Nitopool is running on different algo - YescryptR16

Coin relaunched/forked on argon2id1024

The pools listed on miningstats are all for previous incarnation

New release 31.2 and 30.4

New feature: header PoW verification cache
This release adds a new header proof-of-work verification cache (pow_cache), built on the
same CuckooCache::cache structure already used elsewhere in the codebase (e.g. the signature
cache). The cache stores the result of already-verified headers (64 MiB by default, configurable
via -headerpowcachesize) and sits behind a single choke point, CheckProofOfWorkCached(), used
by both header validation (CheckBlockHeader()) and block reads from disk
(BlockManager::ReadBlock()). This means the cache speeds up essentially any PoW check on an
already-seen header — including serving blocks to other peers and re-reading blocks from local
storage — not just initial header sync. It starts empty and is warmed up by ordinary node
operation; misses always fall back to a full Argon2id recompute, so consensus rules are
unaffected.

New feature: parallel header PoW check queue
Separately, header PoW checks can now be dispatched across a dedicated parallel check queue owned
by ChainstateManager, instead of being checked one at a time. This only engages once a batch of
headers exceeds an internal threshold (32 headers) — in practice this means it's mainly active
during the presync stage of initial headers sync, where large batches of headers arrive from a
peer and need checking before the chain's cumulative work is confirmed; during normal steady-state
operation, where headers usually arrive in small numbers, checks are still processed sequentially.
Up to 6 threads can be used; on machines with fewer than 3 CPU cores, the queue automatically
falls back to a single thread so it never starves the rest of the node.


also backported fixes form Bitcoin master


Bug fixes (backported from upstream Bitcoin Core, not yet in official 31.x)
Fixed a race in FetchBlock() where a peer disconnecting between GetPeerRef() and
BlockRequested() could hit an assert on a missing CNodeState, remotely triggerable via the
getblockfrompeer RPC.
Fixed an undefined-behavior bug where a pruned node in a deep reorg could insert the same
(parent, block) pair into m_blocks_unlinked twice, corrupting the ordering invariant of
setBlockIndexCandidates.
Enforced BIP152's requirement that sendcmpct's announce field be 0 or 1; peers violating this
are now misbehaved and disconnected.
Hardened private-broadcast handling: newly queued transactions get a 5-minute grace period
before being flagged stale (previously they could be flagged within seconds), and sending to the
same node id twice for the same transaction is now prevented.
Fixed mempool package cleanup on an internal, not-known-to-be-reachable
"PolicyScriptChecks succeeded but ConsensusScriptChecks failed" path, so a failing transaction's
dependents are also removed instead of being left behind.
Fixed a cosmetic Taproot-deployment reporting issue in RPC output (Taproot itself was already
active at the validation level; only the RPC status report was affected).