|
| 1 | +--- |
| 2 | +title: Weekly progress summary – January 20, 2025 |
| 3 | +authors: |
| 4 | +- will |
| 5 | +tags: [progress, update, weekly] |
| 6 | +--- |
| 7 | + |
| 8 | +## Simulation Improvements |
| 9 | + |
| 10 | +### Haskell Implementation |
| 11 | + |
| 12 | +- Enhanced parameter handling with support for reading configurations and |
| 13 | + topologies from disk |
| 14 | +- Added new `generate-topology` command for random topology generation |
| 15 | +- Aligned Leios sortition with algorithms from sortition benchmarks and |
| 16 | + technical report |
| 17 | +- Completed analysis comparing Praos simulation with benchmark cluster |
| 18 | + - Adoption times within 10% of measured values |
| 19 | + - Review of simulation parameters pending |
| 20 | + |
| 21 | +### Rust Implementation |
| 22 | + |
| 23 | +- Implemented block-level visualization to complement existing graph view |
| 24 | +- Updated topology files to include baked-in latencies |
| 25 | +- Improved output with human-readable names from shared topology format |
| 26 | +- Enhanced simulation output comparability across different simulations |
| 27 | + |
| 28 | +## Analysis & Research |
| 29 | + |
| 30 | +### Sortition Analysis |
| 31 | + |
| 32 | +- Completed analysis of "Fiat Accompli" sortition scheme using mainnet stake |
| 33 | + distribution |
| 34 | +- Key findings for 500-vote committees: |
| 35 | + - 406 largest stake block-producers would be deterministic voters |
| 36 | + - ~88 voters would be randomly selected |
| 37 | + - Significant certificate size reduction achieved through deterministic voter |
| 38 | + selection |
| 39 | + |
| 40 | +### Downstream Impact Assessment |
| 41 | + |
| 42 | +- Initiated comprehensive analysis of Leios's ecosystem impact |
| 43 | +- Key areas identified: |
| 44 | + - Ledger and node data structure changes affecting indexers and explorers |
| 45 | + - Transaction construction impacts on dapps and wallets |
| 46 | + - Physical layer querying requirements for sophisticated use cases |
| 47 | + - Event filtering optimization needs for high throughput scenarios |
| 48 | + |
| 49 | +### DeltaQ Analysis |
| 50 | + |
| 51 | +- Successfully matched ΔQ model for IB diffusion across both simulation |
| 52 | + implementations |
| 53 | +- Identified key differences in simulation approaches: |
| 54 | + - Haskell simulation includes bandwidth effects (328ms network delay per hop |
| 55 | + at 1MB/s) |
| 56 | + - Rust simulation currently excludes bandwidth effects |
| 57 | +- Completed new comparisons using JSONL event logs |
| 58 | +- Investigating diffusion timing differences between implementations |
| 59 | + |
| 60 | +## Cryptography Performance |
| 61 | + |
| 62 | +- Completed comprehensive benchmarking of the Leios cryptography suite |
| 63 | +- Notable performance achievements: |
| 64 | + - VRF operations: 240 µs for proving, 390 µs for verification |
| 65 | + - Leadership checks efficiency: 0.17 µs per slot/pipeline |
| 66 | + - Vote processing: 3.8 µs per pipeline calculation |
| 67 | + - BLS operations: 1.5 ms for key verification, vote operations ranging 280 µs |
| 68 | + to 1.4 ms |
| 69 | + |
| 70 | +## Design Optimizations |
| 71 | + |
| 72 | +- Successfully optimized vote signature size down to 192 bytes |
| 73 | +- Validated committee certificates (500 votes, 60% quorum) compatibility with |
| 74 | + Praos blocks at ~58 kB |
| 75 | +- Explored integration possibilities with KES rotation and Praos VRF BLS keys |
| 76 | +- Finalized initial technical report cryptography sections |
| 77 | + |
| 78 | +## Simulation Progress |
| 79 | + |
| 80 | +### Multi-Platform Development |
| 81 | + |
| 82 | +- Haskell Implementation: |
| 83 | + - Achieved target diffusion latency matching benchmark cluster data |
| 84 | + - Added JSON-based event logging |
| 85 | + - Implemented 'short-leios' variant for mainnet ranking |
| 86 | + - Enhanced visualization for per-node data transmission |
| 87 | + |
| 88 | +- Rust Implementation: |
| 89 | + - Improved CPU simulation granularity |
| 90 | + - Resolved clock simulation race conditions |
| 91 | + - Implemented shared configuration system with standardized parameter files |
| 92 | + - Established default simulation parameters in YAML format |
0 commit comments