Charts
A chart reads real data and draws it when the page is shown. Line, bar, area, pie, scatter — configured in the back office, placed like any other media item.
Live, and what that costs you
A chart on a page is a query that runs when somebody looks. That is its value and its one hazard:
A chart is a listing that does not look like one
Before publishing one, ask the question you would ask of a listing: should a visitor see this? A chart of monthly revenue is as revealing as the table it came from, and it is easier to put on a page by accident because it looks like a picture.
Charts obey the same visibility rules as everything else — unpublished records are not in them, and another organization's data never is. The question above is about what YOU choose to publish, not about a leak.
Where the data comes from
| Source | Use for |
|---|---|
| A listing | Anything already in the system: orders, stock, tickets |
| A data file | A series that is not platform data — a measurement log |
| Entered series | Small, fixed comparisons that will not change |
Prefer the listing, for the same reason a map should be fed by one: it cannot go stale.
Configuring one
Choose the type, the series, the labels and the axes. The chart library is vendored rather than loaded from a CDN, so a chart works on an installation with no outbound internet access — which is the normal case behind a factory firewall.
In documentation
A documentation page can carry a chart the same way it carries a live table —
through a view block in the Sourcebook dialect.
The page stores the instruction and the chart is drawn when the page is
shown, so a manual's figures are the current ones.
On a phone
Charts resize to their container and scroll inside their own frame when they cannot. The page never scrolls sideways.