction over all disputes and not some random province in China.
That said, you have ignored EVERY SINGLE company press release and quotes from executives who univerally claim that labor costs IS their reason for moving operations overseas. NONE cite regulations or legal costs. But whatever.
GM is losing billions of dollars due to sheer stupidity.
The biggie is that they REFUSE to develop reusable modular parts. Instead, every freakin model of car has unique parts that change almost every year for no damn reason. This results in significatly higher engineering costs because you are designing the same damn part 30 times instead of one, significantly higher production costs as you need 30 times more tooling, lower quality since your costs are higher and you are only producing 100,000 parts instead of 3,000,000 (lose the economies of scale but still have to compete in price), etc.
GM also refuses to really get serious about fuel economy. It's not that they can't produce economical engines, they just don't WANT to because they have a bizzare idea of what consumers desire.
GM, if it WANTED to, could do a 180 in 5 years and start blowing the competition away if they made those two simple changes.
Let's be VERY clear here: the ONLY reason globalization and outsourcing is expanding is because LABOR costs in India and China are significantly lower. If we only paid our people $1,200 / year like they do in China, then we would blow the chineese away. Product quality of outsourced manufacturing is also typically lower mainly due to the huge turnover rate in chineese companies - they have increasing competition for qualified workers and pay bottom dollar for them.
The FACT is that US productivity continues to rise due to the FACT that people are working harder, longer, and developing new business systems and technologies. Check the Bureau of Labor Statistics web site for the data.
As an FYI, there is a big difference between Debian, which has an open release process from the very start, and Windows, which is closed to most until right before the release. The release processes are also very different in many other ways that are too numerous to get into here which you obviously don't understand given your comment, but one of those is that Debian is not on a profit and marketing driven release schedule.
Lastly, IMHO your PIII comment is nothing but pure flamebait, but I would also mention that I won't be the #1 first day adopter of hardware either. I like to see a few months of user reports before I buy to make sure there aren't any critical major flaws. Price also tends to come down after initial release. A good example is core2 motherboards. Many in the first generation have/had serious issues.
But let's stop the BS and get back to the meat of the issue. Vista, while it DOES have new features, has NO compelling features that businesses will say "I can't live without that." Given the history of Windows releases, it's insane to be day-1 adopters. It's one thing to buy licenses at a discount on day one, another to actually deploy (although I would be that MS ties the discount to actual deployment.) Given the potential catastrophic problems that first releases of windows has historically had, it wouldn't be worth the lost productivity / revenue which could easily be 100 times the cost savings of the discount in ONE problem alone.