I was going to jump in with the same point. Google's hired people from DEC and their primary concern in their data warehouses is undoubtedly performance and energy efficiency. It's very likely any innovation in either area that would be dramatically superior to the best offerings from Intel and AMD would not be x86-compatible.
Unless the chip was so outrageously fast that they could run an x86 emulator in it without a performance loss, it would be dead in the home and office PC market.
Or Google wants to release its own line of PCs with its own CPUs and their own Google Operating System. Not very likely.
"There's more than one way to do it" should not be considered a personal challange to find them all
Yes, but when you do things one way, she does things another way, and he does things a third way, you better be damn competent at all three or you can't maintain the code. Of course, that's true in most languages. It's equally easy to intentionally or accidentally obscure things by getting too fancy with the C preprocessor or pointer arithmetic.
Java has a lot of faults, but it replaces pointers with single layer references and removes the preprocessor. In my experience, reading other people's Java code is worlds easier than reading other people's C, C++, and Perl - even if the code itself is poor or buggy. I'm sure there are other languages with a similar straightforward layout and syntax, but I've only worked with the four I listed.
With the source code to the entire operating system at their disposal, the ReactOS people could do a lot to fix the security bugs.
-How about an extra set of control panels or.conf files that let the Administrator lock down certain services. -Intercept certain system calls and ensure that, say, applications can only write to certain folders and only read or write certain pieces of the registry. -Running the web browser as a seperate application instead of an inherent piece of the operating system would help. In fact, there's probably nothing stopping them from making Firefox the default browser. That would plug security holes all by itself. -Additional logging options for basically every service, so that if something starts to go wrong, the user can get more information than the Windows standard Event Viewer provides.
This won't help a home user that runs everything wide open, but for someone genuinely interested in security it would offer a large amount of additional options. Even Linux (and I'm a Linux fan) can be run insecurely by a careless user.
You don't need to learn Linux to use ReactOS. End users wouldn't need to be retrained. More importantly for businesses, Systems Administrators and Network Administrators that were used to Windows wouldn't need to be retrained. Installing Windows software would work, ideally, the exact same it would on an actual Microsoft operating system.
You would get the ability to use all of your Windows-compatible software PLUS no cost for the operating system PLUS the ability to view the source code of the operating system PLUS the added security and stability because anyone could look at the source code and add their own patches and fixes.
I think it's a spectacular idea. That said, Windows is huge. According to this http://www.dwheeler.com/sloc/ Windows XP has 40 million lines of source code. We're talking tens of thousands of volunteer man hours of work.
Obviously the BSD Unixes, Mac, and Windows have good driver ABI models - they all work fine on their respective hardware.
But the advantage of a changeable ABI model is that you can try new things. I'm not aware of any exhaustive benchmarks one way or the other, but it's at least possible that the changeable driver model for Linux gives superior performance or may some day give greater performance than the competition. If the Linux kernel team finds out that a particular change offers a performance advantage, they can make that change. The other operating systems can only do that for changes that dont' break the ABI.
In the reasons for 50 dependency layers, you forgot:
You created or purchased and incorporated a host of different technologies, your customer install base incorporates tens of thousands of applications installed in millions of instances, and as a result your business model requires tremendous backward compatibility.
Imagine trying to modify Debian or Suse (or any other distribution) so that you could install and run any Linux rpm or deb file created since 1999 without having to modify your system. And imagine modifying it so that an upgrade to the next version would not break any of your installed software.
I see Microsoft as a victim of their own success. If they really want to innovate with regards to ease of use, security, performance, or pretty much anything else, they need to start from scratch. Won't happen.
But the adrenaline rush of fighting is fun in its own right. It doesn't matter if it is boxing, kickboxing, Tae Kwan Do, Tae Kwan Leap, or wrestling. The advantage of wrestling is the lower risk of injury. There's still a good chance of getting hurt, but it's lower than the risk of injury from exchanging punches and kicks with your adversary.
