Jacob Williams

* Would there be any harm for allowing it on all platforms? * I don't think `CONDA_BLD_REGEX_CHUNK_SIZE` is a good name for this variable. It's really about the max command...

I'm not a CMake expert so I'm not sure what the proper solution is. If this works, I can merge it in. See also: #31

See this: https://gitlab.com/EmergentSpaceTechnologies/OpenFramesInterface/-/commit/e53f2dabb04bb069d69b85415fd79ef748d217b7 Can we do a similar thing when using from Python?

I did the same thing! https://github.com/jacobwilliams/PCHIP

Test case (same result from refactored version and the one in scipy): ``` n = 2 mp = 5 x = 1.2345600000000001 sj = 0.76464742242968342 0.35211808607480077 9.1005002907604604E-002 dj = -0.35211808607480077...

Here's another one (same result for both): ``` n = 4 mp = 6 x = 10.234560000000000 sj = -7.0754079663227554E-002 6.0471793911910328E-002 8.8479842347283441E-002 -1.7245782071985880E-002 -0.10027521747859354 dj = -6.0471793911910390E-002 -8.2571254785931469E-002 3.4536186807955657E-002 9.5220056708032075E-002...

Interestingly, this case also failed with [json-fortran](https://github.com/jacobwilliams/json-fortran/issues/388) but for a different reason. I just fixed it in that library.

See here for a discussion: https://fortran-lang.discourse.group/t/fun-with-40-year-old-code/3726

* Octave code here: https://github.com/gnu-octave/octave/blob/default/scripts/general/quadgk.m * Julia code here: https://github.com/JuliaMath/QuadGK.jl/blob/master/src/adapt.jl

Yep a comment would be fine. We could merge quadpack into this library for a unified interface? Or just keep it separate?