Hello William!

Aug 02 01:43 03, William McBrine wrote to Bo Simonsen:

 WM>>  for (int x = 0; x < 100; x++)

 BS>> That's not ANSI C/C++, it should be:
 BS>> int x;
 BS>> for(x=0; x < 100; x++)

 WM> It's not ANSI C, but it _is_ standard C++. Since this is the C++ 
 WM> echo, I thought that would be OK. :-) Admittedly, the rest of the message 
 WM> would apply just as well to C as to C++, if one replaced
"new" with 
 WM> malloc(). I almost referred Neil to the K & R book, which devotes a 
 WM> chapter to pointers and arrays.

Yes it's standard C++ :-) But the Kerninghan and Richie book is only about ANSI C ;-)

 WM> You're probably thinking of the change in scope specified in ANSI 
 WM> C++, where the variable now goes out of scope at the end of the loop. But 
 WM> it's still valid _within_ the loop, as I used it.

I guess 'i' would be aviable after the loop?

 BS>> I know borland c++ compile it, but gcc/g++ doesn't.

 WM> Yes, it does. (More precisely: g++ does; gcc doesn't. That is to say, 
 WM> it works in C++ mode, and not in C mode.)

Okey.. :-) 

Regards,
Bo

--- Msged/LNX 6.1.2