Implement Relation and Rollup column for Database Tables

[domenico-ruggiano]

Hello team,

thank you for creating AFFiNE, I'm loving this tool for organizing and collaborating in my organization. I’d like to propose adding Relation and Rollup columns, which would allow users to create connections between data in multiple tables and leverage those relationships for deeper insights.

Relation Columns

Relation columns allow users to link entries between tables.

• Example: Connect a “Customers” table with an “Orders” table to see which customers placed which orders.
• Two-way relations: Automatically update both sides of the connection. For example, if a customer is added to an order in the “Orders” table, the order will appear in the “Customers” table.
• Self-referencing relations: Allow a table to reference its own entries, such as creating parent-child relationships within a “Tasks” table to track dependencies.

Rollup Columns

Rollup columns aggregate data from related tables based on relations.

• Example: In the “Customers” table, calculate the total amount spent by each customer based on data from the related “Orders” table.
• Aggregation options: Include sum, average, count, min, max, earliest date, latest date, and more.

Use Cases

1. Project Management: Link a “Tasks” table to a “Projects” table, allowing teams to see task breakdowns and project progress.
2. CRM: Connect a “Clients” table to an “Interactions” table to track client activity and engagement.
3. Inventory Management: Link a “Products” table to a “Sales” table to track stock levels and revenue.
4. Task Hierarchies: Use self-referencing relations in a “Tasks” table to create parent-child task relationships and track dependencies.

Benefits to Users

• Streamlined Data Organization: Relating tables reduces duplication and improves efficiency.
• Dynamic Data Insights: Rollups provide real-time summaries and calculations for better decision-making.
• Expanded Use Cases: Makes AFFiNE applicable to complex workflows, including project management, CRM, and inventory tracking.

I’d love to hear your thoughts and collaborate on refining this idea further.

[zigarot]

Yes Please,
I use a table for games with it's own fields, but I want the relational to the developers, which has it's own fields, contacts etc. this is also then linked to a publisher/strategist. This allows me to find details from a publisher that might have multiple developers that have multiple games.

[Lorex-ia]

Piggybacking what has been said above. You guys are doing amazing work !

[justinpik]

This two-way sync (or transclusion) is a must have feature for me and for others who will use databases (e.g., project management, PKM & second brain uses) often - Notion and Loop do this in some degrees. Without this feature, there is double and triple work and data integrity issues that result. If Table B (e.g. "Projects") is related to Table A (e.g. "Tasks") because some of the Tasks in Table A relate to various projects in Table B, the user should be able to make changes to A table from B table (update Task due date, start date, etc.) without having to flip back to table A. It's not enough to be able to display results from a related table - the user doesn't want to have to click into it to make changes then flip back. Otherwise love the product and direction. I feel like I'm in limbo between wanting to use Affine now and wanting to buy-in 100%, but I and many others need this feature.

[gavinmcfall]

+1 — this is the single biggest gap for anyone migrating off Notion.

I just moved a large personal/work Notion setup onto self-hosted AFFiNE: ~40 databases and ~1,250 rows across three workspaces. Almost everything came across beautifully (the block model + edgeless canvas are fantastic) — except the relations. I had several relation-heavy clusters that I was forced to flatten to plain text or drop entirely:

• A 6-table "racing scene" graph: drivers ↔ clubs ↔ venues ↔ events ↔ rulesets, including self-references (peer links, club affiliations, ruleset supersedes).
• A People ↔ Assessments / Goals / Teams / Work-History "career hub" — one Teams table alone had 10 relation columns pointing into People.
• A filament-catalogue ↔ print-profiles ↔ materials cluster.
• A task database with parent/child + "blocked by" self-relations.

Without relation columns, all that cross-linking is lost — you end up with disconnected tables and manual text references that drift out of sync. Rollups are the natural follow-on, but relations — especially two-way and self-referencing — are the must-have that would let AFFiNE fully replace Notion for structured data.

Big +1, and happy to help test whenever this lands on the roadmap.

[darkskygit]

We will advance and refine this feature in the next major version(0.28)
Converted to issue #15157 , so close this discussion