Release Quine 1.5.6

Enhancements:

• Added support for STS assumed-role authentication to experimental AWS Keyspaces persistor

Bugfixes:

• Removed "total memory" gauge from metrics dashboard when maximum total memory can't be calculated based on the running JVM
• Fixed a "stream stopped before async invocation was processed" error message that could occur when resuming ingests on startup

Updates:

• webjars-locator to 0.47
• sbt-jmh to 0.4.5
• guava to 32.0.1-jre
• dropwizard metrics to 4.2.19
• logback-classic to 1.4.8
• scala-library to 2.12.18
• cats-effect to 3.5.1
• scala-collection-compat to 2.11.0

Release Quine 1.5.5

Enhancements:

• Ingest status is now preserved across restarts when restore-ingest configuration is set
• Improved error handling within Keyspaces persistor

Bugfixes:

• Valid Kafka configurations (e.g., ssl.truststore.location) are now accepted by the KafkaIngest stream configuration object
• Remove unnecessary log line

Updates:

• bootstrap to 5.3.0
• flatbuffers-java to 23.5.26
• pureconfig to 0.17.4
• commons-io to 2.12.0
• guava to 32.0.0-jre

Release Quine 1.5.4

Upgrading to Quine 1.5.4:

Temporal types have had a spring cleaning! Time values created in Quine 1.5.3 and earlier (without timezones) are automatically converted to LocalTime values in Quine 1.5.4. No user action is required.

Enhancements

• Added support for localtime function and values to the Quine graph and Cypher
• Updated documentation for Kinesis ingest to reflect the new name "Kinesis Data Streams"
• Added AWS Keyspaces persistor for preview
• Improved configuration linting and error reporting for Kafka ingest
• Added support for purgeNode procedure to RocksDB and MapDB persistors
• Added automatic recovery (up to 3 attempts) of ingests that fail to start/restart due to upstream issues
• Simplified reify.time-generated nodes to use static labels for easier styling
• Improved documentation for administrative APIs
• Added scheduler checkpoint settings maxBatchSize and maxBatchWait to Kinesis Data Streams ingests

Bugfixes

• Resuming a failed ingest now reports an HTTP error instead of timing out
• Time values in the Quine graph now correctly store and use offsets from UTC
• The date function now always correctly returns a date instead of a datetime.
• Errors that occur while submitting an invalid standing query are now reported as "Bad Request"

Updates

• Base docker image to JVM 20
• sbt-scalajs-crossproject to 1.3.1
• scalajs to 1.13.1
• circe to 0.14.5
• pureconfig to 0.17.3
• endpoints4s to 3.8.15
• logback-classic to 1.4.7
• scala-collection-compat to 2.10.0

Release Quine 1.5.3

Enhancements:

• A distinct "advertised"/canonical HTTP address used for self-referencing links may be set via quine.webserver-advertise
• Added purgeNode procedure, which may be used to permanently delete a node and all data owned by that node (including historical) from Quine, including the persistence layer
  Bugfixes:
• Error message from Cypher compilation failures involving int.add now report the correct types
• Several typos throughout the codebase have been corrected (contributed by @stavares843)
• Slack notifications from Standing Queries now render with the intended rich formatting
• API documentation for DebugOps now correctly renders the usage disclaimer for those APIs
  Updates:
• logback-classic to 1.4.6
• akka-http-backend, slf4j-backend to 3.8.14
• dropwizard metrics to 4.2.17
• circe-config to 0.10.0
• memeid4s to 0.8.0
• proto-google-common-protos to 2.14.2
• sbt-assembly to 2.1.1
• flatbuffers-java to 23.3.3
• rocksdbjni to 7.10.2

Release Quine 1.5.2

Enhancements:

• Added first-class support to the Quine engine and the cypher interpreter for new temporal types: Date, Time, LocalDateTime, and Duration
• Added int.add (replacing incrementCounter), float.add, set.insert, and set.union procedures for atomically updating properties of different types
• Added castOrNull and castOrThrow cypher functions for providing type hints to the query compiler for values
• Replaced remaining uses of ujson with circe JSON for better support of 64-bit integers
• Replaced LiteralOps API with a new DebugOps API, focusing on presenting human-readable versions of node state
• Removed auto-commit option from Kafka ingest API: To continue using autocommit, pass "autoCommitIntervalMs": 5000 in the kafkaProperties option instead

Bugfixes:

• Complex strId-based queries are now executable without an all-node scan (contributed by @harpocrates)
• MERGE queries now correctly combine the scopes of their CREATE and MATCH possibilities (contributed by @harpocrates)
• Log messages from IngestRoutes no longer report as being logged by a different class (e.g., "QuineApp")

Docs:

• Added "servers" field to openapi.json
• Added missing documentation to metrics API
• Added should-resume-ingest to documented config
• Remove "Standing Query Output Format" from named schemas
• Updated style of ingest APIs
• Standing query Propagate endpoint status code changed from 200 Success to 202 Accepted

Updates:

• flatbuffers-java to 23.1.21
• jquery to 3.6.3
• kafka-clients to 3.3.2
• proguard-base to 7.3.1
• pulsar4s-akka-streams to 2.9.0
• rocksdbjni to 7.9.2
• sbt-jmh to 0.4.4
• scalajs-compiler to 1.13.0

Release 1.5.1

Enhancements:

• Switched JSON parser from ujson to circe to avoid loss of
  precision in large integers in standing queries (Further support for
  large integers coming soon)
• REST API documentation style and readability updates
• Added documentation for graph AI use cases
• Improved performance of Cassandra persistor by reducing load and
  latency during ingest

Bugfixes:

