Open source mining + AI vibe coding + talented, passionate grantees from @256FOUNDATION = lightning fast mining development.
All while the pub-cos throw in the towel.
We are going to win.
How to stack Bitcoin like a sovereign nation:
1. Mine Bitcoin - Bhutan
2. Buy Bitcoin - El Salvador
3. Confiscate Bitcoin - US
4. Steal Bitcoin - North Korea
5. Close the Strait of Hormuz and demand a toll in Bitcoin - Iran
@Leishman I had no idea River is mining bitcoin at decent scale and ~61% gross margin. Well done! Any other mining details you could share around hashrate, BTC production cost, etc. ?
Confirmation is an ordering process. An output must be in the ancestry of any confirmable spender. Only the first spender of an output can be confirmable in a given branch. Locktime and maturity require context.
Validation (everything else) is independent of order. It requires only knowledge of a block/tx and its prevouts. Some validation is even context free.
Consider how simple confirmation is relative to validation. Consider that validation can be evaluated fully concurrently across the entire block tree, and confirmation requires a total ordering by block.
Very different design constraints. A utxo store architecture forces validation to take on the total ordering constraint of confirmation, which absolutely destroys performance.
Almost every piece of software can now be reverse engineered.
Proprietary locking as moat about to be decimated by Open Source.
Adobe, Solidworks, etc...
🍿
AI workloads are characterized by elephant flows when all of the GPUs in a cluster exchange data through collective communication operations to synchronize data for distributed workloads. These flows can often lead to congestion and load balancing issues. (1/6)🧵
I get a lot of questions, so here's the latest. libbitcoin-node is functionally complete as a blocks-only node, sufficient for non-mining. It syncs to 900k in < 1 hour on a $350 mini over 2.3Gbps Internet. Work has moved to public query via libbitcoin-server for RPC compat.