For most people, homosexuality doesn't enter into it. Most guys don't box in order to put their hands on other men or play football in order to tackle other men. Wrestling is no different.
Why isn't red-hat more agessively marketing and expanding RCHE certification?
They are aggressively marketing it. It's full page ads in the industry magazines, it's right on the Red Hat website, it's the first link for "Linux Certification" if you do a Google search on the topic.
It's a simple chicken and egg dilemma. Useful certification programs are demanding and expensive. Microsoft software is prevalent, so your average IT fellow is already comfortable with the tools and has a comparatively easy time convincing his employer to fund MCSE training. Red Hat is a Linux juggernaut, but it's still tiny in the industry at large. Most people considering the Linux training will be starting near the ground floor, paying the expensive tuition entirely out of pocket, and using their own vacation time to attend. Hopefully the balance will shift as Linux marketshare grows.
I agree with everything you're saying, though. There needs to be a larger pool of competent Linux system administrators for companies to draw upon.
I find it amusing that a decker-mage worked as a navy sysadmin. When you weren't setting up the black ICE, were you tossing manabolts around?
More seriously, I think your blessed situation was probably too rare. I've only worked at two different places, but I've seen four different sysadmins face this exact problem. They wanted to lock everything down, but the upper management inclination to do whatever the heck they wanted prevented the sysadmins from being able to do their job.
At the last place, the sysadmins took a lot of heat for being unable to protect the executives from their own mistakes. My current place of employment is much smaller, and the CEO and President are fine with periodic server and network problems in return for greater freedom.
Other people have had the same problem. This distribution: http://gobolinux.org/index.php?page=faq
Is an attempt to address your concerns, although it is not particularly newbie friendly. I haven't tried it in a few years, so I can't give a review. But it may be the answer you want.
My 1950 square foot house cost $186,000 in 2002 and was a 45 minute drive (sorry, no public transit available) from my previous job. I pay $3100 per year in property taxes. Mortgage + property taxes + car payments on my sedan + homeowner's insurance + car insurance + gas (even at $3.00 per gallon) = $2500/month.
A similarly sized house along a route with easy access public transit to my old place of work cost over $300,000 with a proportionately higher property tax. Mortgage + property taxes + car payments + homeowner's insurance + car insurance + (less) gas = $3300/month.
A similarly sized house actually in the neighborhood where I used to work cost $400,000 in 2002 with more than $6,000 in annual property taxes. Mortgage + property taxes + car payments + homeowner's insurance = $3600/month, and I would probably have a car anyway for when I needed to travel off the public transit routes.
I know people that live in eastern Pennsylvania and commute 2.5 hours each way every single day into New York City because they can get a nice single family home for $50,000 in the economically depressed coal regions.
Trust me, I don't like wasting 10+ hours a week sitting in a car, bored out of my mind. If I could afford a nice home within a short walk or a public transit commute of my job, I would take it. It's a safe bet that a lot of other Americans - millions, even - find themselves in a similar situation. Until public transit is a lot more common (and who's going to pay for it?) or housing gets a lot cheaper, it won't change.
I've got a 60-odd hour work week and two kids under age 3. I genuinely enjoy puzzling out whatever it is that isn't working in my latest Linux distribution, but I just don't have the time any more.
Right now I'm having a reasonably easy time with Suse 10. If I run into too many snags, maybe my next attempt will be Linspire.
It's a pain in the behind, but I always do this. Start a command prompt as a normal user, and type
runas.exe/user:Administrator cmd.exe
Enter the Admin password when prompted. The command prompt that pops up runs as Administrator, and any process you start from that command prompt also runs as Administrator. I use the Administrator command prompt for starting and stopping services, opening the Administrative Tools, and launching the.exe files for software installers.
It's a little clunkier than su - but it works. On machines that I use frequently, I install cygwin and the first command inside the Administrator command prompt is C:\cygwin\cygwin.bat. Then I've got a handy suite of Unix tools I can run as Admin at my disposal, too.
