Hmm. On one hand, we have the people responsible for Linux, a nice, fast, loved system.
On the other hand, we have the man who is responsible for Java. You want a language that's a "business phenomenon" because of marketing and has plenty of technical flaws, look no further.
Joy is a bright and creative guy. If there's one thing I've learned from working with bright researchers, though, it's that doing a particular thing well doesn't make you right or even particularly well-informed or unbiased about other things. Joy likes building new things that tickle his fancy. Now he wants to throw out Java and make something new, and is pissed off because he can't because of compatibility contraints. He's already tired of Java. He got tired of Unix a long time ago. He's a nice guy to plop in a research lab and let lose. When he goes and makes blanket statements like the above, though, he speaks with little authority.
Joy doesn't like Linux because he currently in a fan of distributed computing systems. Linux is not as nice a distributed computing system as Joy's Java platform, so Linux is boring to him. The fact that general purpose distributed computing has failed to engender a hell of a lot of interest outside of the research community is no doubt a source of great irritation to many researchers, who find distributed computing a source of many facinating problems.
Joy is overlooking a number of interesting points. Joy's big accomplishments in the past have been design of huge systems -- not modules of systems. Linux is a fantastic system from a research perspective of wanting to try out new OS-level ideas. It's easy to extend and poke at Linux -- much, *much* nicer than poking at and extending Windows. Want to try a new filesystem approach? How about an idea in network scheduling? Linux makes a great research platform.
Linux is a great "product", a practical package. Joy doesn't care about something that's practically useful as long as it isn't technically new.
The point is that people like Perl so much they'll actually buy a t-shirt with the Perl camel on it or whatever to show their affinity for Perl. Nazi flags were flown by the German government.
The Nazi doctrine was not forcibly imposed on everyone (especially the early stuff). The Nazis were quite popular, though their aggressive tactics helped them get more influence faster than they normally would have. Not all of the Nazi doctrine was about stomping Jews because they were Jews -- the Nazi party was a major political party with a lot of different issues. Sure, their most egregious characteristics ended up defining them in people's minds.
A lot of people liked Nixon when he was running for election, remember. He didn't turn out that well either (though decidedly better than Adolf did).
It'd be quite nice to find someone that could answer the first questions, but g++ -Wall on a remotely current gcc will definitely pick out all the above problems. If the position is for a compiler or language analysis position (which I would assume *not*, given the other question), the first questions would be a good idea. For simply a C++ developer...I just don't know.
I'd be interested to see whether the applicant could fix the problems with a compiler handy and then be able to explain *why* the compiler should ask you to fix those problems, though.
The second is certainly fair. Given enough time, it'd be nice to have a third that would test more C++-design features.
"Ask me anything about C++, I will know it. I am an expert".
He struck out 3 for 3.
Many people underestimate the size of C++. I thought I knew C++ well at one point, then decided to pick up Bjarne's book. There's a *lot* of things I had never heard of in C++.
Kind of the same thing in perl, but perl is so phenomenally massive and nonunderstandable that *nobody* makes claims like being able to answer any question about perl.
Your understanding of the economic issues involved is incorrect.
The exchange rate is mostly irrelevant. Playing with currencies does little for purchasing power in anything but the quite short term.
The relevant factor is buying power. If suddenly everyone is paid in US dollars...then folks in India simply would make their current salaries in dollar amounts -- perhaps a thousand dollars a month. Basic goods would still cost perhaps a tenth US costs in India.
I'm dubious. I don't know his goals, but he seems to have done a rather effective job.
Tactically, the jobs his organization has pulled have been stunning successes. 9/11 was amazing in terms of damage/amount of money spent. The attacks on the embassies and the USS Cole didn't require many resources either, and reverberations were felt around the world -- the United States could be hurt. Bin Laden is *still* running around out there, as far as anyone knows, despite an incredible amount of resources being spent on finding him.
