Monthly Contribution - August 2024
💡
Opened 4 PRs, and 3 PRs were merged to Prysm
Opened 1 PR to EIPs
Opened 1 PR to EIPs
Prysm
#1
- Fixed test files that has some misleading lines.
- e.g. in
beacon-chain/rpc/prysm/v1alpha1/beacon/validators_test.go
, there is a wrong assertion for length of the validator list. - My first contribution to Prysm! 😄
- Opened at August 2nd, Merged at August 3rd.
#2
- Fixed flaky tests in
beacon-chain/p2p/testing/mock_peersprovider.go
- Since iterating a map in Go is not guaranteed to be always same(see StackOverflow), tests are randomly failed.
- Opened at August 8th, Merged at August 13th.
#3
- Added a line of comment to be consistent with the actual implementation. (One line fix 😂)
- Opened at August 13th, Merged at August 17th.
#4
- An issue(#13586) raised: p2p
metaData
file "never" updated after initialization. This matters to the users who provide--p2p-static-id
flag. This flag is used to anchor their ID in the p2p network. - Referring to the consensus-specs,
MetaData
consists of two fields:
(
seq_number: uint64
attnets: Bitvector[ATTESTATION_SUBNET_COUNT]
)
- NOTE: There is an additional field in post-altair(
syncnets
). seq_number
is used especailly when a node pings to its peers. If another node received a ping message with lower sequence number than it has, that node rejects this ping message.- NOTE: Each node maintains its own peer table, and the sequence number is also maintained in the table.
- However, in the current implementation of Prysm, the sequence number is never updated: it means it remains
0
in the whole lifecycle. This can cause a significant issue with respect to p2p. - Opened at August 30th, Not yet merged.
- Seems like they are refactoring p2p package.
EIPs
#1
- I was investigating ePBS by reading EIP(link), found some typos.
- Folks often wrongly type
withdrawal
towidthdrawal
. (I did often too lol)