I never log out and log back in as another user now.
This is probably the same guy that says drive an electric car save the enviroment, when you boil it all down to generation losses, power generation polution, and envirmental damage from toxic chemicals used in the batteries an electric car is about TWICE as polluting as a modern compact.
That's a commonly held misconception.
1. Modern batteries for hybrid cars are recyclable.
2. Power generation from most current power plants, even coal burning ones, are less polluting per watt of power output than an internal combustion engine. The automotive combustion engine trades a lot of efficiency for the ability to be mobile, quick to start and stop, and run at a broad range of RPMs.
3. As more solar/wind/geothermal/tidal/whatever else environmentally friendly power generators are used, the electric car can use the power they generate without modifications.
The real problem with electric cars is the same problem we had with them at the beginning of the 19th century. Battery tech just hasn't improved enough to give them a long range.
in fact, you'll be tempted to think you're not progressing because of it; that's when you start increasing reps (from 5 reps to 12 or more in 2-rep increments per session) and weight (add an increment when you hit 12 or more reps and go back to 5 reps), and improving your strength; you still won't feel sore, but you'll know from the numbers that you put up that you're getting something out of it
Off topic, but I think that people who can continue to add weight and/or repetitions for a long period of time are a genetic fluke. In my experience it is always fast progress for a month or two, slow progress for at most two or three more months, and then indefinite stagnation.
High Intensity Training, Hypertrophy-Specific Training, Power Factor Training, Super Slow, Heavy Duty, routines by Ellington Darden... I've done them all, some for years, and it was still the same. In my (limited, personal) experience, after the first few months you need to do a high volume of moderately demanding but not brutally difficult work to make any further progress. Or take steroids, I suppose.
I think that's why the military gets good results having the recruits do calisthenics several times a day every day for weeks.
Many people [are evil]? I seriously doubt this. Most people are egotistical, which is good because it allows them to maximize their benefit to themselves, which means putting them in a competitive position that in the long run means each of us is doing what we're best at.
But egotistical concerns also lead people to cheat. Some people just plain aren't "best at" a damn thing. They either have a live a pretty poor life, or turn to crime. Not everyone has the potential to be an entrepreneur or skilled professional. More importantly, until robot labor takes over there will always be a high demand for unskilled (and low pay) labor, like picking fruit, janitor work, and waitress (and waitor)'ing. In a capitalist market, the vast majority of the populace will always be stuck doing low paying menial work.
Children worked in factories because it allowed their households to reap income it never would have had.
Just because child labor in factories made their lives better than it was before the Industrial Revolution, it does not mean that it made their lives good. If you take a starving man, give him meals with cocaine in them, and then exploit his addiction for the rest of his life you are no hero.
Nothing is wrong with profit -- it is the incentive that all humans use to judge their time preference.
There's nothing inherently wrong with profit. But like Ayn Rand, you're missing that many people are quite willing to break the rules of fair dealings in order to get ahead. Con artists, thieves, extortionists, liars, and defrauders can and do get away with their crimes in a capitalist market free and clear. Many other people are punished but not nearly in proportion to the extent of their crimes. How many families did Ken Lay help to bankrupt?
No, it wouldn't. The world learned so much from those who sold themselves short in the industrial revolution. Now that we have the information revolution, information is shared about rough working conditions and millions of workers are able to find better jobs in easier work environments. It isn't OSHA or any other regulation that creates a safe workplace, it is open competition of employers looking for good employees that makes the difference -- pure capitalism.
But it isn't open competition. Countries use currency manipulation to make exports cheap. Immigration laws prevent workers from legally traveling to where they can find work. A company in one country uses child labor in dangerous conditions while employers in another country are legally mandated to use adults and safety laws.