Strategically, al Queda has worked well as well. It has polarized international opinion on the United States and stirred up anti-US radicals. Since its attacks, the US has invaded and occupied Iraq, a stunningly poor PR move, and now has the worst international regard since probably well before World War II. It has produced interest in resisting the United States, and has tied down a good deal of United States military power. For the amount being expended in resources by al Queda in Iraq alone, the United States and citizens of the United States pay a phenomenal sum of money in military costs. Al Queda has been largely responsible for the disasterous state of the United States air industry (the end of the economic bubble and SARS PR probably played a role as well). Economically, al Queda has had unqualified successes -- though whether by sheer luck or good reading of US leadership psychology is certainly arguable.
Does terrorism work? Sure does, though perhaps not in the traditional sense of directly toppling a government -- or Ireland and Israel would be quite different.
countless young people, desperate and frightened, cannot.
I have a sneaking suspicion that a good number of people that are promoting Linux are taking not the role of the young, confused, people, but the role of Nike. They are attempting to exploit said confusion in others. Many people see Microsoft as a problem, and Linux as a good, and realize that making people embarassed to admit that they run Windows is likely to sway many techies towards Linux. The methods may be unpleasant, though the ends (may) be good.
For $350 I could buy a pair of English hand made brogues which would be considerably more comfortable. Since the point of paying $130 for a pair of shoes is to advertise the fact that you can afford to pay $130 for a pair of shoes a pair of Church's would work far better.
Well...I agree with your overall point, but I'm quite dubious about this. Rubber and canvas sneakers are awfully comfortable. Also, while that's subjective, I'd like to ask you to look at yourself and check that you are entirely innocent of buying those $350 shoes as a luxury item?
Incidently...while I couldn't care less about what brand my shoes are (even if I was concerned about the style, nobody looks at the things anyway, and if they did, they'd hardly care), Nike does have the nice benefit of making wide sneakers. A lot of sneakers are awfully narrow on my feet, which gets to be quite uncomfortable.
I don't see how you're rebutting any of the points in the parent's post.
Enumerating the horrors of corporate presence in third world countries is meaningless if you then simply ignore the alternatives that people have to working for them.
The jobs we ship to people "willing" to work cheaper do not, for the most part, improve the lives of those they go to. We destroy the land the local populace used to at least manage to survive on, make them wage slaves for a pittance far below liveable, and when Nike, or Union Carbide, or Walmart, or whatever company, finally gets bored and moves elsewhere, they leave slums and wastelands.
You do realize that the image of sweatshops as some sort of horrible torture system that people in other countries are tricked into entering and then prevented from escaping is almost entirely an elaborate fantasy produced by influential union organizations in the United States that are concerned about losing jobs?
This is one of the best posts I have read on Slashdot in a long, long time. It is well-written, insightful, thoughtful, and convincing. It is not offensive. It even has decent grammar. I salute you, sir.
Great, so we're in a situation so overpopulated that *everyone* should live like crap and feel the same - that they are expendable, are no better than any number of others at anything, and whose worth is no more than the cost of replacement...
As opposed to a few living like kings, and the rest living even worse? Is that what should be the case?
"Should" is such a pointless thing to argue about.
If all you do is programming, then I suppose that your job is not hard. But if your job includes analysing your company's business and using information technology to solve business problems that affect the bottom line, then your job is very hard, very valuable, not exportable, and very secure.
In the ideal, this is true. The number of people that I have seen that consider themselves crucial IT employees that "analyze their company's business" and actually do an impressive job is, regrettablly, very, very low.
If you're interested in people who actually care about saving a company money, look no further than small business. However, those companies don't *have* people that do nothing but analyze business processes because they simply don't waste enough on their business processes for this to be an issue.
(A) Long, 1 year minimum turnaround time on anything new that can go into a new model going out. With safety testing and the like, probably longer. If you have to wait two years while a competitor has a feature that you don't, that could be a big deal.
