Add content for 4.5: Database and Merklized Storage

[sacha-l]

First pass at adding content for 4.5: Database and Merklized Storage lesson.

Still need to flesh some things out with a little help of SMEs.

Things to note when reviewing:

• Please include ideas for diagrams you'd like to see / add notes with TODOs.
• Not sure how the "Cost to Read and Write" slide fits (this was material from the original lesson plan)
• Most of this content is taken from Shawn's storage deep dive.

[shawntabrizi]

Where is merkle proof on this slide?

[cheme]

@sacha-l I did review a bit in https://github.com/paritytech/polkadot-blockchain-academy/compare/sl/module-4-db-storage...ec-4-5?expand=1 .
Some is just possible additional content. Tell me if you want me to create proper github comment.

[cheme]

Idea of slide for node description (general message is 'radix tree' is a well knowned struct and merklizing it is rather direct and elegant):

## Anatomy of a node

Tree contains nodes, in radix tree node is partialkey ++ n children ++ maybe_value

Merkle version is
partialkey ++ n children hash ++ maybe_value

Add constraint that terminal node always contains a value
Specialized this node for merklizing to these variants
Header branch ++ partialkey ++ n children hash ++ maybevaleu
Header leaf ++ partially ++ value hash
Value
And you got current trie
With the subtleties that if content under hash is smaller than it s hash, then it is written in the parent node instead of hash

[cheme]

Been thinking, I got this branch that is a very incomplete, quite bad implementation of mmr as child trie, I was thinking of removing content from it and make it a 'complete code' exercise.
Could be ok for this module (not directly related but it can touch the overlay logic, and the encoded node storage model).

[sacha-l]

Thanks for the review @cheme. Your exercise seems like a great idea. Please share it and we can add it as a workshop option for this lecture.

As for the ec-4-5 branch, yes please make a PR against this branch with comments on what stuff you're adding. 🙏🏻

[cheme]

The branch with child mmr is this one paritytech/substrate@master...cheme:substrate:child_other_contract2 that's pretty bad code (just drafting things).
The idea would be to work locally on the substrate branch (there is two test case so running both cargo test -p ... would take some time but only once so seems somehow manageable).
Still needs comments to help and more importantly removing the content to complete (not a huge work).
I did not implement the cumulus part which could be interesting as a followup.
Will try to do a PR (I also have to do one for 4.4 this afternoon (will be a bit off next week)).