Update migration guide

Author: abey79

docs/content/reference/migration/migration-0-28.md

@@ -63,3 +63,30 @@ rotation and scale.
 Naturally, if any update to a transform always changes the same components, this does not cause any changes other than
 the simplification of not having to clear out all other components that may ever be set, thus reducing memory bloat both
 on send and query!
+
+## Python SDK: "partition" renamed to "segment" in catalog APIs
+
+<!-- TODO(ab): as I roll more API updates, I'll keep that section up-to-date --> 
+
+In the `rerun.catalog` module, all APIs using "partition" terminology have been renamed to use "segment" instead.
+The old APIs are deprecated and will be removed in a future release.
+
+| Old API | New API |
+|---------|---------|
+| `DatasetEntry.partition_ids()` | `DatasetEntry.segment_ids()` |
+| `DatasetEntry.partition_table()` | `DatasetEntry.segment_table()` |
+| `DatasetEntry.partition_url()` | `DatasetEntry.segment_url()` |
+| `DatasetEntry.download_partition()` | `DatasetEntry.download_segment()` |
+| `DatasetEntry.default_blueprint_partition_id()` | `DatasetEntry.default_blueprint_segment_id()` |
+| `DatasetEntry.set_default_blueprint_partition_id()` | `DatasetEntry.set_default_blueprint_segment_id()` |
+| `DataframeQueryView.filter_partition_id()` | `DataframeQueryView.filter_segment_id()` |
+
+The DataFusion utility functions in `rerun.utilities.datafusion.functions.url_generation` have also been renamed:
+
+| Old API | New API |
+|---------|---------|
+| `partition_url()` | `segment_url()` |
+| `partition_url_udf()` | `segment_url_udf()` |
+| `partition_url_with_timeref_udf()` | `segment_url_with_timeref_udf()` |
+
+The partition table columns have also been renamed from `rerun_partition_id` to `rerun_segment_id`.