Re: Reason for -fno-strict-aliasing?

[cblake]

One other gotcha in this "test" to see how close we already are is that it 
seems nim catches the `stderr` output of the gcc invocations. In that sense the 
nim `gcc` is as if it were `gcc -w` anyway. If you capture the commands from 
`--verbosity:2` and re-execute them from a `/bin/sh` script, you do see any gcc 
warnings, though. I still don't see any warnings related to aliasing.