Once you understand these five terms, most operations across the Tinia interface become intuitive: node graph, dashboard, template, share, permissions. This page is the foundation for every tutorial that follows.
01 · Node Graph
Tinia models "one analysis" as a directed graph built from nodes and connections. Each node takes inputs, performs one task, and produces outputs; connections feed one node's output into the next node's input.
视频即将上线 · Coming Soon
Node graph editor overview: drop in nodes, connect them, configure parameters, run
≈ 1:30
Nodes come in three kinds:
Data Source Nodedataset_nodesource
Pulls the list of items to analyze from a data source
Analysis Nodeloudnessanalysis
Audio in → metrics / features out
Viewer Nodespectrum_viewerviewer
Renders results for embedding in a dashboard
Nodes can also be "disabled"
Right-click a node → once disabled, the entire downstream branch is skipped while its configuration is preserved. Handy for temporarily switching between two analysis paths without repeatedly deleting connections.
02 · Dashboard
After the nodes finish, results land on each node. The dashboard is where you reorganize those results into an interactive report—drop in a Viewer Block to show a chart, add a Slicer Block for linked filtering, add an Item List Block to let readers click through and compare.
截图占位 · Placeholder
Dashboard editor: available Block types on the left, a 12-column grid in the center, configuration panel on the right
A typical data flow is:
01
Add a dashboard node to the node graph
Connect the viewer nodes you want to display to it. After running the workflow, the dashboard has its "data source."
02
Open the dashboard editor
Click "Edit Dashboard" in the top-right of the run results page.
03
Drag a Viewer Block and bind it to a node output
Each Block picks a node port and reads its data from there.
04
Add linkage
When a Slicer or Item List changes, the other Blocks refresh automatically for the selected item.
Not every node can be embedded in a dashboard
Only nodes that implement the dashboard_view output port (Level 2/3 integration) can be bound to a Viewer Block. All official viewer nodes are supported; custom nodes must conform to the protocol.
03 · Template
A node graph + dashboard you've built can be packaged into a template—for colleagues to reuse, or published to the Store for any Tinia user to download. A template is the smallest unit for sharing "methodology" in Tinia.
A local template is saved in your own Tinia instance—edit and version it however you like. Ideal for accumulating internal team knowledge: capture a senior engineer's analysis workflow so newcomers can reproduce it in one click.
04 · Share
Tinia's share links bypass sign-in: anyone with the link sees the results directly, no Tinia account required. This is essential for sending conclusions to customers, letting line colleagues check anomaly reports, or handing results to external experts for review.
The "standalone page" design of share links
When you open a share link, the page has no sidebar and no back button—this is intentional. A share link is a "final presentation view," not an entry point into the product. To return to Tinia, just visit the Tinia domain.
Share links support:
Public (anyone) or private (password required)
Validity: permanent / 1 day / 7 days / 30 days
Revoke at any time
Sharing both workflow results and dashboards
05 · Permissions
Tinia's permission model is "module-level capabilities + owner isolation." Module level: whether you can create workflows, whether you can upload data sources. Owner isolation: even if everyone has the "view workflows" capability, you still see only the workflows you own or that are shared with you.
No resource-level ACLs
Tinia does not support fine-grained authorization like "grant only one specific user view access to one specific workflow." When you need to share a resource, use the "share link" mechanism (external) or "credential sharing" (internal).
Common Confusions
These concepts recur throughout
Quick Start, Templates, and the Administrator Guide all build on these five abstractions. If something later feels confusing, return to this page for a refresher.