FAQ

What are the advantages of using this over git for an AI brain?

Git is designed for code merging. Demarkus is designed for knowledge.

Git solves branch/merge problems you don't have with documentation. For a knowledge base, those features create friction — demarkus eliminates it:

No merge conflicts — Every write creates a new immutable version (v1 → v2 → v3). No branching, no divergent histories, no reconciliation. With git, two people documenting the same function means a merge conflict. With demarkus, it just works.
Content-addressed discovery — Demarkus is building federation with hash-based fetch. Instead of "where is the context for function X?", you ask "what's documented about hash-of-function-X?" — across multiple servers, without central git coordination.
Graph-based context — The persistent graph enables backlinks: "show me everything that references this function." Pure git gives you git grep and hope. Demarkus builds a queryable, navigable graph.
Distributed by design — Git assumes everyone pulls the same repo. Demarkus assumes servers — your docs, your team's docs, public docs — federate and discover each other via content hashes.

Making per-function knowledge available to everyone is exactly what demarkus is built for:

Document per function — write a living doc for each function, link related context, and find it via backlinks and graph traversal — not file paths that move
Link related context via backlinks — what tests use this? what docs mention it? what bugs are related?
Fetch by content hash — anyone can ask "show me everything about function-X" without knowing where it lives
Mirror across servers — if someone on a different team documents a function you use, federation discovers it automatically

Pure git can't do this. It's centralized (one repo) and unindexed (text search only).

Git is designed for code merging. Demarkus is designed for knowledge.

Git solves branch/merge problems you don't have with documentation. For a knowledge base, those features create friction — demarkus eliminates it:

No merge conflicts — Every write creates a new immutable version (v1 → v2 → v3). No branching, no divergent histories, no reconciliation. With git, two people documenting the same function means a merge conflict. With demarkus, it just works.
Content-addressed discovery — Demarkus is building federation with hash-based fetch. Instead of "where is the context for function X?", you ask "what's documented about hash-of-function-X?" — across multiple servers, without central git coordination.
Graph-based context — The persistent graph enables backlinks: "show me everything that references this function." Pure git gives you git grep and hope. Demarkus builds a queryable, navigable graph.
Distributed by design — Git assumes everyone pulls the same repo. Demarkus assumes servers — your docs, your team's docs, public docs — federate and discover each other via content hashes.

Making per-function knowledge available to everyone is exactly what demarkus is built for:

Document per function — write a living doc for each function, link related context, and find it via backlinks and graph traversal — not file paths that move
Link related context via backlinks — what tests use this? what docs mention it? what bugs are related?
Fetch by content hash — anyone can ask "show me everything about function-X" without knowing where it lives
Mirror across servers — if someone on a different team documents a function you use, federation discovers it automatically

Pure git can't do this. It's centralized (one repo) and unindexed (text search only).