What's Changed
- fix(clickhouse): support scd2 on replicated cluster tables by @iremcaginyurtturk in #1311
- fix(duckdb): create append tables without a primary key constraint by @iremcaginyurtturk in #1312
- fix(clickhouse): don't use ReplacingMergeTree for tables without primary keys by @iremcaginyurtturk in #1313
- fix(databricks): skip missing interval bounds in incremental reads by @iremcaginyurtturk in #1316
- fix(multitable): preserve terminal router errors for consumers by @karakanb in #1226
- fix(bigquery): read arrays and unparameterized NUMERIC from the source correctly by @iremcaginyurtturk in #1317
- fix(clevertap, hubspot): write NaN and Inf floats like nulls by @iremcaginyurtturk in #1318
- fix(clickhouse): create arrays with nullable elements instead of nullable arrays by @iremcaginyurtturk in #1320
- fix(iceberg): replace and merge tables with array columns by @iremcaginyurtturk in #1319
- BigQuery: warn when partition_by/cluster_by changes on an existing table by @iremcaginyurtturk in #1321
- cdc: support durable full refresh during streaming by @turtleDev in #1322
Full Changelog: v1.1.68...v1.1.69