What matters is that I go to Dixons (UK electronics store), approach a shelf with subnotebooks and see a sign "Linux notebooks will not work with mobile Internet".
And my point is, "useful to make money with" does not correlate to "actually useful", or "net positive impact on humanity".
Oh please, can we bury this hubris-overloaded programmer notion of "my code will save the world"? If the code makes someone money, it's useful for them, period. And debating on the net impact on humanity of this or that will lead us nowhere, there are too many variables involved.
COBOL is still widely used, and COBOL programmers are highly paid. I really wouldn't want to work with COBOL, though.
You are free to like or dislike any language, it doesn't tell me anything about it's usefulness.
As long as I don't have to edit the source, I really don't care much what language a program is written in.
Me too, unless an unlikely situation arises in which I hit a showstopper bug and the time to fix the bug is increased by the problems of the language.
No, it will check that the program is correct. Types are a laughably small part of that.
We both know that it's rare when 100% of the program is being tested. Static typing guarantees you that 100% of function signatures and variable assignments are tested for correct typing. In a dynamic language, you're on your own.
Generally tests do not break in the same way that the program itself breaks,
Sure, the first (or last) thing to do with a broken tests is to remove it;-)
Second: You're going to need those tests anyway. On a smaller program, you might get away with no tests -- but you probably don't really need the types there, either. On a lager program, it's not enough to know that "public int foo()" returns an integer -- you need to know that it actually does what it's supposed to do.
Look, I *know* what are the benefits of unit tests. I just don't like writing them when I could get it done by the compiler.
Once you know that, why do you care whether it returns an integer?
Because I might be awfully surprised when this integer turns into a float.
And every time I look at it, it seems like the only thing Java has that Ruby doesn't (language-wise) is more restrictions -- static type checking, actually-private values and methods,
Have you ever worked on a large codebase, like 1E5-1E6 lines of code? if you did, you would appreciate the benefits of private values and methods, of having a clean interface and knowing that you can reorganize the inner workings of a class without breaking the code of this guy two floors above you. Without private variables and methods, you risk creating A Giant Ball Of Mud, where everything depends on everything and it gets the point where you are afraid to change *anything* because it might break things in ten places. Been there, done that. It ain't pretty and it ain't pleasant. If your Ruby doesn't allow me to encapsulate my implementation, then I'll never use it for anything more involved than a small script or quick prototype.
"Irrelevant. Lots of companies make money selling antivirus software."
1. You're comparing apples with oranges here. From the software developer's POV, Java is a tool while antivirus program is a product. The fact that I make lots of money selling some product does not have to correlate with the product's usefulness, but if I make lots of money using a tool, it means that this tool is indeed USEFUL TO MAKE MONEY WITH. And if Java is useful to make money with, that's splendid for me.
2. Companies also make money by USING software written in Java. I seldom hear someone claiming that "hey, I've made money on using Norton Antivirus!". Antivirus software helps you not to lose money rather than make them.
"That's what we have unit tests for."
If I unit test module A and module B, but then module A feeds module B with different type of data then B expects, then unit testing these modules won't help you much (since you will take care in your tests to supply the correct types). Oh, you say that I should unit test module A and module B together, to make sure that cooperate correctly -- but what happened to "unit" in "unit tests"? We will end up with one giant blob of a test, which checks whether the types are correct.
"And honestly, type checking is just a subset of unit testing."
I love such impractical analogies. You miss out a crucial point that with static typing, you get these "unit tests" FOR FREE, generated by the compiler without ANY EFFORT ON YOUR PART and without creating any testing code to debug and maintain.
"What's more, it's a lot easier to implement static type checking on top of a more dynamic language than the other way around."
Maybe.
"Those 20 minutes is probably considerably less time than it would take for my first attempt in Java"
Should it tell me something about Java, or about you?
BTW, I've been coding quite a lot in Matlab, which is also dynamic, and I've never need "eval" there. Anonymous functions, on the other hand... I would like to see them in Java. Them and the const correctness.
Lots of companies make lots of money writing software in Java or using software written in this language. I have developed software in C, in Java and in Matlab (which is a dynamic language of sorts). Each of them has its uses. Saying that dynamic languages are always better than static is just stupid. Dynamic languages are faster to develop in because they place less restrictions on the programmer. For the same reason, they are also less safe, especially when the codebase becomes more complicated. I've had weird errors in Matlab prop up because instead of passing a number, I've passed a string to my procedure. In a statically typed language, it wouldn't happen. Anyway, Java does allow you to dynamically modify your code, by using reflection (of course, it is less safe than the "normal" way of programming, as you turn compile-time errors into run-time errors).
"The Wikipedia article [wikipedia.org] notes that MS was one of the companies that helped draft the standard."
So what?
"The Criticism [wikipedia.org] subsection is also informative about some of the problems with the standard."