You're asserting that workers can find better, safer work without government intervention, but you don't have proof. Many of the major airline carriers in the US are driving down rank and file employee costs and then giving upper management bonuses for negotiating the concessions. Union organization in the Pennsylvania coal regions didn't start winning reasonable rights for workers until the government (finally) stopped turning a blind eye to mine bosses' violent union busting. Jobs are constantly being relocated from where workers make a good living to where workers are desperate to accept any pay rather than dying. This lowers the standard of living of the first group and exploits the plight of the second.
I'd like to believe that freedom from government intervention would ultimately turn a pure capitalist market place into a beneficial environment for almost everyone. But all I can see is rampant abuse, deceit, and for the lower class eventually serfdom and indentured servitude.
You're missing two aspects of the situation entirely.
First and foremost, Microsoft can't bundle Anti-Virus software with its operating system for fear of yet another anti-trust lawsuit.
Second, Microsoft maintains its monopoly through backwards compatibility. There will be very few programs that run on Windows 2000 or XP that fail to run on Vista.
So if Microsoft changes the way libraries are stored in the System32 folder because it's a security risk, they break backwards compatibility.
If Microsoft changes the way the registry works so that malware cannot insert itself or corrupt data, they break backwards compatibility.
If Microsoft changes file permissions, read permissions, or privilege seperation in processes in a major way, again they break backwards compatibility.
If Microsoft defaults to 'off' all of the services and security openings that currently default to 'on', a lot of programs won't install or won't work once installed.
They're a victim of their own success. The only way to make a future version of Windows more secure from the ground up is to redesign many of the central features. If they do that, all the 'killer apps' that people use on Windows will stop working, and Microsoft is dead in the water.
In three years, this thing will cost $200, or maybe less. Then the specs will mean something to Joe average consumer, when he's cross shopping it against whichever AMD chip is in that price range.
And the buyer can rest assured that when he's tackling those intense home computing tasks like tabbed browsing and turbo-tax, system responses will be nearly instantaneous.
I'm more concerned with the ethical bar, not the legal one. If very private medical or financial data about me was going to be carted around in a laptop, I would want the security to be ironclad.
It's exceptionally difficult to do that, and I don't see much of a way around it.
I think you're right, but then where do we draw the line between reasonable and unreasonable? The company I work for started using encryption to cover all the private customer files on our laptops. The laptops are used to collect information in areas that don't have wired or wireless internet coverage.
But unless we disable booting from floppy/CD, it won't prevent someone from popping in a CD, starting up the machine in another operating system, and installing a keystroke logger. Then put the laptop back where you found it, and go about your business. Two days later, they can collect the laptop and keystroke log, and do anything they want with the data.
It's frustrating, but I don't see a way to prevent it.
If you mean it is always good to have a skilled, intelligent manager, then I agree.
If you mean you would like a manager that is skilled in the tools and programming languages that you use, I disagree. My current boss and his superior are both good managers. They prioritize items, they help plan projects and provide reasonable time estimates, and they're willing to spend the required money to get the tools, software, and other items we need to get the job done. My boss has never programmed anything, and his boss hasn't done any software development in 15 years.
My last boss was, hands down, the best software developer out of the 10 developers at the company. He would modify the Linux kernel for different hardware, write a display driver for Windows XP on another device, create network tools, anything. However, getting detailed project specifications out of him was harder than squeezing blood from a stone and he would never, ever sit down and take part in project planning. You would get a one page feature description out of him, and the next time you had any input would be harsh criticism on the failings of the finished version. More importantly, you had to fight like crazy to get funds for anything out of him. The amount of time we spent waiting in line - literally - for a turn to use a particular piece of hardware was absurd. This genius software developer probably wasted $75,000 a year in salary for unproductive employees for each of the four years I was there because he was too cheap to spend $15,000 on two extra servers and a handful of spare hardware items.
As an aside, it really is disappointing to see how commonplace pirating is. I know half a dozen people that use Adobe Photoshop for editing pictures from their digital cameras, and I'm the only one that paid for a copy.
Slashdot Mirror
I was going to jump in with the same point. Google's hired people from DEC and their primary concern in their data warehouses is undoubtedly performance and energy efficiency. It's very likely any innovation in either area that would be dramatically superior to the best offerings from Intel and AMD would not be x86-compatible.
