The fourth chapter of the Tara · find a practice track, rewritten in the Eve
edition on 11 September 2026. It follows the plan in EVE-V1-REDESIGN-TODOS.md
and the track brief in
tara-find-a-practice-track.md. The receipt is
under verification/.
The job#
This is the service-backed way into a practice, and the last of the ways in. A recommendation route, a canonical search adapter that assembles pools and filters them, saved-item operations, and a launch resolver that turns a card into a described destination.
It is also the chapter where seams are easiest to mistake for a system. Three read paths share a product word and read different inputs. A role selector returns the same object to every role. Saving sends a desired state and returns no reconciled view. A page walk can finish without being complete. And the launch resolver's whole job is to describe a destination it never opens.
Running example. One low-intensity recommendation — rec-low, with
meditation med-low — found through canonical search, saved for later, and
resolved into a launch descriptor that opens nothing.
Three learning promises (the cover)#
- Three read paths, one name.
- Saving is a command, and the readback is a separate operation.
- A launch descriptor describes — and navigates nothing.
Three things to remember (the close)#
- Trace the caller first.
- Saving is a desired state, and its readback has a scope that must travel with it.
- A descriptor describes: an action, a target, a query and an intent, and nothing visited.
Bridge. Moods, themes, modalities and teachers — chapter five, which closes the track.
Sections#
Four sections of six, five, eight and seven.
| Section | Slides |
|---|---|
| Three ways in, and who is admitted | the case, three paths, the guard, roles, pools, the order of operations |
| What a result actually says | field filters, three identities, object kinds, freshness, moment ordering |
| Saving, and reading it all back | intent, the route, the record, scope, the worked pair, failure, pages, duplicates |
| Opening an item that nobody opened | actions, precedence, the query, the parser, the destination, two closes |
Treatment decisions#
The plan again defaulted twelve slides to card grids. Each was re-read:
- Three became compare panels, where the rows already asked one question twice: what a role declares against what it receives, what an object says against what its action does, and how two entry points treat the same duplicate.
- One became a table. The bridge-moment ordering is three scoring tables of three numbers each, and the numbers are the rule. It is the chapter's only table and it is quantitative, so the values are typed and the caption names them.
- One became a layer stack. Build, filter, slice is an order with a rule at each boundary, and the argument of the slide is that the boundaries cannot be reordered by a small result limit.
- One became a timeline. The page walk is a bounded sequence along offsets where each point decides something, and the last point is where a finished walk is still not a complete view.
- One became a graph diagram — the conditional pools, with the two that need a configured member marked as unverified.
- Four stayed card grids where the rows really are peers: the three read paths, the three card kinds and their fields, the three freshness labels, the four card-to-action mappings, the three precedence tiers and the four parser families.
- Four gated flows became decision trees, each keeping exactly two outcomes, because the source has no third branch in any of them.
The capture, and where it came from#
The chapter has no surface of its own: it is adapters, routes and helpers. What it does have is a last section that spends five slides describing a destination, so the capture is that destination — the practise track's own pinned capture of the shipped sitting player, cited here rather than re-rendered. Three pins: the five-minute sitting the path names, the controls that nothing in this chapter pressed, and the ambient sound request the player reports as refused rather than playing silence and claiming success. That last one is the concrete reason a resolved path is not a completed journey.