## 4. Required changes in the two source repositories
Prefer server-side cleanliness over Manager hacks. Proposed (each is an issue/PR on the owner's side, none is
made without approval):
| # | Repo | Change | Why |
|---|------|--------|-----|
| S1 | mod-coa-playerbots | Master switch `CoaBots.Enable` (0 = module registered but inert; no login/hooks/commands) | Lets a **single** worldserver ship for everyone. "Disable bots" becomes a config flip with restart, not a binary swap; drops the 2-binary matrix and most of §D8 risk. |
| S2 | mod-coa-playerbots | Commit **all** core patches in the core repo; delete "uncommitted working-tree" state from docs | CI must build from clean SHAs (R1). |
| S3 | mod-coa-playerbots | Real Releases + `dist/release.json` replaced by the shared `manifest.json` schema; `TalentBuildsPath` default = `Core/reference/…`, remove the hard-coded dev path fallback | Manager and installer share one truth; no absolute dev paths. |
| S4 | mod-coa-playerbots | Ship a machine-readable config schema (`conf/…schema.json`) next to `.dist`, or generate it in CI from the `.dist` comments | Avoid schema drift (71 vs 30 keys today). |
| S5 | core | Windows build+release workflow (or the build repo builds it); publish `worldserver --version --json` (core commit, bots commit, DB schema id) | Reliable version detection instead of log scraping. |
| S6 | core | Health endpoint alternative: worldserver already prints `ready...`; add a stable machine line (`COA-READY {...}`) or a tiny loopback status file `Core\Logs\ready.json` | Health without regex on human text. |
| S7 | core/repack | Ship a `manager-hooks.json` in the repack describing service names, ports, exes, config templates, RA settings | Layout detection without hard-coded knowledge; unknown custom builds can opt in. |
| S8 | core | Keep the auto DB updater disabled but expose `--dry-run-updates` listing pending updates | Manager migration UI without re-implementing the updater. |
| S9 | repack | Per-download random DB passwords and `Settings\database.json` generated on first start, not packaged | Security (§1.1). Manager does it itself for new installs meanwhile. |
Everything still works without S1–S9 (Manager has fallbacks), they just remove fragility.
| 2 | Read-only scan/classify/import; process discovery; Start/Stop via delegate driver; health | Import of `C:\games\CoA-Repack` leaves it byte-identical (proved by snapshot diff); Start/Stop verified on **fixture copy, offset ports** |
- [Screenshot gallery](SCREENSHOTS.md) — explore the interface and its main features.
- [Host your own server](HOST-A-SERVER.md) — install a server, invite friends and publish a realm listing.
- [Portable characters · Experimental](PORTABLE-CHARACTERS.md) — save an existing character and use compatible realms.
- [Wildcard](WILDCARD.md) — switch realm profiles and manage separate characters.
- [Warcraft Reborn](WARCRAFT-REBORN.md) — create and switch to a Warcraft Reborn realm.
- [Updates, backups and repair](REPAIR_AND_REALMS.md) — keep your server updated and recover from problems.
**Portable characters are experimental.** The Manager is in active development and bugs may still occur. Please [report problems](https://github.com/Corfirean/coa-server-manager/issues/new/choose).
[Download the Manager](https://github.com/Corfirean/coa-server-manager/releases/latest) · [Manager changelog](../CHANGELOG.md) · [Server changelog](https://github.com/Corfirean/coa-server-build/blob/main/CHANGELOG.md)
Settings includes **Check files and database**, **Repair**, and **Start CoA and Wildcard together**.
## Update your server
Database checks inspect both realm schemas when Wildcard exists, the migration ledger, character-save columns and CoA starting class data. Results are saved in the installation metadata's `logs/database-checks.json` and included in exported diagnostics. Hash checks skip mutable MySQL data and configuration files.
Select your server in the Manager. When an update is available, choose **UPDATE** and follow the prompts. Keep the Manager open until it finishes: it stops the server, creates a recovery point and checks the updated server.
