f128 requires an additional third party library on many platforms

[brson]

This feature was recently added and has support in the standard library, but has few tests, and is suspected not to work at all on platforms other than Linux and Windows.

[alexchandel]

Apparently it was removed in 7a93bee.

And that's too bad. Native support for quad precision floats made Rust special to me, like Fortran's real(real128) or Python's numpy.float128. Even if there aren't native instructions for 128 bit operations, there could be in the near future, and I like the sound of i128, u128, and f128.

If these features aren't anticipated then Rust might have to staple them on later, like C to C++.