Tags give the ability to mark specific points in history as being important
  • v1.24.7 protected   This version is an update for API node operators. Both the get_account_history and get_account_history_rocksdb plugins now consistently use 0-based indexing of operations. The get_block_api plugin was also rewritten to eliminate excessive mutex locking. A new api call get_block_range(starting_block_num, count) was added to support fetching multiple blocks with a single api call. It was implemented in such a way that it can read multiple blocks in a single disk IO call.
    f8409240 · Update version to 1.24.7 ·
    Release v1.24.7

    This version is an update for API node operators (it is not a required update for consensus nodes and exchanges). Both the get_account_history and get_account_history_rocksdb plugins now consistently use 0-based indexing of operations. The get_block_api plugin was also rewritten to eliminate excessive mutex locking. A new api call get_block_range(starting_block_num, count) was added to support fetching multiple blocks with a single api call. It was implemented in such a way that it can read multiple blocks in a single disk IO call.

  • v1.24.6 protected
    Release v1.24.6
    • Fixed handling of limit by account_history_api.get_account_history & condenser_api.get_account_history calls

    This version was prepared to fix an issue specific to the way Bittrex processes Hive transactions. It's also strongly recommended for API node operators. It has no consensus changes (no changes needed by witness nodes) and is not a required upgrade for other exchanges that already have properly functioning deposits/withdrawals.

  • v1.24.5 protected
    Release v1.24.5
    • fixed fatal bugs in API locking implementation leading hived to hang on some machines under huge load
    • condenser_api.get_account_history call (wrapping account_history_api.get_account_history) is also lockless API
    • account_history_api.get_account_history allows operation filtering to reduce response size
    • account_history_api.get_account_history has limited response size to max 1000 operations at once (previously it was 10000)
    • fixed curation_rewards, author_rewards calculation in the Low memory node. All nodes built and replayed before 8e1a3777 require another replay.
    • fixed legacy code to enable API support for new virtual operation: sps_convert_operation added in the HF24. Previously cli_wallet and condenser_api.get_account_history, condenser_api.get_ops_in_block could report incomplete history of accoutns involved in such operation.

    As this version contains substantial operational improvements, we’ve tagged this version as v1.24.5, and it’s a strongly recommended upgrade for all API server nodes. However, there’s no need for exchanges or consensus nodes to upgrade to this version, since the performance improvements mostly come into play when a node is handling significant API traffic.

  • v1.24.4 protected
    Release v1.24.4
    • Fixes related to list_proposal_votes: #81
    • Fixes related to cli_wallet operating on new Hive chain-id
    • Fixes related to condenser_api::get_accounts in part specific to querying for account reputation
    • Fixed --rpc-endpoint option in cli_wallet
    • AH-rockdb plugin supports optional query for reversible operations (added include_reversible option)
    • cleanup of contrib related files like example config.ini files etc.
  • v1.24.3 protected   Optional upgrade for API nodes
    Release v1.24.3

    Fixed bug in state snapshot loading Added another build option COLLECT_ACCOUNT_METADATA to enable/disable account_metadata collection. Option is enabled by default.

  • v1.24.2 protected   HF 24 date moved to 06 October 2020 Fixed hived version reporting to match official tag.
    Release v1.24.2

    Hardfork date set for Oct 6th, 2020.

    New command-line options:

    • --dump-snapshot snapshot_name
    • --load-snapshot snapshot_name
    • --snapshot-root-dir directory_for_storing_snapshots (creates directory if it doesn't already exist)
    • --exit-after-replay

    Node management improvements

    • Snapshot dump/load: A node’s state information can be stored to disk, then used to setup a new node, without requiring a time-consuming replay of the blockchain. Snapshots are stored in the snapshot root directory as a subdirectory with the snapshot name supplied when creating the snapshot. By default, the snapshot root directory is a directory called snapshot in the application's data directory.
    • Replays can be stopped via Ctrl-C and resumed. To resume an interrupted replay, relaunch hived, and hived will auto-resume the replay if the state's headblock is less than the block_log's headblock.

    Performance Improvements

    • A full replay currently takes 16-18 hours. Previously this took several days.

    • Huge reduction in memory requirements. A fully featured API node now only requires 16GB RAM.

    • New virtual operations to supply more information to hivemind and other microservices:

    Governance changes

    • Hive must be staked for 30 days before it counts towards witness and proposal voting tallies.

    Updates to Decentralized Hive Fund

    • steem.dao account has been replaced by hive.fund account: #25
    • Hive in the treasury fund (hive.fund) will automatically be converted over time to Hive-backed Dollars (HBD). !97
    • The HBD in the decentralized hive fund doesn't count towards the debt ratio !97
    • HIVE donations to the Decentralized Hive Fund are now possible, the hive will automatically be converted to Hive-backed Dollars (HBD) and made available for proposals. !97
    • Proposals can be updated after creation time using update_proposal operation. Changes include title changes and lowering the proposal payout request. !48
    • Increased creation fee for long running proposals, the base price (10 HBD) increases by 1 HBD per extra day past 60 days !48
    • Fixed bug in descending order in DHF API (another regression from when worker proposal code was originally merged in): #31
    • Restored support for last_id parameter needed for proposal pagination (this was lost when Steemit merged in the original worker proposal code): #32
    • Proposal votes are summed up and available via API queries before a proposal becomes active (previously this data was only available after the proposal became active): #33

    Miscellaneous changes

    • Airdrop for accounts voted for via proposal system.
    • Major rebranding, replace all Steem/SBD references with Hive/HBD
    • Updated code to allow for major release version changes (e.g. 0.23.0 → 1.24.0)
    • Chain id changed from 0x0000… to 0xBEEABODE…
    • Improved validate_invariants to detect any future code changes that might leave the ledger in an inconsistent state.

