by randi » Thu Apr 02, 2015 9:01 pm
Here:
FluidSynth in 64-bit glory
Compiled with the default VS2013 toolchain, so it only works on Vista and up. Turns out I should have paid attention to the warnings instead of assuming they were benign (since the majority were about ptrdiff_t -> int truncation). It wasn't using config.h, so <math.h> was not getting included, which meant all the fp math routines were using the default prototypes that returned ints. Disaster ensued.
Building in 64-bit also revealed a minor problem I had with the original 32-bit build, so I redid that. Anybody using it might want to redownload it.
An interesting project to try sometime might be to "fix" it so it can use native Windows API calls instead of glib so that it can be built more easily for Windows systems.
Here: [url=http://zdoom.org/files/libfluidsynth64.7z]FluidSynth in 64-bit glory[/url]
Compiled with the default VS2013 toolchain, so it only works on Vista and up. Turns out I should have paid attention to the warnings instead of assuming they were benign (since the majority were about ptrdiff_t -> int truncation). It wasn't using config.h, so <math.h> was not getting included, which meant all the fp math routines were using the default prototypes that returned ints. Disaster ensued.
Building in 64-bit also revealed a minor problem I had with the original 32-bit build, so I redid that. Anybody using it might want to redownload it.
An interesting project to try sometime might be to "fix" it so it can use native Windows API calls instead of glib so that it can be built more easily for Windows systems.