What makes a catalog open? Unity Catalog Advances the Open Lakehouse

Databricks published a post arguing that support for open table formats is no longer enough to call a data catalog open, framing its Unity Catalog product as the piece that advances what it calls the open lakehouse. The excerpt distinguishes between platforms that handle open formats and what actually makes a catalog open, though the details behind that distinction are not laid out in the fragment available.
The distinction the headline draws is a useful one for teams building on lakehouse architectures. Open formats such as Parquet, Delta Lake, Apache Iceberg and Apache Hudi govern how tabular data is written to object storage, so any engine can read the files. A catalog sits a layer above: it holds the table names, schemas, permissions and lineage that determine who can find and use which data. In practice, a team can store everything in open files while still locking access and metadata inside a single vendor's control plane, which is the gap the post appears to be pointing at. That framing is my reading of the headline and excerpt rather than a reported claim.
No version numbers, benchmarks, pricing, availability dates or customer results are included in the material, so the substantive argument and any product changes remain to be learned from the full post.
Why it matters: developers and platform teams evaluating lakehouse stacks are being pushed to judge catalogs on governance and portability, not just format support. If catalog-level openness becomes a buying criterion—an inference, not a stated outcome—vendors will face pressure to show that metadata, permissions and lineage can move between engines as freely as the underlying files.
Based on reporting from the original publisher. Visit the source for full context and later updates.
Publisher excerpt
While almost every platform now supports open formats, what makes a catalog truly...