On 06/18/2012 04:46 AM, Jonathan Wakely wrote:
The problems arise when user code that uses the "inline-only" code is linked to other user-code that has a different definition of that inline-only code.
What problems arise then? As long as the interface of the user code is the same, it doesn't matter which definition of the inline-only code is used.
On 06/18/2012 06:14 AM, Gabriel Dos Reis wrote:
A related question is whether for GCC-4.8 we should still continue to claim that C++11 is only experimental? That is at which point are
My hope is to nail down the ABI enough for 4.8 that we can drop that claim. Jason
