Is C really faster than c++??

[PanGalactic]

There is no reason that

template <typename T>
float operator^(float, T) // where T is appropriately SFINAE'd for numeric types 

and

template <typename T>
double operator^(double, T)

and

template <typename T>
long double operator^(long double, T)

should not mean exponentiation. Why the heck would anyone XOR a float? There is precedence for using the carat symbol for exponentiation in other languages.

[rocketboy9000]

Right, but what about when somebody does:
a=2.0^5-1;
And expects it to work, i.e. get 31? You'd actually get 16. The problem is with the precedence, which was my point about bit-shift.

[PanGalactic]

When in doubt, parenthesize. They are free. I almost always use parens when there are more than two terms. It makes my intentions explicit.

[rocketboy9000]

Ya, well, this would just be another stumbling block for newbies. Much like the fact that string s="Hi, ";s+="John!"; works, but string s="Hi, "+"John!"; does not. Many ugly things are produced because C++ attempts to be backward compatible with C. The precedence issues for << is even worse:
cout << hex << setw(2) << setfill('0') << bitfield&0xff << dec;
Its meaning is obvious, but it fails miserably. If it were up to me, I would use the += operator for output, and the -= operator for input.

[stereoMatching]

iit2009154, I forgot to type the title of Item23 of more effective c++

Consider alternative libraries

[moorecm]

Its meaning is obvious, but it fails miserably. If it were up to me, I would use the += operator for output, and the -= operator for input.

And that would solve the problem of precedence?

[rocketboy9000]

Yes. Assignment operators have the highest precedence. It also is a better metaphor. When you print to a file, you *add* the entry into it.

[Kyon]

Well, not really, the scope resolution operator (::) has the greatest precedence. After that is the assignment operator.

[Athar]

The assignment operators actually have the lowest precedence right after the comma operator.

[moorecm]

There has to be a reason. Let's consider:

cout << a += b;
// vs.
cout += a += b;

Is this possible? Are they evaluated the same?

Are there any other scenarios that stand out?

[guestgulkan]

Rocketboy9000 wrote:

Assignment operators have the highest precedence

?????
Care to explain??

[Albatross]

Enough debate, boys; Athar's right. Hopefully the following link will end the debate.
http://www.cppreference.com/wiki//language/operator_precedence?redirect=1

-Albatross

[rocketboy9000]

Oh, right, I forgot which end is "highest". Maybe "loosest" would be better.
anyway my point is that
cout += x&0xff;
would be evaluated in the obvious manner, rather than trying to and 0xff with a reference to cout.