The criticism is childish. It boils down to two complains:
1. "ACPI is complex" 2. "hardware does not always completely support ACPI"
Both problems are things with which real software has to deal all the time. Especially the complain no. 2 is silly, because such problems may crop up *anywhere*. The operating system should, within reason, work around incomplete adherence to the standard. You may think that by saying "sorry, you're not 100% compliant, go away", Linux kernel developers are punishing the manufacturer and giving them the incentive to improve compliance, but it's not true. They're punishing the guy who thought Linux is going to run on his box and giving him the incentive (rather, more of it) to stick to OS that works. That's not how you do things when you write an OS which has marginal market share on desktops and notebooks.
"There are also numerous [mixx.com] other [wordpress.com] examples [google.com] of how Microsoft has been quite deliberately poisoning the ACPI well."
They are numerous references to the same case of some guy's clash with Foxconn.
"Slasdot user leoxx posted a comment [slashdot.org] the other day in the Foxconn mobo thread that you might also find elucidating."
I wonder why they haven't thrown in "And I would like to rape a few ten-year old girls, too" for good measure.
"Sugar and other Linux versions on the XO do take longer to boot; but once the suspend and hibernation features are completely working (and the current Update.1 Release Candidate has most of it working)"
How many years will pass until Linux gets suspend and hibernate right?
Obama, considered to be the the last hope for freedom in the USA and all that;-) voted "Yea" while the reviled Clinton voted "Nay". Just goes to show how impredictable politics are when you don't know what was the real purpose behind this or that action.
Hmm. My PhD advisor, a brilliant mathematical physicist, is a devout Catholic.
Just to make things clear, I'm an atheist. But I have the capability to understand how some people may be good engineers/scientists and at the same time believe in God. They don't believe in ghosts or angels visiting them, just in the existence of a Supreme Being which gives them moral guidance. In fact, what I abhorr most in religion is not its irrationality (we're all irrational anyway), but the fact that you have to submit your conscience and moral sense to what a robed priest tells you.
Slashdot Mirror
Munich is nearly 900 years old
Why limit ourselves here? Earth is several billion years old! That's a time frame.
10% of desktops running Linux 6 years after the migration started?? It's an utter failure.
"Except that Apple laptops are junk. None of them have nipples"
As a true geek you probably were unaware of this fact, but most nipples are found on women.
Can one develop addins to Open Office, like for Excel?
"Well, first of all, what has been shown to be wrong is the guesstimate of correlation that was input into the model. G.I.G.O"
The problem is that Gaussian copulas cannot model current market prices for ANY correlation between 0 and 100%. Read about "fat tails" problem.
What matters is that I go to Dixons (UK electronics store), approach a shelf with subnotebooks and see a sign "Linux notebooks will not work with mobile Internet".
Go figure.
And my point is, "useful to make money with" does not correlate to "actually useful", or "net positive impact on humanity".
Oh please, can we bury this hubris-overloaded programmer notion of "my code will save the world"? If the code makes someone money, it's useful for them, period. And debating on the net impact on humanity of this or that will lead us nowhere, there are too many variables involved.
COBOL is still widely used, and COBOL programmers are highly paid. I really wouldn't want to work with COBOL, though.
You are free to like or dislike any language, it doesn't tell me anything about it's usefulness.
As long as I don't have to edit the source, I really don't care much what language a program is written in.
Me too, unless an unlikely situation arises in which I hit a showstopper bug and the time to fix the bug is increased by the problems of the language.
No, it will check that the program is correct. Types are a laughably small part of that.
We both know that it's rare when 100% of the program is being tested. Static typing guarantees you that 100% of function signatures and variable assignments are tested for correct typing. In a dynamic language, you're on your own.
Generally tests do not break in the same way that the program itself breaks,
Sure, the first (or last) thing to do with a broken tests is to remove it ;-)
Second: You're going to need those tests anyway. On a smaller program, you might get away with no tests -- but you probably don't really need the types there, either. On a lager program, it's not enough to know that "public int foo()" returns an integer -- you need to know that it actually does what it's supposed to do.
Look, I *know* what are the benefits of unit tests. I just don't like writing them when I could get it done by the compiler.
Once you know that, why do you care whether it returns an integer?
Because I might be awfully surprised when this integer turns into a float.
And every time I look at it, it seems like the only thing Java has that Ruby doesn't (language-wise) is more restrictions -- static type checking, actually-private values and methods,
Have you ever worked on a large codebase, like 1E5-1E6 lines of code? if you did, you would appreciate the benefits of private values and methods, of having a clean interface and knowing that you can reorganize the inner workings of a class without breaking the code of this guy two floors above you. Without private variables and methods, you risk creating A Giant Ball Of Mud, where everything depends on everything and it gets the point where you are afraid to change *anything* because it might break things in ten places. Been there, done that. It ain't pretty and it ain't pleasant. If your Ruby doesn't allow me to encapsulate my implementation, then I'll never use it for anything more involved than a small script or quick prototype.
"Irrelevant. Lots of companies make money selling antivirus software."
