Tagged: Architecture
System design, architecture decisions, and data flow
Pages with this tag
Obsidian app internals — research needed
37High-level logic flow, rendering loops, and constraints of the Obsidian app that will limit Crosswalker's plugin capabilities.
Volatility property and entity registry
38Introducing formal content volatility tracking and a registry of institutional actors — building toward the marketplace.
Homepage navigation and site identity decision
39Should the docs sidebar be accessible from the homepage? Is the docs site also the marketing site?
Ontology evolution: first principles revisited
40Breaking down the ontology evolution problem from data structure primitives, not organizational conventions. Where we are, what we're deciding, and the paths forward.
Research: atomic operations on graphs — are the 13 primitives primitive enough?
41Literature survey of graph edit distance, ontology diff, tree-diff, schema evolution, and category theory to determine the provably complete set of atomic operations for describing structural changes.
How primitive are the primitives? Pluggable layers and open questions
42Questioning whether the 13 structural primitives are fundamental enough, exploring detection as pluggable logic, taxonomy-over-taxonomies for decisioning, custom transfer logic, and graph vs. hierarchy scope.
User-first ontology maintenance: what does the experience look like?
43Starting from the user (human or agent) and working backwards — what does it feel like to keep ontologies up to date? What does the system actually DO when frameworks change?
Evidence-link edge model synthesis — junction notes as the committed architecture
44Distilling 2 research sessions directly targeted at Challenge 07 (the link metadata / edge model for evidence mapping). Both converge strongly on junction notes (edge-as-note reification) as the only architecture compatible with Obsidian Bases' flat-YAML constraint, with a concrete 13-field schema validated against 8 GRC tools and OSCAL round-trip.
Foundation research synthesis — testing our decisions against 4 parallel research sessions
45Distilling 4 parallel research sessions against Foundation-phase decisions. Convergences, challenges, new formal concepts, and the decisions this forces — with visual anchors for the math.
SEACOW + folder-tag-sync — prior-art the project owner built before Crosswalker
46Documenting two of the project owner's existing tools (SEACOW meta-framework in cyberbase, obsidian-folder-tag-sync plugin) as Foundation-phase prior-art that should inform Crosswalker's hierarchy-vs-graph decisions — and flagging a formal reuse investigation as a research item.
Direction third-wave shifts — Tier 3 default flipped, Tier 2 confirmed-and-extended, audit-trail 4-tier model adopted
47Third research wave (Ch 14/15/16) landed 2026-05-02 and produced three substantively directional shifts. This log captures the deltas vs the same-day TL;DR commitments log. Ch 14 confirms the Ch 11 layered Tier 2 stack as production and adds Tier 2-Lite (sqlite-wasm) plus Comunica federation; Ch 16 flips the Tier 3 default from Apache AGE to Apache Jena Fuseki / oxigraph-server with layered Fuseki+DuckDB as power-user upgrade; Ch 15 adopts a 4-tier audit-trail model (T0/T1/T2/T3) with OpenTimestamps as the new T2 default and the Ch 08+13 git stack repositioned as one of three T3 options.
Two-mode import architecture decision — bundled projector + direct emission, both first-class
48Decision: Crosswalker supports two architectural modes for producing Tier 1 vaults — Mode 1 (hand structured rows to bundled engine) and Mode 2 (bypass engine, emit Tier 1 Markdown directly). Both are first-class. ChunkyCSV/JSONaut compose naturally with Mode 1. Streaming is at the engine boundary so huge inputs work. Supersedes earlier 'three producer paths' framing.