(B) Massive economies of scale. There is a phenomenal amount of money spent on autos. Almost anything constant cost (like designing a new feature) can be done. You only need to sell a couple of units extra a year to make another engineer worth it.
(C) A huge set of bundled technologies. Autos are, arguably, the largest single set of forcibly bundled products that most people will buy in their lifetime. Even a house can be done by a number of different contractors; you cannot combine a Ford body with a BMW dashboard (as a typical consumer). *All* these areas must be kept current -- you're going to be spending a lot of money on improvements no matter what.
(D) Autos are highly visible, and a status symbol. People will buy things for the latest-and-greatest features to show off to their friends. They are a luxury item, not a commodity.
These four points are not present in the overwhelming number of industries in the United States. Blowing lots of money on R&D on paperclips, for instance, is just silly.
Re:Intelligent growth and application.
on
Does IT Matter?
·
· Score: 1
I cannot begin to tell you how many hundreds of thousands of dollars were thrown around for silly and stupid reasons, mostly so Pointy Haired Bosses could play "buzzword bingo" in order to sound important and get promotions.
Absolutely. Big companies are built on middle-and-above managers starting "initiatives" which can then succeed with a dollar amount saved on a graph somewhere so that said manager can get a promotion and then even more money.
I have a sneaking suspicion that Big Business was full of deadwood since long, long before IT came around, though. IT just happens to be an area where a lot of money was being rapidly allocated to in the past few years, so a lot of incompetent buzzword throwers wound up there.
Except for the CEO's that make the most important decisions, and also work the hardest, and are the most intelligent, no one should make more than maybe $12.50/hr.
Demand and supply.
It may be really, really unpleasant and painful work to poke yourself with needles all day long. However, society doesn't really *want* people to poke themselves with needles, so there isn't any great pay for it.
I agree that programming wages will probably drop, but there's a much smaller pool of people that can code than lift boxes.
As for uncertainty...medical doctors are paid a good deal, and work in a decidedly uncertain field. Yet, should their wages be cut?
As for CEOs...I'll give you that CEOs tend to be ambitious and driven, but as for important and intelligent...well, let's take a quick glance at Carly, shall we?
Most businesses suffer from a lot of redundancy that could be automated, and that redundancy differs from business to business.
True. However, I'm not entirely sure that all this is an IT issue -- I think a good chunk of it is just big business culture. There's an awful lot of overhead that could be eliminated.
Re:I intended a Zen
on
Does IT Matter?
·
· Score: 5, Insightful
If you can have a better mousetrap built, then you better do it before your competitors do.
I disagree.
This may hold water if you're a mousetrap manufacturer. If it's your core product at stake, sure. But IT is a support system, almost always a cost center. The question is whether the latest-and-greatest from HP, Sun, and Oracle is really worth latest-and-greatest prices. Is having a product that's two years newer going to save your company how much you're dropping on it?
The huge gains that came from computers were when, say, Levi decided to computerize inventory management. Now it's computerized. They don't have to worry about warehousing junk where it isn't needed or paper mistakes costing them huge sums of money. Are they going to save more money by using a new HP system with WhizBang-3d-with-sound GUI analysis instead of IBM's old system? Maybe some, but nothing like the gains that have already been realized.
Furthermore, even if the IT spending is worthwhile, spending it on *products* may not be a good idea. If a CIO decides to drop five million dollars on software licenses, that's 100 man years of IT work that he's just exchanged for latest-and-greatest. If he does this every five years, he's losing 20 support personnel. Twenty people can get an awful lot of work done.
Walmart moving to RFID tags could save a *lot* of money, because they eliminate a lot of human work. However, there are two reasons jumping ahead like this ain't necessarily a great idea. First, RFID is a potentially big money saver, but advances like it also don't come along very often. Second, a lot of IT purchases and decisions turn out to be *bad* moves. If you let your competitors lead, and let them soak up the costs of development, only copying them when they do something that works well, yes, you lose a bit of lead time. You might have to absorb some losses. But you also gain a lot of money, and have time to let competing vendors enter the market space.
