What's changed
Fix: the details pane dropped most of what it read. Codec, uncompressed size, row groups and hive partition layout were read from the Parquet footer, written to the cache, and then discarded on the way to the screen — so a partitioned dataset showed no partitions and a compressed one no codec. Only network roots, which take a different path, were unaffected.
- Home screen demo, and a note on how it is recorded. (docs)
modified now agonow readsmodified now.
Full changelog: v0.3.0...v0.3.1