[EXPERIMENTAL] Use DynamicQuantities.jl as default units backend

[icweaver]

Experimenting with how this might look (far future, not committing to anything now)

What's changed

• Set DQ as default units backend
  • Uses symbolic units by default for convenient unit display.
  • Future: Can switch this to regular units if this becomes a performance bottleneck. Trade-off: prints units in base SI
• Moved Unitful support to a minimal-maintenance extension
  • Can still be accessed through the usual <cosmology func>(u::UnitLike, args...; kws...) interface
  • Reduces programmatically defined functions to just what is in the public API.

Example usage: https://learn.juliaastro.org/tutorials/cosmo-plots/

Larger plan

• [SPECULATIVE] Replacing Unitful.jl with DQ.jl system-wide fireplace#3

To-do

• Unitful.jl extension
• tests
• docs

[cgarling]

Do you like returning basic numbers by default here? My 2 cents is that I would prefer us to pick a default units package (DQ if we decide to switch) and return those quantities by default. We can have a Unitful extension so the function signatures that take types can also take Unitful types, but I don't really like returning basic numbers by default as it forces you to check the docstring to see what the units on the basic numbers are.

[icweaver]

I don't. Agreed. I just did unitless returns for now for the interim to experiment with transitioning to DQ by default long term. Growing pains =]

Edit: Since this is a speculative far future draft PR, I'll go ahead and rip off the band-aid

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 100.00%. Comparing base (f7396bf) to head (8aa02fa).

Additional details and impacted files

@@            Coverage Diff            @@
##              main       JuliaAstro/JuliaAstro.github.io#62   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files            1         2    +1     
  Lines           74        78    +4     
=========================================
+ Hits            74        78    +4     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.