On 4/17/2016 7:05 AM, Niall Douglas wrote:
On 16 Apr 2016 at 7:58, degski wrote:
On 16 April 2016 at 01:15, Edward Diener
wrote: I have posted bug 27380 to the llvm/clang bug tracker...
Thanks for not giving up (also @ Niall D.)!
Thanks for saying so, but we could do a lot better.
I flicked through the clang source code and as far as I can tell from a ten minute inspection, getting C99 preprocessor behaviour on MSVC ABI targets is as easy as flipping a boolean on. Most of the other MSVC preprocessor special behaviours only turn on with -fms-compatibility=on, and we no longer need that turned on to use Windows or indeed the MSVC CRT and STL. A source code patch ought to be mostly figuring out how to add a command line argument to control the C99 processor boolean, a few unit test cases to ensure rigour and lack of regression, and that's about it.
The real effort behind such a patch would be navigating the internal politics on getting the patch into trunk. I have no need in my own code for a C99 preprocessor, hence I am not willing to invest that effort. But others with a more pressing need may wish to do so.
Why not bring up the issue on the clang developers mailing list ? Clang source code is not an area which Boost developers can affect in this mailing list, but in the clang developers mailing list you might be able to persuade a clang developer to listen to your suggestion(s).