Feature Updates
- New users now see sign-up by default, with a clearer switch between account creation and login.
- Mobile record comments now use the full drawer width, respect device safe areas, and hide record navigation while open.
- Improved Markdown readability with tighter spacing, more consistent lists, and clearer heading hierarchy.
- Added a new first-run onboarding experience for eligible new Cloud users, guiding them through profile setup, data import, AI-recommended automations/apps, and team invitation so they can reach a usable base faster.
Bug Fixes & Improvements
- Improved schema-change safety by rejecting attempts to delete protected system-managed columns, preventing potential record data loss.
- Fixed two-way one-to-one link deletion so removing a relationship field no longer corrupts shared table structures or disrupts record requests.
- Improved multi-row paste reliability and performance, including pastes exceeding 17 rows. Missing or inaccessible target records now stop the operation instead of applying data to the wrong records.
- Improved performance for value-preserving field conversions, such as converting single select fields to text.
- Invalid or inaccessible record links now load the table normally after removing the invalid record parameter instead of returning a server error.
- Fixed App Builder generations remaining stuck in “running” after completing successfully.
- Improved cross-base conditional lookup and rollup reliability and performance, including correct refreshes when source records change.
- Improved App Builder GitHub integration with inline repository-name validation, clearer creation errors, safer handling of sensitive environment files, and automatic cleanup of stale merged branches.
- Fixed grouped views sorting date values incorrectly across different years.
- Improved schema update reliability for computed fields such as lookups, formulas, links, and rollups by preserving correct value types during rebuilds and backfills.
- Table creation with inline records now validates required fields before saving, preventing incomplete records from leaving table creation in an unusable state.
- Improved field deletion performance on large tables while preserving existing undo behavior.
- Improved duplicate-view performance for wide tables by reducing unnecessary data processing.
- Fixed duplicate-value API errors for legacy unique fields so responses include the conflicting field and localized details.
- Fixed crashes when configuring automation conditions with date fields and improved date filter reliability across triggers and condition editors.