Add docs fixes

[lukem12345]

Close #127

[codecov]

Codecov Report

Patch and project coverage have no change.

Comparison is base (d2b2a79) 88.40% compared to head (7f03237) 88.40%.

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #128   +/-   ##
=======================================
  Coverage   88.40%   88.40%           
=======================================
  Files           9        9           
  Lines        1009     1009           
=======================================
  Hits          892      892           
  Misses        117      117           

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

[jpfairbanks]

It looks like this closes just the climate model docs problems, but we still need to update the old examples

[jpfairbanks]

@lukem12345 ready to merge this?

[lukem12345]

@jpfairbanks Not yet, but I'll push docs changes that I have made. Some of the encoding of boundary conditions in the Poiseuille flow were handled strangely before we settled on current best practices. Also, need to double-check Diffusion-Advection and Boundary Conditions pages

[lukem12345]

Going to push fixes for overview.md and boundary_conditions.md in a few.

[jpfairbanks]

Nice, I am working on a new docs page using simple equations to go after overview for ASKEM.

[lukem12345]

Great!

On an editorial note:

I am wondering what the value we are getting out of using the star d star d formulation for diffusion is in our docs.

Defining diffusion/ the Laplacian this way:

• educates readers somewhat on the Exterior Calculus
• demonstrates the elegance of the DEC
• it is only a difference of 3 symbols to write one instead of the other
• the star d star d formulation is nicer if you want to build up simulations using only things you find directly in the de Rham complex
• not using Delta reduces the number of distinct symbols on the screen

However:
I think it should be used in examples only where the author wishes to educate readers on the definition of the Laplacian:

• this symbol is already recognizable to readers
• this is how users write Decapodes in practice
• via overload resolution, the Laplacian may automatically take the d del + del d form when the argument is a 1 form

I think making this change is a demonstration of the principle that composition should be performed:

• to reduce cognitive load
• in anticipation of components to be swapped out
• to express some physical intuition (e.g. magnetics is separate from hydrodynamics in MHD)
• for education's sake

Without any of the above reasons, it delves somewhat into composition for composition's sake.

[jpfairbanks]

I think it would be fine to use Laplacian everywhere we do diffusion and add a new docs page showing composition and aliases. I'll make a new issue to track composite operators. I'd like to get the old docs restored to main ASAP and then address general docs improvement in additional PRs

[jpfairbanks]

I did the merge conflicts

[lukem12345]

Added in versions of BC and Overview that should be correct barring git pulling and pushing doing odd things.

Poiseuille is partially fixed. The way that boundary conditions happened in the original Poiseuille doc was against current conventions, so this page has taken longer to work through. It will be shorter and clearer after that process is done, which is good.

In the meantime, BC and Overview docs can be merged through whatever git strategies of choice. (cherry-pick overview and bc changes onto a different branch, make a commit that removes changes to Poiseuille in this branch, etc.)

[lukem12345]

Looks like the environment on the GitHub action is slightly different than my local system for whatever reason. Some paths are getting mixed. I'll change some relative paths.

[jpfairbanks]

I'm going to merge this and if you have more docs updates, they can go in a new PR.