Peter Nose
Peter Nose
It looks like `p386` crate doesn't zeroize all sensitive data. The crate uses `crypto-bigint` crate which constructs stack-allocated big integer types `Uint`s using an array of `Limb`s or wrapped `Word`s...
Fixed in #5920
> Will this enable running queries on a local Sapphire client node that doesn't have access to c10l state and it will automatically relay it to a remote RPC? This...
https://github.com/oasisprotocol/oasis-core/blob/master/CONTRIBUTING.md
> Idealy, we move [reindexing part](https://github.com/oasisprotocol/oasis-core/blob/master/go/consensus/cometbft/roothash/roothash.go#L680-L718) from `roothash.ServiceClient` to something like `worker.runtime.xxx`. Agree. I tried to do that in #6089.
> will not allow pruning of consensus blocks and or runtime storage diffs younger then expired history. Could you also provide the expected disk usage for historical checkpoint, consensus blocks...
I would prefer having creation with exact round and not start day. Maybe name `oasis-node storage checkpoint create` and `oasis-node storage checkpoint import` .
> This probably should wait for 25.10/26.0? Yes, this is blocked until we push 25.8 or higher to mainnet.