Apart from the unnecessary obscurity of the code, the code actually simply computes the wrong values with current versions of gcc and optimization turned on. See here:
I stated quite plainly that HP commited identity theft, theft as in crime, and therefore it is appropriate that so many senior level people lost their jobs.
Losing one's job is not punishment for a crime, it's an expression that one is not fit for the job anymore. That's also why there is no due process or "innocent until proven guilty" restriction on losing your job. HP would be perfectly justified in firing these people even if there were no criminal or civil charges against them--the bad press that has resulted from their actions alone is sufficient to make these people unsuitable for their jobs. Conversely, by losing their jobs, these people have not been punished for criminal misconduct in any way.
As for the criminal part, in fact, Dunn and the other executives have been charged by the state AG. The possible punishment for those crimes includes both fines and jail time. The fines are so laughably small ($10,000) that they have no deterrent effect on people like Dunn.
Or is that simply your perception, which appears to be rather influenced by the fact that she's rich?
My perception is that white collar criminals often get away without jail time, when other defendants committing the same kind of crime would go to jail. I'm simply saying that she should get some jail time, because that is necessary as a deterrent, because it fits the severity of the crime, and because this case should be used to set a public example that this kind of behavior is not acceptable.
The code may or may not be a good way of solving the problem, but it's lousy code nonetheless:
* it's badly named: "reciprocal square root" would be better (however, that's a problem with graphics lingo) * it lacks a comment as to where it came from * it lacks a comment as to how it works or how the constant was determined * it lacks a comment as to how well it works * it lacks a comment as to the range of arguments it's valid for * it lacks (optional) error checking * it lacks a comment as to what environment it works in * it lacks a straightforward base implementation * there don't seem to be any test cases either
The code posted in the article is actually not exactly the Q3 source code. The Q3 source code actually contains two slightly different versions of the function, one called "Q_rsqrt" commented (with "// what the fuck") and a single iteration, and second one that's uncommented and named incorrectly "SquareRootFloat" and uses two iterations. Neither of them say what the function actually does, and the fact that there are two of them is also not so great.
Yeah, that kind of code causes me "to go wobbly in the knees"--by annoying the hell out of me--because the programmer who wrote it is wasting my time. Pretty predictably, sooner or later, this code will break something, and it will first take time to figure out that this code is responsible, then it will take some time to figure out what it is supposed to do, and finally, it will be a pain to figure out how to fix it. Or, I'd just rewrite it.
lying is and always will be considered acceptable to get information (ask a journalist or a security expert)
You're playing word games by hiding what amounts to fraud behind the term "lying". Obtaining personal records without authorization by misrepresentation is fraud, no matter who does it. It is not acceptable behavior for anybody.
It's a standard investigative technique. Try watching a TV watchdog show where it is used to catch out corporations, or cowboy builders ripping off old ladies.
That's a different kind of "lying"; whether it is illegal depends on the circumstances (usually it isn't because those people are doing their homework, or because it's staged).
Bitching about jail time just because she's rich is, as far as I know, just bullshit.
No, I'm "bitching" about the fact that it looks like she is getting off easy because she is rich.
You're making a good point. But the correct word ought to be "deception", "fraud", or "misrepresentation". "Pretexting", on the other hand, is a weasel word.
As far as you claiming to have a wife on Slashdot, you underestimate the harm you're doing: dozens of nerds are jumping off the roof thinking that they are failures because you have a wife and they don't.
You know what - no-one outside of HP has been harmed.
If this sets a precedent that it's OK for companies to lie in order to get information about their board or employees, then people have very much been harmed.
But if you didn't notice, the *chairwoman* of a huge damn company lost her job. It will be costing her a fortune in legal fees
Well, and she'll still be richer than 99.99% of the US population and she won't be doing any jail time. Why is that a good thing? I mean, she couldn't be trusted to be in the typing pool with her apparent lack of ethics and judgement.
Bill Gates was running an organization that is more like the USSR than like a democracy: he had nearly absolute power to do as he pleased and he was planning his economy centrally. He could get rid of people he didn't like, he could create and kill projects as he pleased, etc. He also got a huge war-chest of money and a monopoly handed to him on a silver plate by IBM.
