News for hackers. Stuff that matters.
Re:Ask HN: Guido van Rossum's comment about go and scratch by dkersten on Tuesday November 17, 2009 @09:44AM (#946476)
Theres that too. To be honest, I'm a little disappointed. I expected them to come up with something amazingly awesome... Ah, we'll see how things pan out. But if it gains any traction, I still think its because of Google hype.
Re:Ask HN: Guido van Rossum's comment about go and scratch by eru on Tuesday November 17, 2009 @10:00AM (#946508)
Which might not be too bad. Even it Go lags behind the cutting edge, it may still be an improvement on industry standards. Java was also heavily pushed by Sun --- and while I don't like it, I guess it's memory management is still an improvement over C++.
Re:Ask HN: Guido van Rossum's comment about go and scratch by jrockway on Tuesday November 17, 2009 @11:42AM (#946699)
But Go is no improvement over Java. (And you have to work pretty hard to not improve on Java...)
Re:Ask HN: Guido van Rossum's comment about go and scratch by dkersten on Wednesday November 18, 2009 @06:47AM (#948148)
Java isn't an improvement over C++ either though.. oops circular logic!
Re:Ask HN: Guido van Rossum's comment about go and scratch by eru on Wednesday November 18, 2009 @09:22AM (#948299)
Citation neeed.
Re:Ask HN: Guido van Rossum's comment about go and scratch by dkersten on Wednesday November 18, 2009 @02:46PM (#948956)
I'd say its about on par. Which is better depends on the problem your trying to solve. Sure, Java adds some nice features, like garbage collection and potential compile-once-run-elsewhere and a bunch of other things, but C++ has its advantages either. Heres a small few:

C++ can have much better startup time. While I no longer buy into the speed thing (the JVM has become quite fast), I still find Java mesurably slow while starting up. So much, in fact, that projects like NailGun were started to improve things.

I hear Java's floating point hurts everyone everywhere: http://www.eecs.berkeley.edu/~wkahan/JAVAhurt.pdf

C++ supports unsigned arithmetic, Java doesn't.

Javas generics use type erasure.

Checked exceptions in Java are a bad idea.

C++ allows tight OS integration.

And so on. While I'll be the first to admit that C++ is far from perfect, I also don't buy that Java improves it. One is better sometimes, the other is better other times, both are terrible in other situations again. I prefer C++ in more cases than I prefer Java, hence my comment.

But.. maybe I missed the point of the comment I replied to, since memory management in Java might really be an improvement over C++.

Re:Ask HN: Guido van Rossum's comment about go and scratch by eru on Wednesday November 18, 2009 @04:28PM (#949156)
Thanks for the clarification. I guarded my original past by only speaking of memory management. I have programmed in Java and C++, but not enough to say which one is better. I tend to avoid both in favor of Haskell, Python, Ocaml, Scheme and the like. D also looks promising.
Re:Ask HN: Guido van Rossum's comment about go and scratch by dkersten on Thursday November 19, 2009 @07:14AM (#950105)
You're welcome. I use both in work, though a lot more Java than C++, but I agree, when I'm given a choice (or working on my own projects) I, too, prefer to avoid them. I've used a lot of Python in the past and, for the past two months, have been using Clojure. Much much nicer to work in than Java or C++!
Re:Ask HN: Guido van Rossum's comment about go and scratch by nearestneighbor on Tuesday November 17, 2009 @12:21PM (#946791)
I expected them to come up with something amazingly awesome

Why?! C and UNIX are from the worse-is-better camp.

Re:Ask HN: Guido van Rossum's comment about go and scratch by dkersten on Wednesday November 18, 2009 @06:46AM (#948146)
I actually agree with that statement. I dunno, I guess its because I used (and loved) C++ exclusively for a few years and I was hoping maybe they could make something similar - just better. But then when Go came out, reality hit :-)