• Reduced log noise when performing more than 2 operations on a node as
  part of a single message
• Fixed an issue with intellij project import

Release 1.5.0

Release 1.5.0

Enhancements

• Many API endpoints now support YAML request bodies: Convenient for copying pipelines from recipes to a Quine server!
• AWS client configurations (SNS, Kinesis, etc) separated "region" from "credentials"
• Added "Graph Algorithms" category to the REST API
• Added random walk graph algorithm (compatible with Node2Vec and GraphSAGE walk generation), REST endpoint, and Cypher procedure
• Administration REST API documentation style and readability updates
• Non-historical nodes can now be kept awake for a minimum duration after each message
• Kafka ingests may now include a settings block for the Kafka client, e.g., to specify credentials. Contributed by @charliemenke
• No longer require all fields to be specified in recipes, improving readability
• Performance improvements in MapDB, Cassandra, DistinctId standing queries
• Improve indexing of node edge collections for increased query performance

Bugfixes

• Switched JSON parser from ujson to circe to avoid loss of precision in large integers in ingest streams (Further support for large integers coming soon)
• Resolved issues with clean shutdown of the Quine application via the API

Updates

• Update sbt-ghpages to 0.7.0
• Update sbt-git to 2.0.1
• Update sbt-assembly to 2.1.0
• Update slinky to 0.7.3
• Update rocksdbjni to 7.8.3
• Update scala-js-macrotask-executor to 1.1.1
• Update pulsar4s-akka-streams to 2.8.2
• Update dropwizard metrics to 4.2.15
• Update flatbuffers-java to 22.12.06
• Update sbt-paradox to 0.10.3
• Update scalajs to 1.12.0
• Update scala-java-time to 2.5.0
• Update endpoints4s to 1.9.0
• Update logback-classic to 1.4.5
• Update mapdb to 3.0.9
• Update org.eclipse.jgit to 6.4.0.202211300538-r
• Update proguard-base to 7.3.0
• Update jnr-posix to 3.1.16
• Update bootstrap to 5.2.3
• Update pureconfig to 0.17.2
• Update classgraph to 4.8.151
• Update scala-collection-compat to 2.9.0
• Update sbt-scalafmt to 2.5.0
• Update protobuf-java to 3.21.10
• Update cats-core to 2.9.0

Release 1.4.2

Release Quine 1.4.2

Enhancements:

• Added canonicalized mapping of QuineIds and DGBs to reduce heap usage
• Simplified node wakeup protocol
• Improved contributor experience when building with node version greater than 16
• Now support amd64 and arm64 architectures for Quine docker image

Bugfixes:

• Node debug endpoint / procedure now includes all DistinctId localEventIndex entries instead of just one

Release 1.4.1

Bugfix:

• Edge case where Akka complains of an actor name being re-used too quickly
  is now retried.

Release Quine 1.4.0

Release Quine 1.4.0

Enhancements

• The reify.time procedure now yields only the finest-granularity reified period node for better query ergonomics
• Removed nullary and unary locIdFrom variants to simplify locIdFrom usage
• Enriched logging in edge cases involving shard resolution
• Restructured ingest streams for better consistency and handling of edge-cases involving charset transcoding
• Optimized storage of DateTime to use less space
• Simplified node wakeup protocol: edge cases involving simultaneous request to sleep and wake should now be more efficient
• Improved API documentation of ingest streams
• Add notification preview text for rich Slack messages used as Standing Query outputs
• Provide a warning when a full-node scan is detected in ingest query
• Added text.urlencode and text.urldecode Cypher functions
• Added atomic "count" return value to incrementCounter procedure
• Added support for decoding steps during ingest (base64, zlib, and gzip)
• Standing Queries can now use idFrom-based ID constraints, provided that all arguments to the idFrom are literal values
• Rewrote the serialization, persistence, and message passing system used by DistinctId Standing Queries
• Change default order of node effects to "persister first", applying in-memory effects only after the persistence operation succeeds
• Removed unused / experimental "node merging" behavior

Bugfixes

• Cassandra persister batched writes now respect configured timeout and consistency options
• Long-typed QuineIdProviders will now interoperate correctly with the getHost procedure
• singleton-snapshot and journals may now be enabled at the same time
• Added a heuristic rate limiter to Kinesis ingest streams to improve stability when Quine reads faster than Kinesis can emit
• Improved cluster stability when cluster members experience temporary disconnections
• Improved shutdown behavior in failsafe case
• UUIDv5 and UUIDv3 id providers now generate correctly bit-masked identifiers
• MultipleValues standing queries are now properly restored on nodes where shutdown is started and then cancelled
• Node edge and property counts will now be correctly reflected in the metrics dashboard
• Improved use case for reify time to ensure reified time nodes have all properties set

Updates

• Replaced snakeyaml with snakeyaml-engine, supporting YAML 1.2
• akka to 2.6.20 (contributed by @He-Pin)
• bootstrap to 5.2.2
• cassandra client to 4.15.0
• endpoints4s to 1.8.0
• endpoints4s (akka-http) to 7.0.0
• endpoints4s (openApi) to 4.2.0
• endpoints4s (xhr) to 5.1.0
• jquery to 3.6.1
• netty-nio-client to 2.17.288
• parboiled to 1.4.1
• scalajs to 1.11.0
• rocksdb to 7.6.0
• scalajs-dom to 2.3.0
• flatbuffers-java to 22.10.26
• dropwizard metrics to 4.2.12
• logback to 1.4.4
• apache commons to 1.10.0
• protobuf-java to 3.21.8
• Novelty: graal to 22.2.0
• Memeid4s to 0.7.0