Perhaps people forget, but the USSR was initially very successful, both economically and scientifically (e.g., Sputnik). But a few decades later, it all fell apart. And it's no accident that Microsoft is slowly collapsing under its own weight, despite still having huge amounts of money. Organizations like the USSR and like Microsoft simply aren't successful long-term.
Overall, their job experience simply doesn't qualify US business leaders for high political office (of course, some of them may still figure it out and end up being good politicians, but that's not because of their business experience). To succeed as a politician in a democracy, people need to negotiate, compromise, build alliances, convince, have charisma, and do something reasonable even in the face of severe budget shortfalls. Gates doesn't strike as being capable of doing any of those well.
For example, Plone ships with its own version of Python and Zope to keep the host OS's versions of either from breaking the application, and lets you update the host OS independently of the application. This is a good thing.
It's only a good thing if you don't actually want to use any add-on packages. As soon as you want a Python package that your Plone package doesn't support, you have to manually install it. And if everybody did what Plone did, you'd have to install the extra package separately for each of the "own versions of Python".
In fact, good Linux distributions integrate Plone with the distribution-standard version of Python. That's a good thing; shipping Plone with a custom version of Python is an evil crutch.
Bullshit if they're releasing it under the GPL then there is no *patent issue*
There is no patent issue for the code they are releasing under the GPL and code derived from that, but the patent issue remains for other implementations of Java. That's great if you can use Sun's code or code derived from Sun's code. But Sun's patents still are valid and in force, and they can still very much sue any other third party implementation if they like. For example, Sun can still assert their patents against Apache Harmony, and theoretically even against GNU gcj and Classpath.
Provided Sun actually delivers (so far, all we have is a promise), it's an OK compromise and a big improvement over the way things have been. I can live with that, as can most other people, and it makes Java a feasible choice again and restores some of Sun's lost credibility. But it's not the same as if Java were an open standard.
This deal is good for Novell at the expense of everybody else
What do other people lose through this deal? It's legally meaningless as far as patents are concerned, and it probably won't get Novell any more business.
Which just shows the mono and c# loyalists who ignored the patent and legal implicantions with it because it was better than java are idiots.
Until very recently, the legal situation surrounding Java was much worse: Sun owns patents and copyrights up the wazoo on Java. That issue won't be addressed by open sourcing Java. The legal situation surrounding.NET is still better, although the FUD Microsoft is spreading using their hokey patent is certainly a problem.
Now since java is GPLed its time to abandon Mono
Java has not been GPL'ed yet, Sun has merely talked about their intent to do so and they have lied before. Until the entire codebase is out under GPL with linking exception, then, and only then, is Java even in the running.
Furthermore, open sourcing Sun's Java implementation still doesn't make the standard an open standard, since every implementation other than Sun's GPL'ed implementation is still susceptible to patent lawsuits from Sun.
In addition, open sourcing Java won't fix its numerous technical problems.
Finally, Microsoft's patent isn't relevant to most users of Mono, since most users of Mono don't use.NET.
Basically, you are part of the Microsoft FUD machine. Do you work for Microsoft by any chance?
Overall, the alternative to Mono isn't Java, it's going back to C++, or moving to D or Eiffel.
It was a rhetorical question; you should look up the concept.
My point was that your list of reasons made it sound like open source developers targeted Microsoft because they had some sort of irrational prejudices.
In fact, the reason to target Microsoft products is rational and economic, just like Microsoft's reason for being in business. Unlike Microsoft, open source developers don't engage in illegal monopolistic practices and instead make fulfilling the needs of users their primary objective. That's because they are the users: many open source developers have been screwed by companies like Microsoft one too many times and want to reduce their exposure to the risk that comes with building their business on proprietary software.
What I was getting at, is that maybe Open Source needs to not try to compete with Exchange, and come up with a whole new thinking behind Messaging\Calendar\Scheduling instead?
Why should open source do that? I mean, Microsoft copied just about every feature in Exchange from others.
In any case, open source has excellent and innovative solutions for messaging, calendaring, and scheduling. Attempts to be compatible with Exchange are there to make it easy for users to migrate off the Microsoft solution, not because they are the preferred open source solution.
What stops Open Source? Lack of a profit motive.
I don't see open source being "stopped" by anything. And you're seriously out of touch if you think that open source isn't motivated by hardcore dollars-and-cents considerations.
