Tables

Table widgets show the rows themselves: as a grid, as one record's fields, or aggregated into a pivot.

Table

The rows as they come, in the query's order.

Acme's overdue orders, the priority as a colored badge and the order's id kept for a click:

SELECT wo.number, c.name AS customer, s.name AS site,
       wo.priority, wo.due_at,
       CASE wo.priority WHEN 'URGENT' THEN 'red' WHEN 'HIGH' THEN 'orange' END AS _priority_color,
       wo.id AS _id
FROM work_orders wo
JOIN sites s ON s.id = wo.site
JOIN customers c ON c.id = s.customer
WHERE wo.status NOT IN ('COMPLETED', 'CANCELLED')
  AND wo.due_at < now()
ORDER BY wo.due_at
LIMIT 50

• Titles and formats come from each column's presentation in Database: priority reads Urgent, due_at in the viewer's date format. A computed column is titled by its alias.
• Types: numbers right-aligned in the viewer's format, booleans centered, dates and times in the viewer's zone, a file as its thumbnail.
• Sorting: a click on a header sorts the rows already loaded: ascending, descending, back to the query's order.
• Colors: _color tints a row; _priority_color shows priority as badges. See SQL Conventions.

Totals

A column's Total, Sum, Average, Minimum, Maximum or Count, adds a footer row. PostgreSQL computes it over every row the query returns, in the same read-only transaction, not over the rows shown, so it stays right when the result is cut at the row cap. It reads in the column's format.

TABLE

The rows as they come - columns starting with _ never show.

Size: 12 × 5 of the 12-column grid, at least 4 × 3 · Also: events, export

SELECT * FROM (VALUES (1001, 'Ana', 250.5, 'PAID', 'emerald'), (1002, 'Luis', 99.9, 'PENDING', 'amber'), (1003, 'Marta', 410, 'CANCELLED', 'red')) AS t(order_no, customer, total, status, _status_color)

Detail

One row as label / value pairs, with its image: the query's first row. The drawer says when the query returns more than one.

A vehicle chosen in a Select variable:

SELECT v.plate, v.make, v.model, v.year, v.status, v.odometer_km,
       v.insurance_until, v.inspection_until, v.photo
FROM vehicles v
WHERE v.id = $VAR.vehicle::uuid

DETAIL

One row as label / value pairs, with its image - the query's first row.

Size: 4 × 4 of the 12-column grid, at least 3 × 2 · Also: events, export

SELECT 'Ana Pérez' AS name, date '2026-10-01' AS last_visit, 12 AS visits, 'GOLD' AS level

Pivot table

The query returns the detail rows; PostgreSQL aggregates them into rows × columns, with subtotals and totals. Nothing is added up on screen, and an average is the real one, never an average of averages.

Billing by customer, a column per month:

SELECT c.name AS customer,
       to_char(i.issued_on, 'YYYY-MM') AS month,
       ii.quantity * ii.unit_price AS billed
FROM invoices i
JOIN invoice_items ii ON ii.invoice = i.id
JOIN customers c ON c.id = i.customer
WHERE i.status <> 'DRAFT'
  AND i.issued_on >= date_trunc('year', current_date)

Rows takes customer on its own. Set Columns to month, and the aggregate of billed to Sum.

Totals come first in the result, so the row cap never cuts them. More than 60 distinct values across the Columns is refused, with the count: group them in the query.

PIVOT

The query's detail rows aggregated by PostgreSQL into rows × columns, with subtotals and totals - never added up on screen.

Size: 12 × 5 of the 12-column grid, at least 4 × 3 · Also: events, export

SELECT * FROM (VALUES ('North', 'Jan', 120.5), ('North', 'Feb', 99), ('South', 'Jan', 80), ('South', 'Feb', 140.25), ('North', 'Jan', 30)) AS t(branch, month, total)

Next steps

• Visualization and Insertion: headings, text and frames.
• Dashboards in the App: downloading a table as CSV or Excel.