Repair requires both worlds and auth to be stopped, no unfinished update, and the installed version to match the currently published signed update. The signed base must match the installed base as well. It verifies packages before changing files, creates a full database/config backup, saves every replaced file and a recovery inventory under `<server>.manager/repairs/<backup-id>/`, then restores official files and applies pending migrations to both worlds. Settings and player databases are not replaced. A failure during file replacement restores the original files. Interrupted repair can be retried; its file inventory and backups remain available.
Updates preserve accounts, characters and settings and cover every realm you have created. If an update fails, read the displayed error before trying again; recovery information remains available in the Manager.
Repair reports unresolved schema damage. It does not guess ALTER statements or replay already applied SQL: those cases need a new corrective migration. Applied SQL whose recorded SHA-256 changed is reported and rejected. Manager failure records take precedence over the core's old `updates` entries.
## Keep a backup
## Release schema contracts
Use **Backups** to save characters, accounts and settings before important changes. Full backups also include world data. Check the realm recorded in a recovery point before restoring it, and stop the server before a restore.
`clean-base` clears the character migration ledger after rebuilding the character database and captures `Scripts/database-schema.json`. It contains expected table/column names and column types. Installation and migration validation inspect this contract; older packages receive the narrower character-save checks, clearly labelled in the UI.
Portable character saved versions do not replace server backups. **Portable characters are experimental.**
For each update, capture a fresh contract **from a disposable database already migrated to the exact packaged core**, into the release tree before `pack-update`:
2. Open **Settings** and run **Check files and database**.
3. Review the results. If official server files need restoring, use **Repair** and follow the prompts.
The first command verifies every entry in the signed base archive and extracts only the runtime needed for a disposable database. The second applies the release SQL with temporary ports, checks critical feature tables independently of migration history, and captures the resulting schema. It requires the disposable fixture marker and never starts a world server.
Repair creates a recovery backup and restores official files while preserving settings and player databases. It may require the installed version to match the currently published signed package. Missing player data requires a backup; some database problems require a corrective server update.
Base and update packaging require a complete contract for all three databases. Cumulative updates always carry it, even when unchanged. Signing and verification check the archive contents and contract before publication. Never capture it from a user's server or from an unmigrated previous release database.
If an operation stays blocked or fails again, use **Report a problem** and include diagnostics and the error message. Do not delete database files or recovery records to clear an error.
Corrective SQL under `repairs/{auth,characters,world}` receives new `manager_repair__` migration IDs and runs after historical SQL. Missing Wildcard tables can therefore be restored without replaying historical data deletion. Existing malformed tables still fail schema checks; creating a missing table does not recover its lost rows.
## Run CoA and Wildcard together
## Simultaneous worlds
In **Settings**, enable **Start CoA and Wildcard together**. Both realms share accounts but keep separate characters and world data. Both statuses appear on Overview, and Stop shuts down both.
On Windows the checkbox creates Wildcard when needed and saves distinct secondary world/RA ports. Both worlds start with one shared authserver/MySQL and separate world/character databases. The second world has its own executable copy, config, supervisor, logs and report spool in `.realms/secondary`; Data is read through the primary installation's absolute path. This runtime folder is excluded from packages.
If friends networking was configured before enabling the second realm, run its setup again so both realms are reachable. Disable simultaneous startup before switching to Warcraft Reborn.
The selected realm remains the primary world used by the Manager's console, companions, settings and Play. Selecting another realm still stops/restarts both to preserve that relationship. CoA companion restrictions on Wildcard remain in force. Both world statuses are shown on Overview.
Stop shuts down the secondary supervisor/world before the primary launcher may stop the shared database. Database restores, updates and profile changes reject a running secondary world. Each start refreshes its executable/config copies, so server updates cannot leave an old secondary binary behind. The secondary world inherits the primary network bind setting; its RA console remains on loopback. Enabling friends networking opens/maps both world ports. If networking was enabled before the second realm, enable it again to add the second port's rule/mapping.
The first checks both processes, ports, shared services and graceful shutdown. The second uses a test-only signing key, corrupts a fixture file, applies SQL to both worlds, verifies backups and account/character counts, and drops a column only in Wildcard to prove schema checks are isolated. These probes modify the fixture; never use a real server.
changelog validation and whitespace checks are recorded in the accompanying
local logs and PR checks.
- Real bundled MySQL regression validates restored rows, users, procedures,
functions, views and triggers, and preservation of custom defaults. It also
confirms SQUID pending imports are reported without modifying their records.