Why do so many think that Microsoft should just play fair and give up? You would have to be some kind of nut to think that Microsoft would not use every advantage available to them to stomp competition. That is BUSINESS!
Business are required by law to compete fairly. Exclusionary or monopolistic behavior is illegal. Unfortunately, it's a costly and lengthy procedure to hold companies responsible for violations, which is why Microsoft can get away with it again and again.
There were about 43000 traffic fatalities in 2002, and the number of annual deaths from the flu is about 30000 to 50000. Note that these figures could be easily reduced with fairly small investments of money.
There are many other sources of preventable death. For example the US infant mortality is somewhere around 6.4-9/1000, while it is around 4-5/1000 in other industrialized nations. There are around 4M births each year (14/1000 * 300m). That means that there are roughly 8000 excess infant deaths in the US every year compared to what we should have if our government delivered health-related services as effectively as other industrialized nations. Or, to put it more bluntly, since Gingrich is actually responsible for keeping back changes to US health care in the direction of other industrialized nations and has failed to bring down infant mortality rates to industrialized levels when he was in power, you might say that Gingrich is a baby killer who is responsible for more infant deaths every year than bin Laden has killed in his lifetime through terrorist attacks.
I'm really just curious at this point what is driving the anti-Exchange bandwagon, because I don't see a real, viable competitor out there.
Yes, and that is exactly why there need to be competitors for Exchange. Maybe in Microsoft group-think, a single proprietary product from Microsoft is the way the world should run, but in reality, we live in a free market and buyers should have a choice. And they need a choice so that the client access license costs of $67/client are driven down.
Of course, the reasons buyers don't have a choice right now is because Microsoft has largely killed all the commercial alternatives through anti-competitive behavior like bundling, tying, and proprietary protocols. Open source is the only entity still capable of challenging Microsoft and giving users a choice.
I would like someone to honestly tell me either that Exchange has problems that need fixing,
You know, this question coming from someone on the Exchange team just leaves people speechless. To answer your question, apart from its anti-competitive design, yes, Exchange has technical problems.
For many open-source idealists, it is a major target because it is (1) commonly used, (2) in a business-critical role, and (3) close-source.
And Microsoft developers want Microsoft software to succeed because... what? The good of humanity? Who are you kidding?
Exchange is a target for replacement by open source software for the same reason a lot of commercial software is: its closed source nature causes numerous problems for end users: it's buggy, it's unreliable, it interferes with free market and customer choice, and it's way overpriced.
Of course, Exchange needs to be replaced. If open source didn't exist, it would have to be replaced by commercial competitors. It's just that Microsoft's monopolistic practices have killed all their serious commercial competitors, so that open source software is the only challenger that's left and that's strong enough to take them on.
Gingrich is the same kind of politician that asserted that waging war on Iraq would make us safer, when, in fact, it has done the opposite. And now, he is making similarly wild assertions about how restricting free speech would make us safer.
The problem here is not any difficulty of dealing with terrorism, the problem is that Gingrich and politicians like him are completely and utterly incompetent.
Tragic as it is, an instance of 3000 deaths does not warrant throwing away our democracy or spending billions of dollars on ill-conceived wars; we have tens of thousands of preventable deaths from the flu and from traffic accidents each every year.
And maybe Gingrich didn't notice, but we did lose a city recently. That loss would have been completely preventable if people like Gingrich had done their job. And it would have been preventable at a fraction of the cost of the current anti-terrorism measures and without destroying our democracy.
So, you're saying that it's "political" and it's about "fear, uncertainty, and doubt". Well, it is, and it should be. The military is not a game, it's about loss, fear, boredom, injury, limited career and advancement options, destruction, bureaucracy, disease, grief, killing, and being killed. If you don't have "fear, uncertainty, and doubt" about that, there is something wrong with you as a human being. And when the military recruits impressionable young people with games that give them a completely unrealistic picture of the choice they are making, it is perfectly justifiable to criticize them.
Note that I'm not saying that the military is an overall bad career or that military service is intrinsically wrong. The military serves an important function in the defense of our democracy, and we should be grateful to the people who choose military service. But we don't do anybody a favor by pretending that it's all a just a fun game.
Most patents, especially for software, are obvious after the fact. Programmers look and say, man, that's so obvious!
Actually, the usual case for software patents in my experience is that people say "hey, I have been using that for years, how in the world can anybody get a patent on that".
