v0.3.1-alpha
EDIT: Note that the binaries and archive files for this release were updated on 2023-11-20. The previous binaries and archive files were produced from commit cdd6707702788cccfe615be3b06fe34a96ce31c6. However, tag v0.3.1 is at commit fec4a4025db21d2e47994ad4bff29aa5d31d02e9. The new binaries and archive files have been produced from the correct staging environment (commit fec4a4).
What's Changed
Daemon Config Enhancements
Users can now configure postgres database config settings such as the total amount of active database connections on the command line:
RPC Enhancements
The txid of the batch is now returned along with the MintingBatch for tapcli assets mint finalize
Users can now specify a custom proof courier on the CLI. In addition, the gRPC max message size has been increased to ensure all data can properly be fetched with teh current set of RPC calls:
All Universe related RPC calls now have proper pagination support:
- Add pagination to universerpc calls by @GeorgeTsagk in #634
The number of assets has been removed from the main Info call to ensure the server dedup logic always succeeds:
The asset type has been moved into the asset_genesis struct:
- taprpc: Move asset type to genesis info by @GeorgeTsagk in #657
A redundant field has been removed from the anchor info RPC:
- RPC: Remove
anchor_txidfromAnchorInfoby @GeorgeTsagk in #663
Universe Sync Improvements
The default Universe sync algorithm has moved to an "on demand" approach. Before this release, clients would connect to a Universe an attempt to sync all assets ever created. This was very wasteful, as most of the time a new client likely only cares about a handful of assets. With this new model, by default no assets are synced until either: a user creates a new address with an unknown asset ID (JIT Universe Sync), or a user manually adds a specific asset config and syncs an asset via the assets universe sync call
- universe: stop syncer from syncing all known universes by default by @ffranr in #631
- address: optimistic asset sync by @jharveyb in #633
A series of new caches have been added to all Universe related RPC calls, to greatly improve the performance of the server under concurrent load:
The call to fetch all the universe roots has been optimized by making some of the more expensive joins (that return additional information) optional:
Less Debug Logging
The Universe sync process generally has less debug logging, which should make it easier to keep up with log activity:
Bug Fixes
A bug has been fixed where the daemon would fail to retry fetching a proof to complete a receive/send:
- Add backoff procedure to universe RPC proof courier by @ffranr in #617
- Add backoff procedure to the proof courier receive procedure by @ffranr in #637
- Ensure receiver node backoff retries finally succeed in asset send via universe proof courier by @ffranr in #644
A bug has been fixed that would cause the asset stats to factor in transfers, instead of issuance (the default intention):
Misc
- github: use the same issue templates as those found in LND repository by @ffranr in #614
- Rename universe functionality to accomodate transfer proofs. by @ffranr in #610
- mod: bump grpc and net libraries by @guggero in #630
- multi: rename RPC AssetLeaf proof field to account for transfer proofs by @ffranr in #640
- tapdb: ensure universe asset stats ticker is stopped before refresh by @Roasbeef in #658
New Contributors
- @GeorgeTsagk made their first contribution in #634
Full Changelog: v0.3.0...v0.3.1
