Well, I bet it will create a whole new level of mess in the language. Though I guess that is consistent with previous design principles of C++, so that makes it OK :).
Oh, man, I'm in love with type inference. Although the way to was added to C++ is rather ad hoc. When you don't include implicit function polymorphism with it, it just becomes a small time saver. Admittedly, I don't know how they would have included IFP into the existing type system.
But combine type inference with IFP, and you have a language that combines the versatility of dynamic typing with the safety of static typing. Just take a look at OCaml, StandardML, or Haskell.
@ Luc Lieber: Of course I knew auto comes from C and only recently its purpose was perverted. But then you'd argue that it didn't have any real purpose to begin with. Which would contradict your previous post. Checkmate! (Don't take this post too seriously.)
Back on topic, I'm confused as to why this thread has so few replies... on a C++ forum. I was expecting to see Marshall Clines and Yossi Kreinins arguing with great ardor over the good and the bad, yet insignificant, new features or lack thereof.
I cannot wait for the standard regular expression library... that's going to be useful on many levels (because I'll get to myself two command line parameters when compiling programs using regular expressions).
Yes, I'm looking forward to C\+\+1\d. What, too cynical?
I'm excited about anything new and shiny.. but mostly looking forward to uniform initialization, move semantics, ranged for loops, and strongly typed enums.
Regular expressions, lambdas, auto and initializer lists. Maybe threads too, though I gotta admit that my knowledge on threads is right now rather... limited (never got in a situation where it made sense to use threads before).
Does anyone know if there are any changes from the Working Draft, Standard for Programming
Language C++ (n3424)[1] and the final published version?
___________________
[1] http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2011/n3242.pdf
Wikipedia says the changes, if any, are minor. You're probably safe using the working draft. I have the working draft saved on my comp and I'm going to use it until the final version is free.
But what do you need the standard document for? I guess it is mostly for compiler / tool creators, not for programmers. If programmers needed to read language specs to write good code, it would be a bad sign (meaning that language is overcomplicated).