Furthermore, granting someone a patent costs society a lot; we should err on the side of granting too few patents, not too many, and the burden of proof that something is "unobvious" should be on the person filing the patent.
Ask anyone who has submitted a patent application whether they felt their patent was frivolous. I imagine you'd find the vast majority of them holding the belief that they did something novel.
I actually have a bunch of patents. I applied for them because my employer required it, and my employer required it for trading with other companies. I think the patent system should be changed so that what I hold patents on shouldn't be patentable. At the same time, I think my patents are better than the great majority of patents that are currently being granted.
I've invented a bunch of things since then, which I think are considerably more useful. I still haven't patented, simply because there has been no point: as an individual inventor, patents wouldn't have helped me at all, and the cost of getting them is high.
The patent system is broken; it doesn't help people who invent things, it only helps lawyers and big companies that want to keep new competitors out of the market.
What is the reward for developing open source software?
Most people who do it get paid, just like for any other kind of software development.
Not everybody operates with an altruistic "I'm giving back to the community" motivation.
Nor do most open source developers or companies.
What am I missing here?
You're making the false assumptions that there is a finite amount of software to be developed, and that in the absence of open source software, people could simply go on making lots of money with the same old products. Neither is true. In fact, the cost of today's expensive software tends quickly towards zero over the next few years, with or without open source licenses. Arguably, open source is simply an efficient free market mechanism for making that happen.
And I'm not being sarcastic with that, I genuinely don't understand why anyone would want to share the fundamentals of their creation in a way that would compromise any potential future earnings.
Again, you're starting from a false set of assumptions. The software you develop today is probably going to be worthless in a few years. If someone pays you to release it open source, or if there are other benefits to you for releasing it open source, that's the economically rational thing to do.
You can be certain that most people who release open source software do it out of economic self-interest.
So, it looks like Microsoft paid for Novell's signature on that "patent agreement". Novell couldn't say "no" to that big of an instant payoff.
Yes, indeed. And that also makes the agreement pretty meaningless. "I give you a net amount of several hundred million dollars, together with a license to all my patents" simply does not demonstrate that someone's patents have any value.
You're right that it's almost like Linux. The crucial difference is how projects get selected and evaluated.
In Windows, if some yahoo inside the company decides that database-backed file systems with ACL and some hokey metadata schemes are the best thing since sliced bread and sweet-talks his managers into putting it on the feature list, then they are committed to shipping it. Furthermore, the company will rally the resources behind the developer to make it happen. That means that people with great PowerPoint slides but dumb ideas will get to ship their stuff. Unfortunately, computer science is full of dumb ideas that look good on paper and in mockups, and that's in part why Windows is so full of crap, under which some nuggets of good engineering are buried.
In Linux, if some yahoo has a dumb idea, they first have to put in their own time and money to develop it (or convince some people to part with hard-earned cash). Then, they have to build a user community around it, which means the software has to work and be useful. For Linux, this mechanism operates at the level of every single package or project and it tends to weed out most of the bad ideas; "weed out" not in the sense that the software is completely unavailable, but in the sense that the bad ideas don't interfere with the good, useful software.
"Upending nearly a quarter-century of jurisprudence at this point would throw into question the validity of millions of issued patents
Good, that's the point: there are millions of bad patents, and they must be invalidated. The longer we wait, the more painful it gets.
Somehow, these companies are using the same kind of arguments little children would use: at first "aw, it's not that bad yet, let's just see what develops" and then later "aw, but we have been doing it so long, we can't change now".
cause the reassessment of patent licenses worth billions of dollars
As well it should: those billions of dollars have been forced out of companies for obvious inventions, and they should be restored to the victims.
make patent litigation more difficult to settle, and inevitably create more litigation for the courts," the brief said.
Well, I expect it will create more litigation to clean up the patent mess. But that argument is kind of like saying "well, we know the guy is innocent, but let's execute him anyway because it's so much cheaper and simpler".
Slashdot Mirror
Apart from the unnecessary obscurity of the code, the code actually simply computes the wrong values with current versions of gcc and optimization turned on. See here:
t ml
t ml
http://gcc.gnu.org/ml/gcc-bugs/2006-03/msg02943.h
http://gcc.gnu.org/ml/gcc-bugs/2006-03/msg02957.h
It also makes unstated assumptions about the values it gets called with; call it with something else and you get bad results.