    Long-standing bug fixes

    • Fixed bugs in node shutdown (by Ctrl+C) that led to intermittent crashes at shutdown.
    • Fixed fc bug unpacking flat sets that resulted in bug when restoring a snapshot.
  • v1.24.1 protected   - HF24 date moved to the 29. Sept - extended api_account_object in the database_api and condenser_api by post_voting_power data to simplify frontend integration.
  • v1.24.0 protected
    a0655d85 · Final fix in p2p-plugin ·
    Release v1.24.0
  • v1.0.12 protected   Fixed typo in the list of accounts to be recovered at HF24 Accounts cleared at HF23 are right now stored persistently, to be available after node restart.
  • v1.0.11 protected   Fixes related to bug in big::vector unpack (loading from stream) leading to fail state snapshot loading.
    Release v1.0.11

    New command-line options:

    • --dump-snapshot snapshot_name
    • --load-snapshot snapshot_name
    • --snapshot-root-dir directory_for_storing_snapshots (creates directory if it doesn't already exist)
    • --exit-after-replay

    Node management improvements

    • Snapshot dump/load: A node’s state information can be stored to disk, then used to setup a new node, without requiring a time-consuming replay of the blockchain. Snapshots are stored in the snapshot root directory as a subdirectory with the snapshot name supplied when creating the snapshot. By default, the snapshot root directory is a directory called snapshot in the application's data directory.
    • Replays can be stopped via Ctrl-C and resumed. To resume an interrupted replay, relaunch hived, and hived will auto-resume the replay if the state's headblock is less than the block_log's headblock.

    Performance Improvements

    • A full replay currently takes 16-18 hours. Previously this took several days.

    • Huge reduction in memory requirements. A fully featured API node now only requires 16GB RAM.

    • New virtual operations to supply more information to hivemind and other microservices:

    Governance changes

    • Hive must be staked for 30 days before it counts towards witness and proposal voting tallies.

    Updates to Decentralized Hive Fund

    • steem.dao account has been replaced by hive.fund account: #25
    • Hive in the treasury fund (hive.fund) will automatically be converted over time to Hive-backed Dollars (HBD). !97
    • The HBD in the decentralized hive fund doesn't count towards the debt ratio !97
    • Proposals can be updated after creation time using update_proposal operation. Changes include title changes and lowering the proposal payout request. !48
    • Increased creation fee for long running proposals, the base price (10 HBD) increases by 1 HBD per extra day past 60 days !48
    • Fixed bug in descending order in DHF API (another regression from when worker proposal code was originally merged in): #31
    • Restored support for last_id parameter needed for proposal pagination (this was lost when Steemit merged in the original worker proposal code): #32
    • Proposal votes are summed up and available via API queries before a proposal becomes active (previously this data was only available after the proposal became active): #33

    Miscellaneous changes

    • Airdrop for accounts voted for via proposal system.
    • Major rebranding, replace all Steem/SBD references with Hive/HBD
    • Updated code to allow for major release version changes (e.g. 0.23.0 → 1.24.0)
    • Chain id changed from 0x0000… to 0xBEEABODE…
    • Improved validate_invariants to detect any future code changes that might leave the ledger in an inconsistent state.

    Long-standing bug fixes

    • Fixed bugs in node shutdown (by Ctrl+C) that led to intermittent crashes at shutdown.
    • Fixed fc bug unpacking flat sets that resulted in bug when restoring a snapshot.
  • v1.0.10 protected   sbd_interest_rate/sbd_exchange_rate take precedence over hdb_ equivalents for better compatibility with old nodes.
    Release v1.0.10
  • v1.0.9 protected   effective_comment_vote_operation extended by total_vote_weight field
  • v1.0.8 protected   Internal version changed to 1.24.0
    6cf2f52e · Merge fix ·
  • v1.0.7 protected   Fixed bug in premature activation of DHF funds conversion. HF24 time set to 08. September 2020
    Release v1.0.7

    Fixed bug in premature activation of DHF funds conversion. HF24 time set to 08. September 2020

  • v1.0.6 protected
    Release v1.0.6

    Further changes related to hivemind data supply process New blockchain operation to convert DHF funds.

  • v1.0.5 protected   Another improvements needed to supply hivemind database.
  • v1.0.4 protected   Another fix of data provided by vop to the hivemind consuming operations during sync.
  • v1.0.3 protected   Other fixes to bugs detected during testing.
  • v1.0.2 protected   Fixes to bugs found during testing of previous version to be published.
  • v1.0.1 protected