Skip to content
ProbeDocs

Sequence vs Structure view

The log table has two view modes: Sequence and Structure. They render the same captured traffic, but the layout changes how quickly you can answer different questions. Sequence wins when you're watching events happen live; Structure wins when you're exploring an unfamiliar API.

You switch between them with the view-mode button in the FilterBar above the log table. The current mode persists across sessions.

Log table in Structure view: captured traffic grouped into a tree by host (api.example.com, api.github.com, api.stripe.com …) with expandable path-segment nodes and a per-node request count

Sequence view

Section titled “Sequence view”

Sequence view is a flat, chronological list — capture-time order, newest at the top. Each row is one request: method badge, host, path, status, duration, size.

Use Sequence when:

  • You're reproducing a flow and need to see the order events fired in.
  • You're chasing a race condition where two requests interleave.
  • You want the lowest-friction view for live debugging — new rows appear instantly without any tree to expand.
  • You're scanning for a single anomaly (an unexpected 500, a long duration) across all hosts at once.

Sequence view orders rows by capture time, newest first. New rows appear at the top as traffic arrives.

Structure view

Section titled “Structure view”

Structure view groups requests into a tree: one node per host, then one node per path segment. A path that was requested once ends in a node with a badge. A path that was requested more than once has no node of its own: each of its requests is a row, showing the method, status and duration, with no badge and nothing to expand.

Hosts show as scheme://host (a non-default port stays in), busiest host first; path segments show without slashes, in alphabetical order. The badge on a host or folder counts the requests under it and everything below it.

Diagram of the log table in Structure view. Hosts are the top-level nodes, shown as scheme://host with the busiest host first: https://cdn.example.com (4 requests) holds the folder static (4), which holds three separate request rows for app.js and the leaf logo.png (1). https://api.example.com (3) holds the leaf health (1) and the folder v1 (2), which holds two separate request rows for users, the first one selected. A path requested once is a node with a badge; a path requested more than once has no node and no badge, and each of its requests is its own row with its method, status and duration. A folder's badge counts the requests under it and everything below it; path segments are shown without slashes. Diagram of the log table in Structure view. Hosts are the top-level nodes, shown as scheme://host with the busiest host first: https://cdn.example.com (4 requests) holds the folder static (4), which holds three separate request rows for app.js and the leaf logo.png (1). https://api.example.com (3) holds the leaf health (1) and the folder v1 (2), which holds two separate request rows for users, the first one selected. A path requested once is a node with a badge; a path requested more than once has no node and no badge, and each of its requests is its own row with its method, status and duration. A folder's badge counts the requests under it and everything below it; path segments are shown without slashes.

Use Structure when:

  • You're exploring an API you've never touched before and want a quick map of its endpoints.
  • You're grouping a long debugging session — all calls to /v1/users listed together under v1, one row per call.
  • You want to see request counts per endpoint at a glance (the badges).
  • You're auditing what an app talks to: every host appears as a top-level branch.

Click a host or folder to expand or collapse it. Click a request row to load it into the Detail panel just like in Sequence view.

Filtering and search behavior

Section titled “Filtering and search behavior”

Both view modes share the same filter bar, search input, and sidebar selection. Whatever's selected in the sidebar (a domain, a Filtered List, the All node) narrows the rows in either view. Search inside the FilterBar also applies to both views.

In Structure view, filters affect which leaves are present, and empty branches collapse out automatically — so flipping on the 5xx chip in the FilterBar collapses entire hosts that have no errors.

Domain Watch gates traffic at capture time, so it affects both views identically: unwatched domains show their teaser only.

When to switch

Section titled “When to switch”

A practical pattern:

  1. Start in Sequence while you reproduce the bug — you want to see exactly when each request fired.
  2. Switch to Structure once you've got the relevant traffic captured — fold up the noise, navigate the API surface, and click into the specific endpoints you care about.
  3. Save the result as a session — the captured rows travel in the .pro file, and your reader can flip between Sequence and Structure on their end.
Section titled “Related”
  • Detail panel — what each row opens into
  • Search — filtering inside both view modes
  • Domain Watch — cuts noise before traffic reaches the table
  • Sessions — save the current view for later