I stated quite plainly that HP commited identity theft, theft as in crime, and therefore it is appropriate that so many senior level people lost their jobs.
Losing one's job is not punishment for a crime, it's an expression that one is not fit for the job anymore. That's also why there is no due process or "innocent until proven guilty" restriction on losing your job. HP would be perfectly justified in firing these people even if there were no criminal or civil charges against them--the bad press that has resulted from their actions alone is sufficient to make these people unsuitable for their jobs. Conversely, by losing their jobs, these people have not been punished for criminal misconduct in any way.
As for the criminal part, in fact, Dunn and the other executives have been charged by the state AG. The possible punishment for those crimes includes both fines and jail time. The fines are so laughably small ($10,000) that they have no deterrent effect on people like Dunn.
Or is that simply your perception, which appears to be rather influenced by the fact that she's rich?
My perception is that white collar criminals often get away without jail time, when other defendants committing the same kind of crime would go to jail. I'm simply saying that she should get some jail time, because that is necessary as a deterrent, because it fits the severity of the crime, and because this case should be used to set a public example that this kind of behavior is not acceptable.
The code may or may not be a good way of solving the problem, but it's lousy code nonetheless:
* it's badly named: "reciprocal square root" would be better (however, that's a problem with graphics lingo)
* it lacks a comment as to where it came from
* it lacks a comment as to how it works or how the constant was determined
* it lacks a comment as to how well it works
* it lacks a comment as to the range of arguments it's valid for
* it lacks (optional) error checking
* it lacks a comment as to what environment it works in
* it lacks a straightforward base implementation
* there don't seem to be any test cases either
The code posted in the article is actually not exactly the Q3 source code. The Q3 source code actually contains two slightly different versions of the function, one called "Q_rsqrt" commented (with "// what the fuck") and a single iteration, and second one that's uncommented and named incorrectly "SquareRootFloat" and uses two iterations. Neither of them say what the function actually does, and the fact that there are two of them is also not so great.
Yeah, that kind of code causes me "to go wobbly in the knees"--by annoying the hell out of me--because the programmer who wrote it is wasting my time. Pretty predictably, sooner or later, this code will break something, and it will first take time to figure out that this code is responsible, then it will take some time to figure out what it is supposed to do, and finally, it will be a pain to figure out how to fix it. Or, I'd just rewrite it.
lying is and always will be considered acceptable to get information (ask a journalist or a security expert)
You're playing word games by hiding what amounts to fraud behind the term "lying". Obtaining personal records without authorization by misrepresentation is fraud, no matter who does it. It is not acceptable behavior for anybody.
It's a standard investigative technique. Try watching a TV watchdog show where it is used to catch out corporations, or cowboy builders ripping off old ladies.
That's a different kind of "lying"; whether it is illegal depends on the circumstances (usually it isn't because those people are doing their homework, or because it's staged).
Bitching about jail time just because she's rich is, as far as I know, just bullshit.
No, I'm "bitching" about the fact that it looks like she is getting off easy because she is rich.
Gates has been good at running Microsoft; GWB would be a failure at that.
But both are bad choices for running the nation.
You're making a good point. But the correct word ought to be "deception", "fraud", or "misrepresentation". "Pretexting", on the other hand, is a weasel word.
As far as you claiming to have a wife on Slashdot, you underestimate the harm you're doing: dozens of nerds are jumping off the roof thinking that they are failures because you have a wife and they don't.
You know what - no-one outside of HP has been harmed.
If this sets a precedent that it's OK for companies to lie in order to get information about their board or employees, then people have very much been harmed.
But if you didn't notice, the *chairwoman* of a huge damn company lost her job. It will be costing her a fortune in legal fees
Well, and she'll still be richer than 99.99% of the US population and she won't be doing any jail time. Why is that a good thing? I mean, she couldn't be trusted to be in the typing pool with her apparent lack of ethics and judgement.
Bill Gates was running an organization that is more like the USSR than like a democracy: he had nearly absolute power to do as he pleased and he was planning his economy centrally. He could get rid of people he didn't like, he could create and kill projects as he pleased, etc. He also got a huge war-chest of money and a monopoly handed to him on a silver plate by IBM.