This is one reason why a lot of successful big companies are pretty conservative. Microsoft doesn't actually try very many new things for a tech company (and when it does, it tends to not do very well). Microsoft does *much* better by sitting around, waiting for masses of tiny companies to try various things, and then buy the one or two that succeed. Sure, they have to pay a hefty price for the one, but they let hordes of VCs fund their development and testing, rather than having to do it themselves. Most of Microsoft's primary products were originally developed by other companies that were then acquired.
It's pretty easy to have a couple of static tools on the system to fix things if an upgrade fails.
I believe RH packages sash and busybox. They don't have any problem, and once, long ago, after borking up ld.so (worse than just libc), I had to use 'em to recover.
How often do you have a libc upgrade fail, though, in this day of package managers and tested distributions?
Another poster asked what the point of this is -- its memory savings.
By including the above notice in a Licensed Implementation, you will be deemed to have accepted the terms and conditions of this license....A bit close to the GPL in some respects, hmm?
The first sentence, at least, is not. Furthermore, at least this sentence probably isn't even valid, in much the same way that I cannot say "in reading this post, you accept the terms of my license below.
1) Pay me a million dollars.
You can always not accept a license. Of course, then you're guilty of patent infringement, but that's your choice.
That is irrelevant. You neatly chopped off the important words in GPL 7:
"For example...".
The preceeding sentence reads:
If you cannot distribute so as to satisfy simultaneously your obligations under this License and any other pertinent obligations, then as a consequence you may not distribute the Program at all.
You cannot automatically grant sublicenses to other people using the software as per Microsoft's restrictions, therefore you cannot GPL software using Microsoft patents. The sentence you were citing (citing *part of*, to be precise) was an example, and had no legal content in the license. So you are not correct.
Slashdot Mirror
Hmm. On one hand, we have the people responsible for Linux, a nice, fast, loved system.
On the other hand, we have the man who is responsible for Java. You want a language that's a "business phenomenon" because of marketing and has plenty of technical flaws, look no further.
Joy is a bright and creative guy. If there's one thing I've learned from working with bright researchers, though, it's that doing a particular thing well doesn't make you right or even particularly well-informed or unbiased about other things. Joy likes building new things that tickle his fancy. Now he wants to throw out Java and make something new, and is pissed off because he can't because of compatibility contraints. He's already tired of Java. He got tired of Unix a long time ago. He's a nice guy to plop in a research lab and let lose. When he goes and makes blanket statements like the above, though, he speaks with little authority.
Joy doesn't like Linux because he currently in a fan of distributed computing systems. Linux is not as nice a distributed computing system as Joy's Java platform, so Linux is boring to him. The fact that general purpose distributed computing has failed to engender a hell of a lot of interest outside of the research community is no doubt a source of great irritation to many researchers, who find distributed computing a source of many facinating problems.
Joy is overlooking a number of interesting points. Joy's big accomplishments in the past have been design of huge systems -- not modules of systems. Linux is a fantastic system from a research perspective of wanting to try out new OS-level ideas. It's easy to extend and poke at Linux -- much, *much* nicer than poking at and extending Windows. Want to try a new filesystem approach? How about an idea in network scheduling? Linux makes a great research platform.
Linux is a great "product", a practical package. Joy doesn't care about something that's practically useful as long as it isn't technically new.
The point is that people like Perl so much they'll actually buy a t-shirt with the Perl camel on it or whatever to show their affinity for Perl. Nazi flags were flown by the German government.
The Nazi doctrine was not forcibly imposed on everyone (especially the early stuff). The Nazis were quite popular, though their aggressive tactics helped them get more influence faster than they normally would have. Not all of the Nazi doctrine was about stomping Jews because they were Jews -- the Nazi party was a major political party with a lot of different issues. Sure, their most egregious characteristics ended up defining them in people's minds.
A lot of people liked Nixon when he was running for election, remember. He didn't turn out that well either (though decidedly better than Adolf did).
