Repository navigation
Retire Neo4j and narrow the AgensGraph 0.7 claim - #189
Merged
Merged
Conversation
Move the Neo4j cabinet onto the leftover archive shelf and bind the factory to fake|agensgraph. Ignore MEMNET_NEO4J_* so a host that still exports it starts. The 0.7 row is adapter plus operator round trip with no runtime caller. Closes #187 Co-authored-by: chouswei <chouswei@users.noreply.github.com>
chouswei
added a commit
that referenced
this pull request
Sep 25, 2026
Hatch, project.toml, and memnet.__version__ move to 0.19.15 so this tip can upload. Changelog records storage role labels, Path-B ingest kinds, and Neo4j retired. Last published PyPI stays memnet-llm==0.19.14 until upload. Co-authored-by: Cursor Agent <cursoragent@cursor.com> Co-authored-by: chouswei <chouswei@users.noreply.github.com>
This was referenced Sep 27, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Closes #187
Neo4j had no runtime caller (agentReachable=false; DurableBuffer.hydrateFlushCallers="tests").
At 80, 229, and 504 nodes it is not faster on an equal query. JVM startup is about 7 s and 811 MiB, against 0.87 s and 33 MiB for MemNet.
Its flush-then-hydrate round trip did not return the same node set it flushed.
Change
SysML first, then code. Neo4jAdapter, Neo4jLibraryPort, and liveNeo4jClaimed move to the leftover archive shelf (
MemNetArchive, off ProjectMemNet load). FactorybindsExactlyOneOfisfake|agensgraph. The 0.7 row is adapter + operator round trip (no runtime caller). No new hydrate/flush caller. No package version bump.MEMNET_NEO4J_*is ignored with one log warning. The engine starts.Removed:
parts/common/memnet/memnet/durable/neo4j.pyparts/common/memnet/memnet/durable/neo4j_library.pytests/test_neo4j_library.pyAlso dropped the
neo4joptional extra and theneo4j_livemarker.Checks
ruff checkandruff format --checkonparts/common/memnet,parts/memnet-mcp/software, andtestspytest: 403 passed, 1 skipped (agensgraph_live)