View presets design

View presets design

Roger asked on 25 October 2026 which orderings, layouts and views suit different brain types, with WCAG kept in mind. He also asked that every choice be saved for later. This page is the plan. The requirements are editor spec section 69, workstream VP1. The research behind it is summed up below.

The short answer

• Presets are named by need, never by diagnosis. People inside one group differ more than the groups differ.
• Every choice saves on its own, and a named preset bundles them.
• WCAG 2.2 AA is a floor. No preset goes below it, and the device's own settings win.
• The default for everyone is a 2D drawing with an outline beside it. 3D, VR and fisheye stay as extras.

flowchart LR
  P["Preset"] --> O["Ordering"] --> L["Layout"] --> V["View"] --> F["Formatting"] --> M["Motion"] --> T["Text and keys"]

Loading

Built so far

The six starters are in packages/core/src/view-presets.ts. View preset heads the Drawing settings: picking one sets every Drawing setting it turns, and each stays yours to change. The floor, the device winning, the nearest preset winning and the one-file export are there too. The hosts read the device's settings, and a graph or vault keeps its own preset, next.

A preset sits beside a personal profile (section 39). The preset says how a graph is drawn. The profile says how much of it a viewer is shown.

Starter presets

Starter	Ordering	Layout	View	Motion	Extra
Default	By flow	Rows, or stress for a network	2D and outline	Follows the device	AA floor
Focus support	Hubs first	Rows, the focus in the middle	2D, the rest blurred	Off	A trail of recent picks
Calm and predictable	One order the person picks	Pinned, laid out again only when asked	2D or grid	Off	A preview before a change
Detail first	By name or property	Grid or matrix	Matrix or outline	Off	Every label shown
Big picture first	By kind or group	Groups as boxes, folded	Semantic zoom	Optional	One key to fit all
Reading support	By flow or kind	Rows, short labels	2D and outline	Follows the device	Wider spacing, read aloud

What the research says

Finding	How sure	Source
Almost no study tests these groups reading node-link graphs	High	Wu 2026, Tran 2024
With ADHD, a distractor holds attention longer once it catches it	Medium	Zhang 2024
A busy but relevant view distracts less than a sparse one	Medium	Forster and Lavie 2023
ADHD readers prefer blurring the rest to hiding it, and saved presets	Medium	Zhu, ASSETS 2025
No evidence that ADHD subtypes need different layouts	Medium	Search found none
Autistic readers see the whole more slowly, not worse at it	High	Van der Hallen 2015
Uncertainty and surprise changes raise anxiety for many autistic people	High	Jenkinson 2020
Fonts made for dyslexia do not help; wider word spacing does a little	High	Wery 2017, Kuster 2018, Galliussi 2020
3D hurts memory of where things are	Medium	Kotlarek 2020
A matrix suits a dense graph, a node-link drawing a sparse one	High	Okoe 2018

The full research, with 34 sources and a confidence for each choice, is kept with the project files (neuro-layouts/research.md).

Safe for everyone

• Contrast at least 4.5:1 for text and 3:1 for shapes and focus.
• Colour never the only cue: a shape or an icon too.
• Motion follows the device; nothing plays by itself; a pause for anything past five seconds.
• Laid out again only on request; pinned things stay put.
• Every action by keyboard, with undo and a way that needs no dragging.
• The outline always one key away.