Perhaps people forget, but the USSR was initially very successful, both economically and scientifically (e.g., Sputnik). But a few decades later, it all fell apart. And it's no accident that Microsoft is slowly collapsing under its own weight, despite still having huge amounts of money. Organizations like the USSR and like Microsoft simply aren't successful long-term.
Overall, their job experience simply doesn't qualify US business leaders for high political office (of course, some of them may still figure it out and end up being good politicians, but that's not because of their business experience). To succeed as a politician in a democracy, people need to negotiate, compromise, build alliances, convince, have charisma, and do something reasonable even in the face of severe budget shortfalls. Gates doesn't strike as being capable of doing any of those well.
For example, Plone ships with its own version of Python and Zope to keep the host OS's versions of either from breaking the application, and lets you update the host OS independently of the application. This is a good thing.
It's only a good thing if you don't actually want to use any add-on packages. As soon as you want a Python package that your Plone package doesn't support, you have to manually install it. And if everybody did what Plone did, you'd have to install the extra package separately for each of the "own versions of Python".
In fact, good Linux distributions integrate Plone with the distribution-standard version of Python. That's a good thing; shipping Plone with a custom version of Python is an evil crutch.
Bullshit if they're releasing it under the GPL then there is no *patent issue*
There is no patent issue for the code they are releasing under the GPL and code derived from that, but the patent issue remains for other implementations of Java. That's great if you can use Sun's code or code derived from Sun's code. But Sun's patents still are valid and in force, and they can still very much sue any other third party implementation if they like. For example, Sun can still assert their patents against Apache Harmony, and theoretically even against GNU gcj and Classpath.
Provided Sun actually delivers (so far, all we have is a promise), it's an OK compromise and a big improvement over the way things have been. I can live with that, as can most other people, and it makes Java a feasible choice again and restores some of Sun's lost credibility. But it's not the same as if Java were an open standard.
What they _should_ have done was to keep things compatible in the first place.
.NET, Mono/GNU, and Java.
Novell did: Mono provides seamless integration of CLR and JVM code. So, on Mono, you can run Microsoft
This deal is good for Novell at the expense of everybody else
What do other people lose through this deal? It's legally meaningless as far as patents are concerned, and it probably won't get Novell any more business.
Which just shows the mono and c# loyalists who ignored the patent and legal implicantions with it because it was better than java are idiots.
.NET is still better, although the FUD Microsoft is spreading using their hokey patent is certainly a problem.
.NET.
Until very recently, the legal situation surrounding Java was much worse: Sun owns patents and copyrights up the wazoo on Java. That issue won't be addressed by open sourcing Java. The legal situation surrounding
Now since java is GPLed its time to abandon Mono
Java has not been GPL'ed yet, Sun has merely talked about their intent to do so and they have lied before. Until the entire codebase is out under GPL with linking exception, then, and only then, is Java even in the running.
Furthermore, open sourcing Sun's Java implementation still doesn't make the standard an open standard, since every implementation other than Sun's GPL'ed implementation is still susceptible to patent lawsuits from Sun.
In addition, open sourcing Java won't fix its numerous technical problems.
Finally, Microsoft's patent isn't relevant to most users of Mono, since most users of Mono don't use
Basically, you are part of the Microsoft FUD machine. Do you work for Microsoft by any chance?
Overall, the alternative to Mono isn't Java, it's going back to C++, or moving to D or Eiffel.
It was a rhetorical question; you should look up the concept.
My point was that your list of reasons made it sound like open source developers targeted Microsoft because they had some sort of irrational prejudices.
In fact, the reason to target Microsoft products is rational and economic, just like Microsoft's reason for being in business. Unlike Microsoft, open source developers don't engage in illegal monopolistic practices and instead make fulfilling the needs of users their primary objective. That's because they are the users: many open source developers have been screwed by companies like Microsoft one too many times and want to reduce their exposure to the risk that comes with building their business on proprietary software.
What I was getting at, is that maybe Open Source needs to not try to compete with Exchange, and come up with a whole new thinking behind Messaging\Calendar\Scheduling instead?
Why should open source do that? I mean, Microsoft copied just about every feature in Exchange from others.
In any case, open source has excellent and innovative solutions for messaging, calendaring, and scheduling. Attempts to be compatible with Exchange are there to make it easy for users to migrate off the Microsoft solution, not because they are the preferred open source solution.