1. You're comparing apples with oranges here. From the software developer's POV, Java is a tool while antivirus program is a product. The fact that I make lots of money selling some product does not have to correlate with the product's usefulness, but if I make lots of money using a tool, it means that this tool is indeed USEFUL TO MAKE MONEY WITH. And if Java is useful to make money with, that's splendid for me.
2. Companies also make money by USING software written in Java. I seldom hear someone claiming that "hey, I've made money on using Norton Antivirus!". Antivirus software helps you not to lose money rather than make them.
"That's what we have unit tests for."
If I unit test module A and module B, but then module A feeds module B with different type of data then B expects, then unit testing these modules won't help you much (since you will take care in your tests to supply the correct types). Oh, you say that I should unit test module A and module B together, to make sure that cooperate correctly -- but what happened to "unit" in "unit tests"? We will end up with one giant blob of a test, which checks whether the types are correct.
"And honestly, type checking is just a subset of unit testing."
I love such impractical analogies. You miss out a crucial point that with static typing, you get these "unit tests" FOR FREE, generated by the compiler without ANY EFFORT ON YOUR PART and without creating any testing code to debug and maintain.
"What's more, it's a lot easier to implement static type checking on top of a more dynamic language than the other way around."
Maybe.
"Those 20 minutes is probably considerably less time than it would take for my first attempt in Java"
Should it tell me something about Java, or about you?
BTW, I've been coding quite a lot in Matlab, which is also dynamic, and I've never need "eval" there. Anonymous functions, on the other hand... I would like to see them in Java. Them and the const correctness.
Lots of companies make lots of money writing software in Java or using software written in this language. I have developed software in C, in Java and in Matlab (which is a dynamic language of sorts). Each of them has its uses. Saying that dynamic languages are always better than static is just stupid. Dynamic languages are faster to develop in because they place less restrictions on the programmer. For the same reason, they are also less safe, especially when the codebase becomes more complicated. I've had weird errors in Matlab prop up because instead of passing a number, I've passed a string to my procedure. In a statically typed language, it wouldn't happen. Anyway, Java does allow you to dynamically modify your code, by using reflection (of course, it is less safe than the "normal" way of programming, as you turn compile-time errors into run-time errors).
http://mjg59.livejournal.com/94998.html
"The Wikipedia article [wikipedia.org] notes that MS was one of the companies that helped draft the standard."
So what?
"The Criticism [wikipedia.org] subsection is also informative about some of the problems with the standard."
The criticism is childish. It boils down to two complains:
1. "ACPI is complex"
2. "hardware does not always completely support ACPI"
Both problems are things with which real software has to deal all the time. Especially the complain no. 2 is silly, because such problems may crop up *anywhere*. The operating system should, within reason, work around incomplete adherence to the standard. You may think that by saying "sorry, you're not 100% compliant, go away", Linux kernel developers are punishing the manufacturer and giving them the incentive to improve compliance, but it's not true. They're punishing the guy who thought Linux is going to run on his box and giving him the incentive (rather, more of it) to stick to OS that works. That's not how you do things when you write an OS which has marginal market share on desktops and notebooks.
"There are also numerous [mixx.com] other [wordpress.com] examples [google.com] of how Microsoft has been quite deliberately poisoning the ACPI well."
They are numerous references to the same case of some guy's clash with Foxconn.
"Slasdot user leoxx posted a comment [slashdot.org] the other day in the Foxconn mobo thread that you might also find elucidating."
I wonder why they haven't thrown in "And I would like to rape a few ten-year old girls, too" for good measure.
Cool. My Windows laptop has 100% support.
"Sugar and other Linux versions on the XO do take longer to boot; but once the suspend and hibernation features are completely working (and the current Update.1 Release Candidate has most of it working)"
How many years will pass until Linux gets suspend and hibernate right?
I meant "support" as in "are they as easy to get as in C".
Obama, considered to be the the last hope for freedom in the USA and all that ;-) voted "Yea" while the reviled Clinton voted "Nay". Just goes to show how impredictable politics are when you don't know what was the real purpose behind this or that action.
Oh oh, we're snobbish, aren't we?
For example, does it support buffer overruns? Double free's?
A way will be found to make it illegal.
Hmm. My PhD advisor, a brilliant mathematical physicist, is a devout Catholic.
Just to make things clear, I'm an atheist. But I have the capability to understand how some people may be good engineers/scientists and at the same time believe in God. They don't believe in ghosts or angels visiting them, just in the existence of a Supreme Being which gives them moral guidance. In fact, what I abhorr most in religion is not its irrationality (we're all irrational anyway), but the fact that you have to submit your conscience and moral sense to what a robed priest tells you.
"One, their love for money trumps their customers interest."
And how is it different from any other company?
But it's already a fork!
"John McCain's got it all wrong on both the torture"
What has he got wrong on torture? You mean he's not strongly enough opposed to it?
"and then sends them home to their toxic filled rivers and lakes."
You're talking about Chinese people or Chinese frogs?
OK, never mind. What's your point?
I don't think so. As far as I know, it's not that the Chinese hate their government, overall.