Unless the chip was so outrageously fast that they could run an x86 emulator in it without a performance loss, it would be dead in the home and office PC market.
Or Google wants to release its own line of PCs with its own CPUs and their own Google Operating System. Not very likely.
"There's more than one way to do it" should not be considered a personal challange to find them all
Yes, but when you do things one way, she does things another way, and he does things a third way, you better be damn competent at all three or you can't maintain the code. Of course, that's true in most languages. It's equally easy to intentionally or accidentally obscure things by getting too fancy with the C preprocessor or pointer arithmetic.
Java has a lot of faults, but it replaces pointers with single layer references and removes the preprocessor. In my experience, reading other people's Java code is worlds easier than reading other people's C, C++, and Perl - even if the code itself is poor or buggy. I'm sure there are other languages with a similar straightforward layout and syntax, but I've only worked with the four I listed.
With the source code to the entire operating system at their disposal, the ReactOS people could do a lot to fix the security bugs.
.conf files that let the Administrator lock down certain services.
-How about an extra set of control panels or
-Intercept certain system calls and ensure that, say, applications can only write to certain folders and only read or write certain pieces of the registry.
-Running the web browser as a seperate application instead of an inherent piece of the operating system would help. In fact, there's probably nothing stopping them from making Firefox the default browser. That would plug security holes all by itself.
-Additional logging options for basically every service, so that if something starts to go wrong, the user can get more information than the Windows standard Event Viewer provides.
This won't help a home user that runs everything wide open, but for someone genuinely interested in security it would offer a large amount of additional options. Even Linux (and I'm a Linux fan) can be run insecurely by a careless user.
You don't need to learn Linux to use ReactOS. End users wouldn't need to be retrained. More importantly for businesses, Systems Administrators and Network Administrators that were used to Windows wouldn't need to be retrained. Installing Windows software would work, ideally, the exact same it would on an actual Microsoft operating system.
You would get the ability to use all of your Windows-compatible software PLUS no cost for the operating system PLUS the ability to view the source code of the operating system PLUS the added security and stability because anyone could look at the source code and add their own patches and fixes.
I think it's a spectacular idea. That said, Windows is huge. According to this http://www.dwheeler.com/sloc/ Windows XP has 40 million lines of source code. We're talking tens of thousands of volunteer man hours of work.
I did mean the whole package. Thank you. (and thanks BrokenHalo, who provided the same info.)
In all seriousness, what would you suggest instead as a way to learn about Linux internals?
Obviously the BSD Unixes, Mac, and Windows have good driver ABI models - they all work fine on their respective hardware.
But the advantage of a changeable ABI model is that you can try new things. I'm not aware of any exhaustive benchmarks one way or the other, but it's at least possible that the changeable driver model for Linux gives superior performance or may some day give greater performance than the competition. If the Linux kernel team finds out that a particular change offers a performance advantage, they can make that change. The other operating systems can only do that for changes that dont' break the ABI.
In the reasons for 50 dependency layers, you forgot:
You created or purchased and incorporated a host of different technologies, your customer install base incorporates tens of thousands of applications installed in millions of instances, and as a result your business model requires tremendous backward compatibility.
Imagine trying to modify Debian or Suse (or any other distribution) so that you could install and run any Linux rpm or deb file created since 1999 without having to modify your system. And imagine modifying it so that an upgrade to the next version would not break any of your installed software.
I see Microsoft as a victim of their own success. If they really want to innovate with regards to ease of use, security, performance, or pretty much anything else, they need to start from scratch. Won't happen.
And put it where the n00b can find it.
And include a really good index.
If you don't enjoy wrestling, that's fine.
But the adrenaline rush of fighting is fun in its own right. It doesn't matter if it is boxing, kickboxing, Tae Kwan Do, Tae Kwan Leap, or wrestling. The advantage of wrestling is the lower risk of injury. There's still a good chance of getting hurt, but it's lower than the risk of injury from exchanging punches and kicks with your adversary.
