Teable
Open-source no-code database and Airtable alternative
A daily release train grinding on BYODB reliability and large-table performance.
◆Recent moves
- 1d ago
Teable sets a 10-minute floor on scheduled automations
Scheduled automations and email polling now require a minimum 10-minute interval for new or updated configurations, with existing ones grandfathered in. The rest is performance work on bulk deletes, view duplication and computed-field cascades — the same large-table reliability theme running through this train.
View source ↗ - 2d ago
Teable adds App Builder metadata and an instance-wide AI key control
Published App Builder apps gain custom favicons, page titles and descriptions, automation script nodes get a 180-second ceiling, and admins can now stop platform-managed AI keys from being handed to apps. The AI key setting is the notable one — it lets operators force apps onto their own approved credentials.
View source ↗ - 3d ago
Teable reworks collaborator management and space row-count handling
Space collaborators are grouped and paginated by person with Base-level permissions shown inline, and record-creation latency improves under high concurrency by reforming space row-count calculations. BYODB connection validation also tightened against internal network addresses.
View source ↗ - 4d ago
Teable adds BYODB health statuses and read-only conflict handling
Explicit health statuses, notifications and remediation guidance for read-only, unreachable, degraded and recovered databases, with writes returning 409 and computed tasks pausing until recovery. This is the core of the BYODB reliability arc.
View source ↗ - 4d ago
Teable consolidates repeated database errors in the admin Outbox
The Admin Outbox anomaly page now consolidates repeated database errors with changing details, making bulk recovery from incidents such as full disks practical. Operator-facing incident tooling rather than end-user capability.
View source ↗ - 4d ago
Teable fixes admin Outbox task counts and field naming conflicts
Corrects inaccurate failed-task counts when underlying tasks no longer exist and prevents rare field naming conflicts from disrupting computed updates. Small corrections inside the same admin-reliability push.
View source ↗