It is still a long way to the next stable release (to replace m87). Probably m14x or even m15x.

• M87 was 27 skipped, 8 xfailed. M134 is likely to be similar to m133, 102 skipped, 18 xfailed. (120 not passed) . Need to get the skipped figure down to 80 or 70 at least. [1]

• would be nice to fix or address to some degree C++ type names appear in the docstrings #262 , regarding documentation. It is a one-off massive code re-org.

[1] v116b2 was 124 skipped, 15 xfailed, 16 removed. (155 not passed) . Some of the removed tests were parametric (parameters no longer works), some of the non-parametric ones were added back later, fixed, xfailed, or skipped. It took 17 months to come down from 155 to 120.

My earliest test record of a buildable m116 was 601 not passed, or only 73%, or more than 1/4 broken.