For most people, homosexuality doesn't enter into it. Most guys don't box in order to put their hands on other men or play football in order to tackle other men. Wrestling is no different.
Why isn't red-hat more agessively marketing and expanding RCHE certification?
They are aggressively marketing it. It's full page ads in the industry magazines, it's right on the Red Hat website, it's the first link for "Linux Certification" if you do a Google search on the topic.
It's a simple chicken and egg dilemma. Useful certification programs are demanding and expensive. Microsoft software is prevalent, so your average IT fellow is already comfortable with the tools and has a comparatively easy time convincing his employer to fund MCSE training. Red Hat is a Linux juggernaut, but it's still tiny in the industry at large. Most people considering the Linux training will be starting near the ground floor, paying the expensive tuition entirely out of pocket, and using their own vacation time to attend. Hopefully the balance will shift as Linux marketshare grows.
I agree with everything you're saying, though. There needs to be a larger pool of competent Linux system administrators for companies to draw upon.
I find it amusing that a decker-mage worked as a navy sysadmin. When you weren't setting up the black ICE, were you tossing manabolts around?
More seriously, I think your blessed situation was probably too rare. I've only worked at two different places, but I've seen four different sysadmins face this exact problem. They wanted to lock everything down, but the upper management inclination to do whatever the heck they wanted prevented the sysadmins from being able to do their job.
At the last place, the sysadmins took a lot of heat for being unable to protect the executives from their own mistakes. My current place of employment is much smaller, and the CEO and President are fine with periodic server and network problems in return for greater freedom.
Other people have had the same problem. This distribution:
http://gobolinux.org/index.php?page=faq
Is an attempt to address your concerns, although it is not particularly newbie friendly. I haven't tried it in a few years, so I can't give a review. But it may be the answer you want.
My 1950 square foot house cost $186,000 in 2002 and was a 45 minute drive (sorry, no public transit available) from my previous job. I pay $3100 per year in property taxes. Mortgage + property taxes + car payments on my sedan + homeowner's insurance + car insurance + gas (even at $3.00 per gallon) = $2500/month.
A similarly sized house along a route with easy access public transit to my old place of work cost over $300,000 with a proportionately higher property tax. Mortgage + property taxes + car payments + homeowner's insurance + car insurance + (less) gas = $3300/month.
A similarly sized house actually in the neighborhood where I used to work cost $400,000 in 2002 with more than $6,000 in annual property taxes. Mortgage + property taxes + car payments + homeowner's insurance = $3600/month, and I would probably have a car anyway for when I needed to travel off the public transit routes.
I know people that live in eastern Pennsylvania and commute 2.5 hours each way every single day into New York City because they can get a nice single family home for $50,000 in the economically depressed coal regions.
Trust me, I don't like wasting 10+ hours a week sitting in a car, bored out of my mind. If I could afford a nice home within a short walk or a public transit commute of my job, I would take it. It's a safe bet that a lot of other Americans - millions, even - find themselves in a similar situation. Until public transit is a lot more common (and who's going to pay for it?) or housing gets a lot cheaper, it won't change.
Thanks for the information.
I've got a 60-odd hour work week and two kids under age 3. I genuinely enjoy puzzling out whatever it is that isn't working in my latest Linux distribution, but I just don't have the time any more.
Right now I'm having a reasonably easy time with Suse 10. If I run into too many snags, maybe my next attempt will be Linspire.
It's a pain in the behind, but I always do this. Start a command prompt as a normal user, and type /user:Administrator cmd.exe
.exe files for software installers.
runas.exe
Enter the Admin password when prompted. The command prompt that pops up runs as Administrator, and any process you start from that command prompt also runs as Administrator. I use the Administrator command prompt for starting and stopping services, opening the Administrative Tools, and launching the
It's a little clunkier than su - but it works. On machines that I use frequently, I install cygwin and the first command inside the Administrator command prompt is C:\cygwin\cygwin.bat. Then I've got a handy suite of Unix tools I can run as Admin at my disposal, too.
