v0.36.0

0.36.0 (2024-09-13)

BREAKING CHANGES

• Remove support for USER_CONTEXT (#8705) (8a796f8) - This functionality was deprecated starting from v0.26.0. Please migrate to SECURITY_CONTEXT.
• chore!: Remove support for checkAuthMiddleware (86eadb3) - checkAuthMiddleware option was deprecated in 0.26.0, because this option was tightly bound to Express. Since Cube.js supports HTTP and WebSockets as transports, we want our authentication API to not rely on transport-specific details. We now recommend using checkAuth.
• feat(cubesql)!: Enable CUBESQL_SQL_NO_IMPLICIT_ORDER by default (f22e1ef) - It's started to be true. it means that SQL API will not add ordering to queries that don't specify ORDER BY, previously it was true only for ungrouped queries.
• chore!: /v1/run-scheduled-refresh - was removed (7213ae7) - Removing this method since Cube is designed and supposed to be run as a cluster, rather than a serverless application, in a production setting. Use the Orchestration API instead.
• chore!: Remove cache & queue driver for Redis (eac704e) - Starting from v0.32, Cube Store is used as default cache and queue engine. Article: https://cube.dev/blog/replacing-redis-with-cube-store
• chore!: Support for Node.js 16 was removed (8b83021) - Node.js is EOL, it was deprecated in v0.35.0.
• feat(docker)!: Remove rxvt-unicode (which was used as TERM) (fb9cb75). It was not useful, and this TERM requires a lot of libraries to be installed. It costs 148 MB of additional disk space.
• cubejs-client/playground: New query builder & new chart prototyping (6099144) - Playground was upgraded to Playground 2.0 and Chart Prototyping, previously available in Cube Cloud only.

Features

• cubesql: Support information_schema.sql_implementation_info meta table (841f59a)
• databricks-jdbc-driver: Support Java higher then 11 (03c278d)
• docker-jdk: Upgrade JDK to 17 from 11 (c3a1ccd)
• docker: Upgrade Node.js to 20.17.0 (9fedf78)
• docker: Upgrade OpenSSL to 3 from (1.1) (97159e7)
• docker: Upgrade OS to bookworm from bullseye (67e9521)
• docker: Upgrade Python to 3.11 from 3.9 (7684579)
• schema-compiler: Support custom granularities (#8537) (2109849)
• schema-compiler: Support custom granularities for Cube Store queries (#8684) (f7c07a7)
• Upgrade aws-sdk to 3.650.0 (0671e4a)

v0.35.81

0.35.81 (2024-09-12)

Bug Fixes

• api-gateway: fixes an issue where queries to get the total count of results were incorrectly applying sorting from the original query and also were getting default ordering applied when the query ordering was stripped out (#8060) Thanks @rdwoodring! (863f370)
• cubesql: Use load meta with user change for SQL generation calls (#8693) (0f7bb3d)
• Updated jsonwebtoken in all packages (#8282) Thanks @jlloyd-widen ! (ca7c292)

Features

• cubestore: Support date_bin function (#8672) (64788de)
• ksql and rollup pre-aggregations (#8619) (cdfbd1e)

v0.35.80

0.35.80 (2024-09-09)

Bug Fixes

• schema-compiler: propagate FILTER_PARAMS from view to inner cube's SELECT (#8466) (c0466fd)

Features

• cubesql: Fill pg_description table with cube and members descriptions (#8618) (2288c18)
• cubesql: Support join with type coercion (#8608) (46b3a36)

v0.35.79

0.35.79 (2024-09-04)

Bug Fixes

• schema-compiler: correct string casting for BigQuery (#8651) (f6ac9c2)
• schema-compiler: correct string casting for Databricks (#8652) (5cf71d9)
• schema-compiler: correct string casting for MySQL (#8650) (9408bb4)
• schema-compiler: correct string concatenation and casting for MS SQL (#8649) (0e9f9f1)
• schema-compiler: Support CURRENT_DATE SQL push down for BigQuery (9aebd6b)

Features

• cubejs-api-gateway: Support returning new security context from check_auth (#8585) (704a96c)

v0.35.78

0.35.78 (2024-08-27)

Bug Fixes

• cubesql: Don't clone AST on pre-planning step (#8644) (03277b0)
• cubesql: Don't clone response, improve performance (#8638) (4366299)
• cubesql: Fix non-injective functions split rules, make them variadic (#8563) (ed33403)
• cubesql: Normalize incoming date_part and date_trunc tokens (#8583) (f985265)

Features

• cubejs-api-gateway: Add description to V1CubeMeta, V1CubeMetaDimension and V1CubeMetaMeasure in OpenAPI (#8597) (1afa934)
• cubesql: Introduce max sessions limit (#8616) (dfcb596)
• cubesql: Upgrade serde, serde_json - performance boost (#8636) (b4754db)
• schema-compiler: detect boolean columns for schema generation (#8637) (178d98f)

v0.35.77

0.35.77 (2024-08-26)

Bug Fixes

• Row deduplication query isn't triggered in views if multiple entry join path is used (#8631) (1668175)
• View builds incorrect join graph if join paths partially shared (#8632) Thanks @barakcoh for the fix hint! (5ca76db), closes #8499

Features

• cubesql: CubeScan - don't clone strings for non stream response (#8633) (df364be)

v0.35.76

0.35.76 (2024-08-24)

Bug Fixes

• Invalid interval when querying multiple negative interval windows within Cube Store (#8626) (716b26a)

v0.35.75

0.35.75 (2024-08-22)

Bug Fixes

• Internal: ParserError("Expected AND, found: INTERVAL") in case of trailing and leading window parts are defined for pre-aggregated rolling window measure (#8611) (18d0620)

v0.35.74

0.35.74 (2024-08-22)

Bug Fixes

• allow non-stream mode for sql over http (#8610) (6a16e67)

Features

• cubesql: Add pg_class self-reference (#8603) (05ea7e2)

v0.35.73

0.35.73 (2024-08-21)

Bug Fixes

• cubesql: Reduce memory usage while converting to DataFrame (#8598) (604085e)
• cubesql: Use date_part => date_part + date_trunc split only with appropriate date_part argument (#8552) (9387072)
• cubestore: Reduce memory usage while converting to DataFrame (#8599) (15f9a0f)

Features

• cubejs-schema-compiler: Support FILTER_GROUP in YAML models (#8574) (f3b8b19)