high risk - early research stage
Impact if succesful
P1 - Urgent
Problems it is solving
- Gas costs –
- Generally – WG across lotus, CNL + CEL?
- JP + Alex + Aayush/Zen + driver
- PSD and ProveCommit are the major issue.
- With cron fix this will get worse
- Nicola – faster blocktimes (15 seconds) so more bandwidth
- Are PSD, Pre-Commit and ProveCommit all necessary?
- ProveCommit “proof” message is quite long– https://filfox.info/en/message/bafy2bzacebtmoboqwiumz2kysmbflodyabeg374vc33cxdpmudp6emoruk6we?block=bafy2bzacecpwc4iighh63p3gw65pcksqnpzu7fwko6xbpqgyfbmbegby3bmw6
- Is most of the gas used for message length or computation?
- Could bundle PSD and pre-commit potentially
- Could simplify software
- Any way to “reserve” for all onboarding messages at once?
- From JP on using gas lanes–
- The idea is that IF there's a lane for PSD with its own base fee and EIP1559-like transaction fee mechanism, the base fee for PSD would (1) be decoupled from other messages, so it wouldn't be affected by, say, FEVM (so no more of that argument) and (2) could, in theory, become smaller, provided that such a lane has sufficient bandwidth (i.e., PSD uses, on average, less gas than what it's allocated)