- no operator overloading because some people abuse it, too bad if code is twice as long and hard to read, we must protect people from themselves.
I think the argument generally goes that operator overloading hides the complexity of an operation. I don't really buy this argument, and I think Java probably should have it. Don't get used to me agreeing with you, it won't last long...
- we'll take care of the memory management because stupid programmers can't handle it.
Well, maybe they are right =) How often do you run a program which leaks memory or crashes (*ahem* netscape)? Besides, I hate to bust up your precious belief system, but garbage collection is a very good thing.
Good GC is faster than hand allocation/deallocation. The main reason everyone thinks it is slow is because some types of GC cause the whole program to pause up. This is not true in general. And the reason Java is slow is not because of GC, it is because Java has all kinds of safety checks on every operation you do.
Not having GC breaks the safety you gain by getting rid of pointers in the first place. If the programmer is required to free memory, he/she may end up freeing something which still has references pointing to it.
You can't prove anything useful about non-trivial programs.
Wow, this is just plain wrong. For one, people who design really important systems (ie, space shuttle systems) often have to go through and prove things about their programs. Secondly, Java actually *does* prove that your program does not break type safety (that is what we are really talking about here). A few elements of Java actually prevent it from being able to do this correctly (eg, allowing references to be null), but it does for the most part. Check out ML if you are interested. It is a strongly typed language, and you can't run a program if its type checker cannot prove that it is type safe.
To me this summarises the entire java/C++ debate. The whole philosophy behind java is you can't trust the programmer,
Even if this is what you believe, it may not be a bad thing. I don't really trust the people writing much of the software I use, because they have proven again and again to be unworthy of my trust (with a few notable exceptions, of course).
Of course, this is only part of the story. Languages with real types (C++ types aren't worthy of the name) and a real type soundness theorem allow the type checker to prove that data actually is what you say it is. This in the end benefits the programmer by making designing and debugging faster and easier.
Software engineering is in a pretty sad state today. Programs crash all the time, or just plain don't work, and we accept that as standard. This is just wrong. There is no reason for people to use languages that make it harder for them to write good programs, which C/C++ often turn out to be. Now, don't get me wrong, I think both C and C++ have a place, and some people are able to use them effectively. I don't think they deserve the great respect and widespread use they currently have however.
2d "space" is a plane with no thickness. So it's not space.
It may not be space as we typically think about it (eg, outer space or something), but a plane can most definitely be a space, and the xy plane is a space in every mathematical way (ie, vector space, metric space, etc) that the xyz 3d space is, simply with one less dimension to worry about.
Of course, the way tech stock IPO's have been going recently, it is probably not a bad bet.
Speaking of which, who wants to join in a business venture with me? We will buy a server, set it up, call the company networks.com, IPO, sell to a big company, and make a FORTUNE! Oh wait, I forgot, the company would have to lose money or it wouldn't be worth anything. Oh well.
2) Patents - most especially software patents - have nothing to do with the free market; they are artifical inventions of the state.
On the other hand, patents, copywrights, and the like provide a service by preserving the ideas and technology behind the inventions and writings. This is actually, AFAIK, one of the reasons these things were developed, so that inventions, books, etc would not be lost forever when they stopped being sold, or something weird happened.
Of course, there are a lot of people who make a living by developing new things, and I think they have a right to profit from the sale of those things.
I think that in our zeal for free software, we easily forget that legal protections of IP do serve a purpose.
Of course, I think this whole GIF thing is total BS, but that is, as you say, because of the "bait and switch," not because they are trying to charge for their patent.
Slashdot Mirror
A few comments...
I think the argument generally goes that operator overloading hides the complexity of an operation. I don't really buy this argument, and I think Java probably should have it. Don't get used to me agreeing with you, it won't last long...
Well, maybe they are right =) How often do you run a program which leaks memory or crashes (*ahem* netscape)? Besides, I hate to bust up your precious belief system, but garbage collection is a very good thing.
Wow, this is just plain wrong. For one, people who design really important systems (ie, space shuttle systems) often have to go through and prove things about their programs. Secondly, Java actually *does* prove that your program does not break type safety (that is what we are really talking about here). A few elements of Java actually prevent it from being able to do this correctly (eg, allowing references to be null), but it does for the most part. Check out ML if you are interested. It is a strongly typed language, and you can't run a program if its type checker cannot prove that it is type safe.
To me this summarises the entire java/C++ debate. The whole philosophy behind java is you can't trust the programmer,Even if this is what you believe, it may not be a bad thing. I don't really trust the people writing much of the software I use, because they have proven again and again to be unworthy of my trust (with a few notable exceptions, of course).
Of course, this is only part of the story. Languages with real types (C++ types aren't worthy of the name) and a real type soundness theorem allow the type checker to prove that data actually is what you say it is. This in the end benefits the programmer by making designing and debugging faster and easier.
Software engineering is in a pretty sad state today. Programs crash all the time, or just plain don't work, and we accept that as standard. This is just wrong. There is no reason for people to use languages that make it harder for them to write good programs, which C/C++ often turn out to be. Now, don't get me wrong, I think both C and C++ have a place, and some people are able to use them effectively. I don't think they deserve the great respect and widespread use they currently have however.
It may not be space as we typically think about it (eg, outer space or something), but a plane can most definitely be a space, and the xy plane is a space in every mathematical way (ie, vector space, metric space, etc) that the xyz 3d space is, simply with one less dimension to worry about.
Of course, the way tech stock IPO's have been going recently, it is probably not a bad bet.
Speaking of which, who wants to join in a business venture with me? We will buy a server, set it up, call the company networks.com, IPO, sell to a big company, and make a FORTUNE! Oh wait, I forgot, the company would have to lose money or it wouldn't be worth anything. Oh well.
On the other hand, patents, copywrights, and the like provide a service by preserving the ideas and technology behind the inventions and writings. This is actually, AFAIK, one of the reasons these things were developed, so that inventions, books, etc would not be lost forever when they stopped being sold, or something weird happened.
Of course, there are a lot of people who make a living by developing new things, and I think they have a right to profit from the sale of those things.
I think that in our zeal for free software, we easily forget that legal protections of IP do serve a purpose.
Of course, I think this whole GIF thing is total BS, but that is, as you say, because of the "bait and switch," not because they are trying to charge for their patent.