tegwick
047a6a2d31
Assistant: codex Assistant-Model: gpt-6.1-sol Assistant-Session: 01a1045b-933d-7aa1-b2db-104557487a79 |
||
|---|---|---|
| activity-definitions | ||
| architectures | ||
| canon | ||
| docs | ||
| feedback | ||
| history | ||
| integrations/activity-core | ||
| patterns | ||
| reference | ||
| registry | ||
| research | ||
| scripts | ||
| sequences | ||
| technologies | ||
| templates | ||
| tests | ||
| workplans | ||
| .custodian-brief.md | ||
| .gitignore | ||
| .repo-classification.yaml | ||
| AGENTS.md | ||
| INTENT.md | ||
| Makefile | ||
| README.md | ||
| REVIEW.md | ||
| SCOPE.md | ||
| START.md | ||
| TREE.md | ||
| WORK-RECORDS.md | ||
frontend-patterns
An evolving research language for frontend architecture and design.
Start with the 0.1 research preview: two decision paths, seventeen draft hypotheses and evidence-qualified implementation mappings. Canonical admission, independent usefulness and the M1/M2/M3 gates remain open.
For ecosystem use, see the reuse and improvement loop and its three-action workplan.
The project treats frontend architecture not as a choice among frameworks or named styles, but as the progressive resolution of recurring forces concerning:
- authority,
- addressability,
- projection,
- interaction,
- mutation,
- synchronization and reconciliation,
- composition,
- execution and delivery,
- quality boundaries.
The repository is intended as an infospace for both human and agentic reference.
Its goal is to make frontend architecture:
- understandable as a set of recurring design problems,
- navigable as a connected language of patterns,
- generative rather than merely taxonomic,
- explicit about trade-offs and failure modes,
- independent of framework fashion,
- grounded in known implementations and sources,
- machine-readable enough for agents to reason over.
Core proposition
Frontend architecture is the progressive resolution of forces concerning authority, addressability, projection, interaction, synchronization, composition, and execution. A frontend pattern language captures recurring resolutions of those forces and the contexts they create for subsequent design decisions.
Repository map
canon/— normative vocabulary and structure of the language.patterns/— individual frontend architecture and design patterns.sequences/— generative sequences for common frontend situations.architectures/— named architectural configurations expressed through patterns.technologies/— mappings from technologies/frameworks to the patterns they implement.research/— research material, source reconciliation, concepts, catalogues, and cases.templates/— authoring templates for new patterns and mappings.
Design principles
Patterns before products
React, Solid, HTMX, Astro, LiveView, Vue, Svelte and similar technologies are treated as implementations and examples, not as the ontology.
Forces before preferences
A pattern should be selected because of the forces present in a context, not because a framework community considers it idiomatic.
Relations before lists
Patterns form a graph. Important relations include:
- requires
- enables
- refines
- generalizes
- complements
- alternative-to
- conflicts-with
- commonly-combined-with
- creates-context-for
- implemented-by
- observed-in
Sequences before checklists
A pattern language should help generate architecture through an ordered sequence of decisions.
Failure modes are part of the pattern
Software patterns must describe when they become harmful, misleading, expensive or fragile.
Status
The corpus currently contains eighteen draft hypotheses, traceable source reviews, bounded runtime experiments and four synthetic/provisional walkthroughs. Independent human validation, canonical admission and demonstrated generativity remain open. The evolution report makes evidence, gaps, learning outcomes and unknowns visible; the method explains how the language continues learning over time.
Research entry point
Start with the research program, its protocol and validation cases. The seed investigation in FEP-WP-0002 is finished with the M1 gate at revise/defer; independent validation of the connected seed corpus continues in FEP-WP-0009. Later phases broaden coverage and evaluate retrieval against real systems.
The research bench now provides executable source
investigations through infospace-bench. Run make bench-test to verify the adapter
and make bench-plan to inspect the synthetic pilot. Source reviews and live pilot receipts are retained under research/.
Run make evolution-check evolution-test to verify evolution records and reporting.