I never log out and log back in as another user now.
This is probably the same guy that says drive an electric car save the enviroment, when you boil it all down to generation losses, power generation polution, and envirmental damage from toxic chemicals used in the batteries an electric car is about TWICE as polluting as a modern compact.
That's a commonly held misconception.
1. Modern batteries for hybrid cars are recyclable.
2. Power generation from most current power plants, even coal burning ones, are less polluting per watt of power output than an internal combustion engine. The automotive combustion engine trades a lot of efficiency for the ability to be mobile, quick to start and stop, and run at a broad range of RPMs.
3. As more solar/wind/geothermal/tidal/whatever else environmentally friendly power generators are used, the electric car can use the power they generate without modifications.
The real problem with electric cars is the same problem we had with them at the beginning of the 19th century. Battery tech just hasn't improved enough to give them a long range.
in fact, you'll be tempted to think you're not progressing because of it; that's when you start increasing reps (from 5 reps to 12 or more in 2-rep increments per session) and weight (add an increment when you hit 12 or more reps and go back to 5 reps), and improving your strength; you still won't feel sore, but you'll know from the numbers that you put up that you're getting something out of it
Off topic, but I think that people who can continue to add weight and/or repetitions for a long period of time are a genetic fluke. In my experience it is always fast progress for a month or two, slow progress for at most two or three more months, and then indefinite stagnation.
High Intensity Training, Hypertrophy-Specific Training, Power Factor Training, Super Slow, Heavy Duty, routines by Ellington Darden... I've done them all, some for years, and it was still the same. In my (limited, personal) experience, after the first few months you need to do a high volume of moderately demanding but not brutally difficult work to make any further progress. Or take steroids, I suppose.
I think that's why the military gets good results having the recruits do calisthenics several times a day every day for weeks.
Many people [are evil]? I seriously doubt this. Most people are egotistical, which is good because it allows them to maximize their benefit to themselves, which means putting them in a competitive position that in the long run means each of us is doing what we're best at.
But egotistical concerns also lead people to cheat. Some people just plain aren't "best at" a damn thing. They either have a live a pretty poor life, or turn to crime. Not everyone has the potential to be an entrepreneur or skilled professional. More importantly, until robot labor takes over there will always be a high demand for unskilled (and low pay) labor, like picking fruit, janitor work, and waitress (and waitor)'ing. In a capitalist market, the vast majority of the populace will always be stuck doing low paying menial work.
Children worked in factories because it allowed their households to reap income it never would have had.
Just because child labor in factories made their lives better than it was before the Industrial Revolution, it does not mean that it made their lives good. If you take a starving man, give him meals with cocaine in them, and then exploit his addiction for the rest of his life you are no hero.
Nothing is wrong with profit -- it is the incentive that all humans use to judge their time preference.
There's nothing inherently wrong with profit. But like Ayn Rand, you're missing that many people are quite willing to break the rules of fair dealings in order to get ahead. Con artists, thieves, extortionists, liars, and defrauders can and do get away with their crimes in a capitalist market free and clear. Many other people are punished but not nearly in proportion to the extent of their crimes. How many families did Ken Lay help to bankrupt?
No, it wouldn't. The world learned so much from those who sold themselves short in the industrial revolution. Now that we have the information revolution, information is shared about rough working conditions and millions of workers are able to find better jobs in easier work environments. It isn't OSHA or any other regulation that creates a safe workplace, it is open competition of employers looking for good employees that makes the difference -- pure capitalism.
But it isn't open competition. Countries use currency manipulation to make exports cheap. Immigration laws prevent workers from legally traveling to where they can find work. A company in one country uses child labor in dangerous conditions while employers in another country are legally mandated to use adults and safety laws.