What stops Open Source? Lack of a profit motive.
I don't see open source being "stopped" by anything. And you're seriously out of touch if you think that open source isn't motivated by hardcore dollars-and-cents considerations.
Why do so many think that Microsoft should just play fair and give up? You would have to be some kind of nut to think that Microsoft would not use every advantage available to them to stomp competition. That is BUSINESS!
Business are required by law to compete fairly. Exclusionary or monopolistic behavior is illegal. Unfortunately, it's a costly and lengthy procedure to hold companies responsible for violations, which is why Microsoft can get away with it again and again.
There were about 43000 traffic fatalities in 2002, and the number of annual deaths from the flu is about 30000 to 50000. Note that these figures could be easily reduced with fairly small investments of money.
There are many other sources of preventable death. For example the US infant mortality is somewhere around 6.4-9/1000, while it is around 4-5/1000 in other industrialized nations. There are around 4M births each year (14/1000 * 300m). That means that there are roughly 8000 excess infant deaths in the US every year compared to what we should have if our government delivered health-related services as effectively as other industrialized nations. Or, to put it more bluntly, since Gingrich is actually responsible for keeping back changes to US health care in the direction of other industrialized nations and has failed to bring down infant mortality rates to industrialized levels when he was in power, you might say that Gingrich is a baby killer who is responsible for more infant deaths every year than bin Laden has killed in his lifetime through terrorist attacks.
I'm really just curious at this point what is driving the anti-Exchange bandwagon, because I don't see a real, viable competitor out there.
Yes, and that is exactly why there need to be competitors for Exchange. Maybe in Microsoft group-think, a single proprietary product from Microsoft is the way the world should run, but in reality, we live in a free market and buyers should have a choice. And they need a choice so that the client access license costs of $67/client are driven down.
Of course, the reasons buyers don't have a choice right now is because Microsoft has largely killed all the commercial alternatives through anti-competitive behavior like bundling, tying, and proprietary protocols. Open source is the only entity still capable of challenging Microsoft and giving users a choice.
I would like someone to honestly tell me either that Exchange has problems that need fixing,
You know, this question coming from someone on the Exchange team just leaves people speechless. To answer your question, apart from its anti-competitive design, yes, Exchange has technical problems.
For many open-source idealists, it is a major target because it is (1) commonly used, (2) in a business-critical role, and (3) close-source.
And Microsoft developers want Microsoft software to succeed because... what? The good of humanity? Who are you kidding?
Exchange is a target for replacement by open source software for the same reason a lot of commercial software is: its closed source nature causes numerous problems for end users: it's buggy, it's unreliable, it interferes with free market and customer choice, and it's way overpriced.
Of course, Exchange needs to be replaced. If open source didn't exist, it would have to be replaced by commercial competitors. It's just that Microsoft's monopolistic practices have killed all their serious commercial competitors, so that open source software is the only challenger that's left and that's strong enough to take them on.
Gingrich is the same kind of politician that asserted that waging war on Iraq would make us safer, when, in fact, it has done the opposite. And now, he is making similarly wild assertions about how restricting free speech would make us safer.
The problem here is not any difficulty of dealing with terrorism, the problem is that Gingrich and politicians like him are completely and utterly incompetent.
Tragic as it is, an instance of 3000 deaths does not warrant throwing away our democracy or spending billions of dollars on ill-conceived wars; we have tens of thousands of preventable deaths from the flu and from traffic accidents each every year.
And maybe Gingrich didn't notice, but we did lose a city recently. That loss would have been completely preventable if people like Gingrich had done their job. And it would have been preventable at a fraction of the cost of the current anti-terrorism measures and without destroying our democracy.
So, you're saying that it's "political" and it's about "fear, uncertainty, and doubt". Well, it is, and it should be. The military is not a game, it's about loss, fear, boredom, injury, limited career and advancement options, destruction, bureaucracy, disease, grief, killing, and being killed. If you don't have "fear, uncertainty, and doubt" about that, there is something wrong with you as a human being. And when the military recruits impressionable young people with games that give them a completely unrealistic picture of the choice they are making, it is perfectly justifiable to criticize them.
