Some quick (and dirty) notes:
Difficulty bomb probably needs to be prolonged
Block time with difficulty bomb:
end of march 14.6 s
end of june 20-21 s
end of august 32 s
Vitalik's preference is to release Metropolis end of june. Maybe even kick out a feature or two to avoid a later release (longer block time). But: Ethereum is so far ahead of other cryptocurrencies "in terms of $/sec" that the risc of hashpower migrating to a more profitabele chain is low.
Vitaliks latest thinking on transition to PoS:
PoW -> PoW with PoS "on top" -> Hybrid PoW/PoS -> PoS
PoS would finalize blocks after PoW has de facto finalized them anyway.
After a couple of month: implement clients that favor PoS.
Cut down PoW block reward, increase PoS validator reward.
Hybrid PoS scheme parametrize between PoW and PoS at 0-100 scale
Implemeting 100% of this roadmap might take longer than a year. Earlier stages of this transition may be ahead of schedule.
First part wouldn't be a hard fork but a very weird kind of soft fork. Second part would be a hard fork which could ease in over the course of 100 days.
Stage 1 is already being implemented (in python).
Excellent summary -- thank you!
This is super exciting!
Edit: I also really like the idea of a multi-stage approach and a series of "smaller bangs" vs one "big bang" in terms of the transition to PoS.
I've always understood that was to be the plan. Now it's nice to hear it actually being discussed and steps actually being taken in that direction.