Introduce Unit Literals

[ryanmrichard]

The resolution of #113 was that we should:

• specify units at user-facing APIs to avoid ambiguity, and
• internally store quantities in atomic units

This is going to be API breaking and should be done before a 1.0 release.

I foresee the following tasks:

• Unit Literal Design #286
• Decide on unit conventions #287
• ensure that the design decisions make it into the documentation for future generations
• add units to classes in Chemist
  • my vote is to do this by operator overloading (i.e., next to the options which don't take units) to avoid breaking downstream repos
• update downstream repos to use units where necessary (this is probably a number of tasks in and of itself)
• remove overloads which don't take units

[twindus]

Since it will be API breaking, we should plan the timing of the check in carefully.

[wadejong]

https://github.com/nholthaus/units#unit-literals

https://www.boost.org/doc/libs/1_78_0/doc/html/boost_units.html

https://www.researchgate.net/publication/328507940_The_next_700_unit_of_measurement_checkers/figures?lo=1

[twindus]

On the Python side, I know people have been using Pint for unit conversion. We probably don't want to go this way, but thought I would throw it out there.

[ryanmrichard]

@wadejong and @twindus I started a design issue, #286, and copied these links there.