|
1 | 1 | # Leios logbook
|
2 | 2 |
|
| 3 | +## 2024-10-02 |
| 4 | + |
| 5 | +### Team discussion of Haskell simulation |
| 6 | + |
| 7 | +- What general approach are we planning to take? |
| 8 | + - What questions? |
| 9 | + - What level of detail? |
| 10 | + - What order to do things? |
| 11 | +- What we don't need to model (at least initially) |
| 12 | + - The content of the ledger |
| 13 | + - Not needed for most questions |
| 14 | + - Can be done later |
| 15 | + - Instead, just model input block in terms of size |
| 16 | + - Cannot answer questions about tx lifecycle |
| 17 | + - Shared formats not needed for now |
| 18 | +- Approach |
| 19 | + - Start with existing network simulation, which is just a traffic simulation |
| 20 | + - Go back an do praos |
| 21 | + - 100s of nodes |
| 22 | + - Visualization |
| 23 | + - Add leios |
| 24 | + - Depart from paper by accounting for asynchrony in implementation |
| 25 | + - Might write an asynchronous version of the leios specification |
| 26 | + - Haskell stm version could be very similar to asynchronous leios spec |
| 27 | + - Stick with gtk visualizations |
| 28 | + - Record videos for dissemination |
| 29 | +- Questions that can be answered |
| 30 | + - Praos questions |
| 31 | + - Validate |
| 32 | + - Leios |
| 33 | + - Time for IB to RB |
| 34 | + - Repeat measurements for leios |
| 35 | +- Other |
| 36 | + - Rust might be the better option for visualization |
| 37 | + - Outputs from both simulations might both be visualizable with the same tool |
| 38 | +- For Thursday |
| 39 | + - [ ] Visualizaiton approach |
| 40 | + - [ ] Are common formats really not needed? |
| 41 | + - [ ] What is the status of the [leios-sim/](leios-simm/) folder? |
| 42 | + - [ ] What questions do we lose by having no simulation of txs? |
| 43 | + |
3 | 44 | ## 2024-10-01
|
4 | 45 |
|
5 | 46 | Report on ΔQ work in Rust so far (Roland):
|
|
0 commit comments