MAXFLOAT should be rewritten as Infinity

[trevordblack]

MAXFLOAT represents the largest possible floating point (32 or 64 bit) number.
This will almost never present a problem in practice, but the correct formulation here is with floating point infinity.

[hollasch]

Agreed.

[ronaldfw]

Fixed on future branch. Will keep this issue open until this is merged with master.

[hollasch]

Merged into the development branch.