On 2018-06-04 4:35 AM, Ole Streicher wrote:
As mentioned before, float128 is not supported on hppa-linux. I get the following with longdouble:https://bugs.debian.org/898776
>>> numpy.arange(3, dtype=numpy.longdouble) array([0., 1., 2.], dtype=float64)
So, it's not the same bug. Regards, Dave -- John David Anglin dave.ang...@bell.net
