V2.1 update

[djrtwo]

Ready for review

There are a few bugs in the spec, noted here. This codebase reflects the fixes to these bugs so you'll see some disparities between the codebase and the spec. I expect Vitalik to fix these in a couple of days.

• Updated models
• Updated state transition
• Base sanity testing in test_full_pos.py

TODO:

• do block processing before state recalc
• add in get_indices_for_slot and get_block_hash. Code already handles these cases but using the functions from the spec will be more illustrative to other teams
• Need to include functionality for "Repeat while slot - last_state_recalc >= CYCLE_LENGTH:" . this handles case for when blocks transition past two state recalc points.

MISSING (saving for another PR)

• RPJ forkchoice

[djrtwo]

Most PR changes addressed except for a couple of questions above. Thanks @hwwhww !!

[djrtwo]

Also, there are some updates to the spec from V already that aren't reflected in here...

[djrtwo]

I think it would be clear if most “height”s in the spec are replaced with “slot”s, or "index"s if it's about the ith slot of this cycle. IMO It’s confusing because we usually use “heights” as the block serial numbers. Does it makes sense?

Hm.. you are right. I'll try to clean it up now

[djrtwo]

@hwwhww I'm thinking about leaving the following for a future PR. I'm pretty afk this week and don't want to hold you or anyone else up on collaborating

• cleaning up the naming around slot and height
• adding get_indices_for_slot and get_block_hash
• changing get_new_recent_block_hashes with suggested update (as it might change with the added helpers)

[hwwhww]

@djrtwo no problem! And I can add more unit tests in the following PR.