You've had Linux *hang*?
You sure it isn't just XFree86 hanging?
This is the *second* time I've found this Adolph_Hitler guy saying things that are incredibly funny, once you consider his choice of username.
I have to agree with twiddlingbits.
It'd be quite nice to find someone that could answer the first questions, but g++ -Wall on a remotely current gcc will definitely pick out all the above problems. If the position is for a compiler or language analysis position (which I would assume *not*, given the other question), the first questions would be a good idea. For simply a C++ developer...I just don't know.
I'd be interested to see whether the applicant could fix the problems with a compiler handy and then be able to explain *why* the compiler should ask you to fix those problems, though.
The second is certainly fair. Given enough time, it'd be nice to have a third that would test more C++-design features.
"Ask me anything about C++, I will know it. I am an expert".
He struck out 3 for 3.
Many people underestimate the size of C++. I thought I knew C++ well at one point, then decided to pick up Bjarne's book. There's a *lot* of things I had never heard of in C++.
Kind of the same thing in perl, but perl is so phenomenally massive and nonunderstandable that *nobody* makes claims like being able to answer any question about perl.
Your understanding of the economic issues involved is incorrect.
The exchange rate is mostly irrelevant. Playing with currencies does little for purchasing power in anything but the quite short term.
The relevant factor is buying power. If suddenly everyone is paid in US dollars...then folks in India simply would make their current salaries in dollar amounts -- perhaps a thousand dollars a month. Basic goods would still cost perhaps a tenth US costs in India.
Terrorism does not work. Just ask Bin Laden.
I'm dubious. I don't know his goals, but he seems to have done a rather effective job.
Tactically, the jobs his organization has pulled have been stunning successes. 9/11 was amazing in terms of damage/amount of money spent. The attacks on the embassies and the USS Cole didn't require many resources either, and reverberations were felt around the world -- the United States could be hurt. Bin Laden is *still* running around out there, as far as anyone knows, despite an incredible amount of resources being spent on finding him.
Strategically, al Queda has worked well as well. It has polarized international opinion on the United States and stirred up anti-US radicals. Since its attacks, the US has invaded and occupied Iraq, a stunningly poor PR move, and now has the worst international regard since probably well before World War II. It has produced interest in resisting the United States, and has tied down a good deal of United States military power. For the amount being expended in resources by al Queda in Iraq alone, the United States and citizens of the United States pay a phenomenal sum of money in military costs. Al Queda has been largely responsible for the disasterous state of the United States air industry (the end of the economic bubble and SARS PR probably played a role as well). Economically, al Queda has had unqualified successes -- though whether by sheer luck or good reading of US leadership psychology is certainly arguable.
Does terrorism work? Sure does, though perhaps not in the traditional sense of directly toppling a government -- or Ireland and Israel would be quite different.
Good post.
countless young people, desperate and frightened, cannot.
I have a sneaking suspicion that a good number of people that are promoting Linux are taking not the role of the young, confused, people, but the role of Nike. They are attempting to exploit said confusion in others. Many people see Microsoft as a problem, and Linux as a good, and realize that making people embarassed to admit that they run Windows is likely to sway many techies towards Linux. The methods may be unpleasant, though the ends (may) be good.
For $350 I could buy a pair of English hand made brogues which would be considerably more comfortable. Since the point of paying $130 for a pair of shoes is to advertise the fact that you can afford to pay $130 for a pair of shoes a pair of Church's would work far better.
Well...I agree with your overall point, but I'm quite dubious about this. Rubber and canvas sneakers are awfully comfortable. Also, while that's subjective, I'd like to ask you to look at yourself and check that you are entirely innocent of buying those $350 shoes as a luxury item?
Incidently...while I couldn't care less about what brand my shoes are (even if I was concerned about the style, nobody looks at the things anyway, and if they did, they'd hardly care), Nike does have the nice benefit of making wide sneakers. A lot of sneakers are awfully narrow on my feet, which gets to be quite uncomfortable.