Note that I'm not saying that the military is an overall bad career or that military service is intrinsically wrong. The military serves an important function in the defense of our democracy, and we should be grateful to the people who choose military service. But we don't do anybody a favor by pretending that it's all a just a fun game.
Most patents, especially for software, are obvious after the fact. Programmers look and say, man, that's so obvious!
Actually, the usual case for software patents in my experience is that people say "hey, I have been using that for years, how in the world can anybody get a patent on that".
Furthermore, granting someone a patent costs society a lot; we should err on the side of granting too few patents, not too many, and the burden of proof that something is "unobvious" should be on the person filing the patent.
Ask anyone who has submitted a patent application whether they felt their patent was frivolous. I imagine you'd find the vast majority of them holding the belief that they did something novel.
I actually have a bunch of patents. I applied for them because my employer required it, and my employer required it for trading with other companies. I think the patent system should be changed so that what I hold patents on shouldn't be patentable. At the same time, I think my patents are better than the great majority of patents that are currently being granted.
I've invented a bunch of things since then, which I think are considerably more useful. I still haven't patented, simply because there has been no point: as an individual inventor, patents wouldn't have helped me at all, and the cost of getting them is high.
The patent system is broken; it doesn't help people who invent things, it only helps lawyers and big companies that want to keep new competitors out of the market.
What is the reward for developing open source software?
Most people who do it get paid, just like for any other kind of software development.
Not everybody operates with an altruistic "I'm giving back to the community" motivation.
Nor do most open source developers or companies.
What am I missing here?
You're making the false assumptions that there is a finite amount of software to be developed, and that in the absence of open source software, people could simply go on making lots of money with the same old products. Neither is true. In fact, the cost of today's expensive software tends quickly towards zero over the next few years, with or without open source licenses. Arguably, open source is simply an efficient free market mechanism for making that happen.
And I'm not being sarcastic with that, I genuinely don't understand why anyone would want to share the fundamentals of their creation in a way that would compromise any potential future earnings.
Again, you're starting from a false set of assumptions. The software you develop today is probably going to be worthless in a few years. If someone pays you to release it open source, or if there are other benefits to you for releasing it open source, that's the economically rational thing to do.
You can be certain that most people who release open source software do it out of economic self-interest.
So, it looks like Microsoft paid for Novell's signature on that "patent agreement". Novell couldn't say "no" to that big of an instant payoff.
Yes, indeed. And that also makes the agreement pretty meaningless. "I give you a net amount of several hundred million dollars, together with a license to all my patents" simply does not demonstrate that someone's patents have any value.
You're right that it's almost like Linux. The crucial difference is how projects get selected and evaluated.
In Windows, if some yahoo inside the company decides that database-backed file systems with ACL and some hokey metadata schemes are the best thing since sliced bread and sweet-talks his managers into putting it on the feature list, then they are committed to shipping it. Furthermore, the company will rally the resources behind the developer to make it happen. That means that people with great PowerPoint slides but dumb ideas will get to ship their stuff. Unfortunately, computer science is full of dumb ideas that look good on paper and in mockups, and that's in part why Windows is so full of crap, under which some nuggets of good engineering are buried.
In Linux, if some yahoo has a dumb idea, they first have to put in their own time and money to develop it (or convince some people to part with hard-earned cash). Then, they have to build a user community around it, which means the software has to work and be useful. For Linux, this mechanism operates at the level of every single package or project and it tends to weed out most of the bad ideas; "weed out" not in the sense that the software is completely unavailable, but in the sense that the bad ideas don't interfere with the good, useful software.
"Upending nearly a quarter-century of jurisprudence at this point would throw into question the validity of millions of issued patents
Good, that's the point: there are millions of bad patents, and they must be invalidated. The longer we wait, the more painful it gets.
Somehow, these companies are using the same kind of arguments little children would use: at first "aw, it's not that bad yet, let's just see what develops" and then later "aw, but we have been doing it so long, we can't change now".
cause the reassessment of patent licenses worth billions of dollars
As well it should: those billions of dollars have been forced out of companies for obvious inventions, and they should be restored to the victims.
make patent litigation more difficult to settle, and inevitably create more litigation for the courts," the brief said.
Well, I expect it will create more litigation to clean up the patent mess. But that argument is kind of like saying "well, we know the guy is innocent, but let's execute him anyway because it's so much cheaper and simpler".