Nothing more, nothing less. C++0x does not change that fact.
Having a memory model is a good thing though:)
As the document shows from page 3 to 7, C++ values efficiency a lot. If that's a requirement C++ is great. Criticism that C++ is not Python/Java totally missed the point!
1) No matter how talented you are, to become an expert will take at least 10 years. 2) Given two people practicing for 10 years, the one with more talent will have better performance.
I don't see how people can generalize further from the above.
It's tough to be a prey, but hey, most animals are at the same time both preys and predators.
Whether free software create or destroy wealth depends a lot on the circumstances. On one hand, without copyrights/patent it's possible that much fewer software would be written. On the other hand, it's not very productive if everyone reinvents the wheel all the time and all companies have to pay a million bucks for a C compiler.
My advice: get back to work! Slashdot DOES destroy wealth for sure:)
"Then there is the Collections API itself which upon first glance seems like it was written by amateurs who have never had to write any performance critical code in their lives."
I think the situation has improved since Java 5.0 thanks to Doug Lea.
"Ultimately, even the computer and operating system you're running on are probably out of your control, yet your program will rely on them to do its job, too."
Sure, but if you know your program is correct then you know where is the problem. And in embedded real-time area, even the OS and the computer can be proved correct to a conservative but precise spec.
Of course, programs proved correct could still contain bugs. But I think we've sort of lost the patience and dedication of programmers before us like Knuth, Hoare and Wirth, which one can interprest as the decline, if not death, of computer science.
I just don't think you can prove MIT curriculum is the best:)
It's a bit tiresome. There are plenty of awesome schools that don't use the scheme curriculum -- some uses ML, some uses smalltalk, while other uses more traditional programming language like C or Modula-2/3. Scheme is not special.
Yes, I like to trash "Java schools" too but that's (mostly:) for jokes only. One good thing about Java is that it can teach students concurrent programming.
And there is nothing wrong with C! Well, one should rather use Pike:)
Just like to say God can't be a murderer by definition. You can disagree with God's decisions and rules, but there is no reason that you should be alive.
> performant
Yes! We need code to run on 216 cores computer!
>Widely adopted
It works on VMS. Enough said.
>eminently portable
It's mostly true :)
I assume Windows Server would work better.
No? :)
Mainframe jobs are being outsourced.
Linux jobs are being outsourced.
Results are the same :)
Nothing more, nothing less. C++0x does not change that fact.
Having a memory model is a good thing though :)
As the document shows from page 3 to 7, C++ values efficiency a lot. If that's a requirement C++ is great. Criticism that C++ is not Python/Java totally missed the point!
http://www.research.att.com/~bs/abstraction-and-machine.pdf
Or open Knuth, spend 30 years learning all included, and then reread the next edition...
I think it's fairly obvious that:
1) No matter how talented you are, to become an expert will take at least 10 years.
2) Given two people practicing for 10 years, the one with more talent will have better performance.
I don't see how people can generalize further from the above.
How many F5 keys do you press a day?
1) 10 (Normal)
2) 100 (Well...)
3) 1000 (What?)
4) 10000 (Calling 911...)
I think you are very wrong.
Whether there is a any materials beyond this world is either unknown, or 50/50.
The chance of the second is almost zero. (That's my guess, unless you have more statistics.)
And not to forget Self, the simple yet powerful language from Sun.
It's tough to be a prey, but hey, most animals are at the same time both preys and predators.
:)
Whether free software create or destroy wealth depends a lot on the circumstances. On one hand, without copyrights/patent it's possible that much fewer software would be written. On the other hand, it's not very productive if everyone reinvents the wheel all the time and all companies have to pay a million bucks for a C compiler.
My advice: get back to work! Slashdot DOES destroy wealth for sure
Well, you did compare PHP with Visual Basic.
I don't think World Bank loan money. IMF does.
But I am very likely to be wrong! Am I?
While I am not a fan of PHP, you need to argue what else you use instead which is so much better than PHP.
Perl? Python?
Well, you need to read the MIT Algorithm books :)
Or search linear median on google!
"Calculus is analysis for idiots."
:)
Such as Newton, Euler and Gauss
So?
:)
Even PhDs could be wrong about the Monty Python Problem
It isn't TOO bad.
At the very least, I don't think C++ as described in the following document is too complex.
http://www.research.att.com/~bs/hopl2.pdf
Sorry, I am too lazy, but you can read the exerpts from here
- 0471798142,descCd-tableOfContents.html
http://ca.wiley.com/WileyCDA/WileyTitle/productCd
But I am not that confident just by looking at the Table of Content.
Kind of agree.
:)
I think calculus is invaluable too, but just not the way it is taught in first year College maths.
Hey, newton method is really like a simple pascal program
"Then there is the Collections API itself which upon first glance seems like it was written by amateurs who have never had to write any performance critical code in their lives."
i ndex.html
I think the situation has improved since Java 5.0 thanks to Doug Lea.
http://gee.cs.oswego.edu/dl/concurrency-interest/
Yes, some Sun engineers are amateur. But I don't think you can claim that for Doug Lea!
"Ultimately, even the computer and operating system you're running on are probably out of your control, yet your program will rely on them to do its job, too."
Sure, but if you know your program is correct then you know where is the problem. And in embedded real-time area, even the OS and the computer can be proved correct to a conservative but precise spec.
Of course, programs proved correct could still contain bugs. But I think we've sort of lost the patience and dedication of programmers before us like Knuth, Hoare and Wirth, which one can interprest as the decline, if not death, of computer science.
I just don't think you can prove MIT curriculum is the best :)
:) for jokes only. One good thing about Java is that it can teach students concurrent programming.
:)
It's a bit tiresome. There are plenty of awesome schools that don't use the scheme curriculum -- some uses ML, some uses smalltalk, while other uses more traditional programming language like C or Modula-2/3. Scheme is not special.
Yes, I like to trash "Java schools" too but that's (mostly
And there is nothing wrong with C! Well, one should rather use Pike
In this special case, yes, you can prove the interpreter's correct.
It's just like proofs in maths. Nothing special.
There are four relevant outcomes -- either there is no God, you suffer, I suffer or both suffer. Which you think is more likely?
Just like to say God can't be a murderer by definition. You can disagree with God's decisions and rules, but there is no reason that you should be alive.