You're asserting that workers can find better, safer work without government intervention, but you don't have proof. Many of the major airline carriers in the US are driving down rank and file employee costs and then giving upper management bonuses for negotiating the concessions. Union organization in the Pennsylvania coal regions didn't start winning reasonable rights for workers until the government (finally) stopped turning a blind eye to mine bosses' violent union busting. Jobs are constantly being relocated from where workers make a good living to where workers are desperate to accept any pay rather than dying. This lowers the standard of living of the first group and exploits the plight of the second.
I'd like to believe that freedom from government intervention would ultimately turn a pure capitalist market place into a beneficial environment for almost everyone. But all I can see is rampant abuse, deceit, and for the lower class eventually serfdom and indentured servitude.
You're missing two aspects of the situation entirely.
First and foremost, Microsoft can't bundle Anti-Virus software with its operating system for fear of yet another anti-trust lawsuit.
Second, Microsoft maintains its monopoly through backwards compatibility. There will be very few programs that run on Windows 2000 or XP that fail to run on Vista.
So if Microsoft changes the way libraries are stored in the System32 folder because it's a security risk, they break backwards compatibility.
If Microsoft changes the way the registry works so that malware cannot insert itself or corrupt data, they break backwards compatibility.
If Microsoft changes file permissions, read permissions, or privilege seperation in processes in a major way, again they break backwards compatibility.
If Microsoft defaults to 'off' all of the services and security openings that currently default to 'on', a lot of programs won't install or won't work once installed.
They're a victim of their own success. The only way to make a future version of Windows more secure from the ground up is to redesign many of the central features. If they do that, all the 'killer apps' that people use on Windows will stop working, and Microsoft is dead in the water.
It doesn't mean anything right now.
In three years, this thing will cost $200, or maybe less. Then the specs will mean something to Joe average consumer, when he's cross shopping it against whichever AMD chip is in that price range.
And the buyer can rest assured that when he's tackling those intense home computing tasks like tabbed browsing and turbo-tax, system responses will be nearly instantaneous.
I'm more concerned with the ethical bar, not the legal one. If very private medical or financial data about me was going to be carted around in a laptop, I would want the security to be ironclad.
It's exceptionally difficult to do that, and I don't see much of a way around it.
I think you're right, but then where do we draw the line between reasonable and unreasonable? The company I work for started using encryption to cover all the private customer files on our laptops. The laptops are used to collect information in areas that don't have wired or wireless internet coverage.
But unless we disable booting from floppy/CD, it won't prevent someone from popping in a CD, starting up the machine in another operating system, and installing a keystroke logger. Then put the laptop back where you found it, and go about your business. Two days later, they can collect the laptop and keystroke log, and do anything they want with the data.
It's frustrating, but I don't see a way to prevent it.
If you mean you would like a manager that is skilled in the tools and programming languages that you use, I disagree. My current boss and his superior are both good managers. They prioritize items, they help plan projects and provide reasonable time estimates, and they're willing to spend the required money to get the tools, software, and other items we need to get the job done. My boss has never programmed anything, and his boss hasn't done any software development in 15 years.
My last boss was, hands down, the best software developer out of the 10 developers at the company. He would modify the Linux kernel for different hardware, write a display driver for Windows XP on another device, create network tools, anything. However, getting detailed project specifications out of him was harder than squeezing blood from a stone and he would never, ever sit down and take part in project planning. You would get a one page feature description out of him, and the next time you had any input would be harsh criticism on the failings of the finished version. More importantly, you had to fight like crazy to get funds for anything out of him. The amount of time we spent waiting in line - literally - for a turn to use a particular piece of hardware was absurd. This genius software developer probably wasted $75,000 a year in salary for unproductive employees for each of the four years I was there because he was too cheap to spend $15,000 on two extra servers and a handful of spare hardware items.
As an aside, it really is disappointing to see how commonplace pirating is. I know half a dozen people that use Adobe Photoshop for editing pictures from their digital cameras, and I'm the only one that paid for a copy.