Lack of order-preserving parallel parquet writing was my biggest gripe with DuckDB. Glad to see they implemented it! Between this duckdb release, DataFusion and

Lack of order-preserving parallel parquet writing was my biggest gripe with DuckDB. Glad to see they implemented it! Between this duckdb release, DataFusion and Apache Arrow (yes, I know that helps power the other projects I listed), I’m hoping I won’t have to touch a spark cluster for a long time.