I don't see how you're rebutting any of the points in the parent's post.
Enumerating the horrors of corporate presence in third world countries is meaningless if you then simply ignore the alternatives that people have to working for them.
The jobs we ship to people "willing" to work cheaper do not, for the most part, improve the lives of those they go to. We destroy the land the local populace used to at least manage to survive on, make them wage slaves for a pittance far below liveable, and when Nike, or Union Carbide, or Walmart, or whatever company, finally gets bored and moves elsewhere, they leave slums and wastelands.
You do realize that the image of sweatshops as some sort of horrible torture system that people in other countries are tricked into entering and then prevented from escaping is almost entirely an elaborate fantasy produced by influential union organizations in the United States that are concerned about losing jobs?
This is one of the best posts I have read on Slashdot in a long, long time. It is well-written, insightful, thoughtful, and convincing. It is not offensive. It even has decent grammar. I salute you, sir.
Great, so we're in a situation so overpopulated that *everyone* should live like crap and feel the same - that they are expendable, are no better than any number of others at anything, and whose worth is no more than the cost of replacement...
As opposed to a few living like kings, and the rest living even worse? Is that what should be the case?
"Should" is such a pointless thing to argue about.
If all you do is programming, then I suppose that your job is not hard. But if your job includes analysing your company's business and using information technology to solve business problems that affect the bottom line, then your job is very hard, very valuable, not exportable, and very secure.
In the ideal, this is true. The number of people that I have seen that consider themselves crucial IT employees that "analyze their company's business" and actually do an impressive job is, regrettablly, very, very low.
If you're interested in people who actually care about saving a company money, look no further than small business. However, those companies don't *have* people that do nothing but analyze business processes because they simply don't waste enough on their business processes for this to be an issue.
I disagree.
The auto industry has:
(A) Long, 1 year minimum turnaround time on anything new that can go into a new model going out. With safety testing and the like, probably longer. If you have to wait two years while a competitor has a feature that you don't, that could be a big deal.
(B) Massive economies of scale. There is a phenomenal amount of money spent on autos. Almost anything constant cost (like designing a new feature) can be done. You only need to sell a couple of units extra a year to make another engineer worth it.
(C) A huge set of bundled technologies. Autos are, arguably, the largest single set of forcibly bundled products that most people will buy in their lifetime. Even a house can be done by a number of different contractors; you cannot combine a Ford body with a BMW dashboard (as a typical consumer). *All* these areas must be kept current -- you're going to be spending a lot of money on improvements no matter what.
(D) Autos are highly visible, and a status symbol. People will buy things for the latest-and-greatest features to show off to their friends. They are a luxury item, not a commodity.
These four points are not present in the overwhelming number of industries in the United States. Blowing lots of money on R&D on paperclips, for instance, is just silly.
I cannot begin to tell you how many hundreds of thousands of dollars were thrown around for silly and stupid reasons, mostly so Pointy Haired Bosses could play "buzzword bingo" in order to sound important and get promotions.
Absolutely. Big companies are built on middle-and-above managers starting "initiatives" which can then succeed with a dollar amount saved on a graph somewhere so that said manager can get a promotion and then even more money.
I have a sneaking suspicion that Big Business was full of deadwood since long, long before IT came around, though. IT just happens to be an area where a lot of money was being rapidly allocated to in the past few years, so a lot of incompetent buzzword throwers wound up there.
But I'll bet .01% could actually build a scalable, well-managed, backed up version that you would bet your business on.
And secure.
Except for the CEO's that make the most important decisions, and also work the hardest, and are the most intelligent, no one should make more than maybe $12.50/hr.
Demand and supply.
It may be really, really unpleasant and painful work to poke yourself with needles all day long. However, society doesn't really *want* people to poke themselves with needles, so there isn't any great pay for it.
I agree that programming wages will probably drop, but there's a much smaller pool of people that can code than lift boxes.
