Plan
- POV Theorique sur blockchains
- BTC wut dat?
History
The ‘pizza day’
Two guys bouht two pizzas for 10k BTCs.
JPMorgan
Ordered employees not to buy. Went out and bought some the same day.
Russia
At some point BTC was #2 most used currency in Russia, next day it was banned.
Wut dat?
TX
Alice sends X amount of something to Bob.
Made of
- Paye :
- A envoie X a B. Ca implique:
- A possede les BTC
- +x BTC pour A dans l’historique
- Coinbase:
- Pour les mineurs 50btc au debut
- Halving tout les 210k blocks
- On arrive a la fin vers year 2140
Payment
- T0: Miner(CoinBase)
- T1: SigMiner(Paye(PubKey_Alice, H(T0), H(T1), …))
- T2: SigAlice(Paye(PubKey_Bob), H(T1))
Payment to multiple
- T0:
Miner(CoinBase)
- T1:
SigMiner(Paye(PubKey_Alice, H(T0), H(T1), ...))
- T2:
SigAlice(Paye( [<PubKey_Bob, 0.25>, <PubKey_Alice, 0.75>], H(T1)))
Merkle trees
Hashes of hashes of hashes of hashes
Mining algo:
b <- block
do
Nonce <- rand
while H(PrevBlockHash + nonce) <= V
La valeur de V est la difficulte. Le but de V est de regler la difficulte sur le reseau, est de l’equilibrer de facon a avoir 1 block tout les 10min.
V c’est une fonction de l’histoire proche. Si le temps de decouverte est trop lent, on augmente le V, si trop court, on le diminue.
L’historique est evalue sur 2016 blocks precedents.
Consensus
- Vote : Impossible car taille de set evolutive
- PoW : Max puissance
- C’est qui qui mine lel (deterministe et non previsible)
- PoH (Hold)
- PoS (Stake)
- Je met en jeu mes coins
- Problem: “Nothing at stake” problem. Miners put money in two branches. If 80% places their money on both branches (malicious users), and only 20% plays fair, all it takes is 20%+
Epsilon to overtake the system.
- PoU (Use)
- On valorise ceux qui utilisent beaucoup la coin
- Breakable by sending back and forth with your friend
- Or huge taxes on txs
- PoI (Importance)
- On est important dans le reseau, les gens passent par nous bcp/nous envoient bcp de thunes/etc