Decide whether to promote all ints to floats in binops

[KhaledAtef]

Decide whether to promote all ints to floats in binops

[KhaledAtef]

In the case of adding floats to ints, did we decide on promoting the ints to floats? Originally, we said we'll just take whatever comes first, but I couldn't remember what we decided on.

[six5532one]

I thought we decided to promote ints to floats (for binops) because when we got together last time, we tried out the llvm library's build_add function with an int and a "float" by casting the int to const_double, then using const_uitofp

[DavidWatkins]

Ah gotcha. But if we have the addition of two integers, it stays type int, correct?

[six5532one]

In the current implementation, yup indeed:
https://github.com/DavidWatkins/Dice/blob/emily/Compiler/analyzer.ml#L115