As for uncertainty...medical doctors are paid a good deal, and work in a decidedly uncertain field. Yet, should their wages be cut?
As for CEOs...I'll give you that CEOs tend to be ambitious and driven, but as for important and intelligent...well, let's take a quick glance at Carly, shall we?
Most businesses suffer from a lot of redundancy that could be automated, and that redundancy differs from business to business.
True. However, I'm not entirely sure that all this is an IT issue -- I think a good chunk of it is just big business culture. There's an awful lot of overhead that could be eliminated.
If you can have a better mousetrap built, then you better do it before your competitors do.
I disagree.
This may hold water if you're a mousetrap manufacturer. If it's your core product at stake, sure. But IT is a support system, almost always a cost center. The question is whether the latest-and-greatest from HP, Sun, and Oracle is really worth latest-and-greatest prices. Is having a product that's two years newer going to save your company how much you're dropping on it?
The huge gains that came from computers were when, say, Levi decided to computerize inventory management. Now it's computerized. They don't have to worry about warehousing junk where it isn't needed or paper mistakes costing them huge sums of money. Are they going to save more money by using a new HP system with WhizBang-3d-with-sound GUI analysis instead of IBM's old system? Maybe some, but nothing like the gains that have already been realized.
Furthermore, even if the IT spending is worthwhile, spending it on *products* may not be a good idea. If a CIO decides to drop five million dollars on software licenses, that's 100 man years of IT work that he's just exchanged for latest-and-greatest. If he does this every five years, he's losing 20 support personnel. Twenty people can get an awful lot of work done.
Walmart moving to RFID tags could save a *lot* of money, because they eliminate a lot of human work. However, there are two reasons jumping ahead like this ain't necessarily a great idea. First, RFID is a potentially big money saver, but advances like it also don't come along very often. Second, a lot of IT purchases and decisions turn out to be *bad* moves. If you let your competitors lead, and let them soak up the costs of development, only copying them when they do something that works well, yes, you lose a bit of lead time. You might have to absorb some losses. But you also gain a lot of money, and have time to let competing vendors enter the market space.
This is one reason why a lot of successful big companies are pretty conservative. Microsoft doesn't actually try very many new things for a tech company (and when it does, it tends to not do very well). Microsoft does *much* better by sitting around, waiting for masses of tiny companies to try various things, and then buy the one or two that succeed. Sure, they have to pay a hefty price for the one, but they let hordes of VCs fund their development and testing, rather than having to do it themselves. Most of Microsoft's primary products were originally developed by other companies that were then acquired.
Let me guess -- you don't understand why the article is important, therefore you feel the need to insult the article poster.
It's pretty easy to have a couple of static tools on the system to fix things if an upgrade fails.
I believe RH packages sash and busybox. They don't have any problem, and once, long ago, after borking up ld.so (worse than just libc), I had to use 'em to recover.
How often do you have a libc upgrade fail, though, in this day of package managers and tested distributions?
Another poster asked what the point of this is -- its memory savings.
By including the above notice in a Licensed Implementation, you will be deemed to have accepted the terms and conditions of this license....A bit close to the GPL in some respects, hmm?
The first sentence, at least, is not. Furthermore, at least this sentence probably isn't even valid, in much the same way that I cannot say "in reading this post, you accept the terms of my license below.
1) Pay me a million dollars.
You can always not accept a license. Of course, then you're guilty of patent infringement, but that's your choice.
That is irrelevant. You neatly chopped off the important words in GPL 7:
"For example...".
The preceeding sentence reads:
If you cannot
distribute so as to satisfy simultaneously your obligations under this
License and any other pertinent obligations, then as a consequence you
may not distribute the Program at all.
You cannot automatically grant sublicenses to other people using the software as per Microsoft's restrictions, therefore you cannot GPL software using Microsoft patents. The sentence you were citing (citing *part of*, to be precise) was an example, and had no legal content in the license. So you are not correct.