fix(#117): adopt Knex migrations as single schema authority#118
Open
vrkothekar wants to merge 9 commits into
Open
fix(#117): adopt Knex migrations as single schema authority#118vrkothekar wants to merge 9 commits into
vrkothekar wants to merge 9 commits into
Conversation
Adds a banner near the top of the README pointing to the published Whitepaper at theailab.org/whitepaper, and updates the Links table with the canonical Whitepaper URL, the PDF download URL, and the errata channel. Also corrects the Protocol Spec link to v5.0. The Whitepaper is the canonical public specification of the TIP Protocol. The Protocol Spec (CTID v5.0) remains the technical reference for endpoint contracts and message formats. License: Whitepaper is CC BY 4.0. Implementations remain under TIPCL-1.0 with conversion to Apache 2.0 on Jan 1, 2031.
…inks table
The TIP Protocol Whitepaper, Version 1.0 is now permanently archived
on Zenodo (operated by CERN) with a citable DOI. This README update
surfaces the DOI as a discoverable badge at the top of the page,
adds the author's ORCID iD alongside the Wikidata QID, and adds
three rows to the Links table for the Version DOI, the Concept
DOI, and the Zenodo record URL.
The full citation in the README banner now uses the canonical
Zenodo title ("Trust Identity Protocol (TIP): An Open Standard for
Verified Human Identity and Content Provenance on the Internet")
and ends with the DOI URL, so anyone clicking the citation can
resolve to the permanent record.
The DOI badge surfaces in the GitHub repository preview and in any
README mirror.
Lifts the entire _ensureSchema() body from knex-adapter.js into a proper initial migration. All 24 tables created in dependency order; platform_links uses its final schema (unlinked_at/unlink_tx_id) with the hotfix alterTable block omitted. down() drops in reverse order.
…eline client-conditional
…n-pg servers knexfile.js now mirrors knex-adapter.js connection logic: reads DB_DRIVER / DB_HOST / DB_NAME / DB_* env vars (same as the app) and exports five named environments (sqlite, pg, mariadb, mssql, oracle) plus a `development` default that follows DB_DRIVER. TIP_PG_* vars removed. 000_baseline.js: ctid column condition changed from `=== "pg"` to `=== "better-sqlite3"`. Only the SQLite path (dag.js SQLiteStore) uses the column name `ctid`; knex-adapter always references `tip_ctid` for every server-side DB, so mariadb/mssql/oracledb must also get `tip_ctid` or every content/prescan query fails on those engines.
…SSQL
entity_id was t.text() → Oracle maps this to CLOB (ORA-02329: LOB cannot
be primary key) and MSSQL maps it to nvarchar(max) (cannot be part of a
PRIMARY KEY constraint). Changed to t.string("entity_id", 128) — TIP IDs
are ≤30 chars so 128 is ample.
Also remove idx_entity_keys_time: it duplicates the PK column list exactly,
which Oracle rejects with ORA-01408 (column list already indexed).
Add docker-compose.mixed-db.yml + scripts/start-mixed-db.sh +
scripts/verify-mixed-db.sh — a 5-engine integration test cluster
(Postgres/MariaDB/Oracle/MSSQL/SQLite) verified at 21/21 checks,
all nodes advancing at identical consensus rounds with spread=0.
Tested with DB_PORT: "3306" fix for node2 (MariaDB): node-2.env carried
DB_PORT=5432 from postgres-backed node1 registration; the compose env
override was missing.
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
2 participants
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.
Summary
Closes #117.
Replaces the dual inline-schema approach (
dag.jsCREATE TABLE blocks +knex-adapter.js_ensureSchema()) with a single Knex migration file as the authoritative schema source.node/src/db/migrations/000_baseline.js— 24-table baseline migration, replaces allCREATE TABLE IF NOT EXISTSand ad-hocALTER TABLEguards from both stores. Column names are client-conditional where needed (ctidon SQLite /tip_ctidon all server-side DBs).node/knexfile.js— full 5-driver config (better-sqlite3,pg,mysql2,mssql,oracledb), mirrors env vars fromknex-adapter.js.node/src/db/knex-adapter.js—_ensureSchema()removed; replaced withknex.migrate.latest().node/src/dag.js— inlineCREATE TABLEblock andALTER TABLEbackfill guards removed; SQLite schema now driven by the same Knex migration file..gitignore— local mixed-DB test artifacts excluded.Oracle / MSSQL compatibility fixes (found during local integration test)
ORA-02329:entity_id TEXTin composite PK → Oracle mapst.text()to CLOB, which cannot be a primary keyt.string("entity_id", 128)— sufficient for all TIP IDs (≤30 chars)ORA-01408:idx_entity_keys_timeduplicates PK columns(entity_type, entity_id, valid_from_ts)exactlynvarchar(max)cannot be part of aPRIMARY KEYconstraintt.string("entity_id", 128)resolves both enginesDB_PORTenv override added to compose (node-2.env carried the postgres port from registration)Verification
Tested locally with a 5-engine cluster (Postgres / MariaDB / Oracle Free 23ai / SQL Server 2022 / SQLite) — all nodes connected to their respective engines, migrations applied, 26 tables created on each,
tip_ctid/ctidcolumn name correct per engine, all 5 nodes advancing in consensus at identical rounds (spread = 0).Test plan
knex_migrationstable created and schema matches prior hand-written schemactidcolumn (nottip_ctid) incontent,prescan_reviews,prescan_jobsknex.migrate.latest()is a no-op when migration is already appliednpm testinnode/— all existing tests pass