Correct test for cpp_dec_float division.

[jzmaddock]

@ckormanyos , @mborland : this should fix the current quaternion / cpp_dec_float failures.

[ckormanyos]

this should fix the current quaternion / cpp_dec_float failures.

Also note that this recent Multiprecision PR also fixes these tests without disabling them for cpp_dec_float. That Multiprecision PR, however, only partially addresses the missing rounding issue in cpp_dec_float.

We could consider reviving Multiprecision's PR, synchronizing develop's new standalone changes into it and going forward with that PR. It did pass CI. On the other hand, I did list several reservations that I have about only partially addressing the rounding issue. These are in my final post in that PR.

I could be happy with either: go forward with Math's PR here that disables the tests, or move forward on Multiprecision's PR.

See also the relevant comments

Cc: @jzmaddock and @mborland

[jzmaddock]

@ckormanyos I'm easy either way, for now I really just want CI clean again so we can better see where we are ;)

[ckormanyos]

@ckormanyos I'm easy either way, for now I really just want CI clean again so we can better see where we are

Indeed @jzmaddock. Let's go that way since there has really been a lot going on.

Maybe we can pick up rounding with a more strong effort in 1.79 or 1.80, as I could not fully figure out how to do it by myself. So for the moment, I'm just happy if everything works. I will let the multiprecision PR ripen, maybe synchronize with develop so as not to get too far behind.