- Published Content Scaling revision e5d8c675 passes the production-budget
regression for 40/100 HP, core multipliers, rank rates, groups and repeated
calculation. This is a code regression, not an in-game Wildcard/PTR comparison.
## Outstanding evidence and release gates
| Report | Current evidence | Next verification |
| --- | --- | --- |
| Manager #98 | Repair journal defect reproduced; report itself has no diagnostic ZIP | Reproduce the complete Custom Races workflow and verify existing affected journal recovery |
| Screenshot: SQUID Start fails but Repair reports zero | Separate SQUID ledger was absent from Repair checks | Retest with a copy of the affected database; recover from a verified pre-import backup |
| Manager #82 / screenshot lock messages | #90 and #91 are already in published 0.6.16; lock regressions pass | Distinguish an active slow check from abandoned maintenance using diagnostics |
| Manager #100 / 54% | Attached archive ends immediately after entering private rehearsal; no failure is captured | Measure the full private-copy run on the affected hardware; new progress identifies its stage |
| Screenshot: stopped-MySQL error | Broken-template shutdown and lost-record/supervisor failures reproduced and fixed on disposable real MySQL | Verify the affected installation has one of these causes; other failures remain diagnostic-specific |
| Manager #25 | Missing defaults have targeted repair; start-data incompatibilities remain separate | Retest imported schema and affected race/class data without suppressing incompatibility checks |
| Manager #54 | Schedule fix is published; Stop now tolerates broken templates and Repair has complete MySQL recovery | Verify the exact unresolved placeholder and original recovery transaction; automatic placeholder guesses are not introduced |
| Scaling screenshot | 2.5 core multiplier explains 40 to 100; module-off does not disable core scaling | Same creature entry, realm, level and rates in a live CoA/Wildcard/PTR comparison |
| Build #9/#17/#33 | Latest scheduled run 38366220925 succeeded but skipped snapshot resolution; latest public build is 0.261013.54 | Qualify a new integrated candidate; old issue state is not current release evidence |
This batch does not publish a Manager installer, promote a server package,
close user reports or claim the complete fresh-install/update/gameplay matrix.
Clean Windows installs now select the full image attached to the same signed stable release used by updates. They never fall back to the legacy base when the current full image is absent. Explicit offline/custom full packages remain supported, while Repair and release upgrade fixtures retain the legacy base source. The release worker composes and validates the complete current database image before publication; users do not run an intermediate base server or an update transaction.
The Manager selects the latest signed package directly for any known older server version. The source-version list records release qualification; it is not an exclusive allowlist and does not prove that a particular installed database can apply the package. Unknown imported versions, downgrades, incompatible platforms and newer Manager requirements remain blocked.
The updater verifies the signed manifest and archive, checks file changes and database history, stops the server, and saves one complete recovery point. It then replaces files, applies pending SQL once, verifies the target schema and starts the updated server once for validation. Shared authentication migrations run once; both realm histories are checked before changing either. Account and character identities are compared before and after migration and startup.
Account migration history is stored in the shared account database. Read-only checks recognize legacy auth records from both realms. Write operations adopt existing records and preserve already-applied core history; unknown legacy checksums stay unknown. History is read in two batched client calls, with explicit physical schemas and consistent text collations. Use the current Manager for subsequent database operations.
Applied migrations are skipped; changed checksums and interrupted/failed records stop the update. Duplicate SQL identifiers for the same schema are rejected before execution. SQL or startup failure restores files and databases together. If recovery fails, startup remains blocked and the recovery artifacts remain available.
The per-update nested server rehearsal and intermediate-package routing have been removed. Full-copy upgrade/startup testing belongs to release qualification, on independent disposable installations. Cleanup of interrupted rehearsals from older Manager versions remains for compatibility. Signed packages, recovery verification, configuration merge and startup validation stay enabled. No migration guard is bypassed and no unexecuted SQL is marked applied. Retired source archives do not prevent a direct update, but the database history must still be consistent and target checks must pass.
## Verify with an actual installation copy
`cargo run -p coa-core --example upgrade_fixture -- <copy-root> <signed-package-directory>` exercises the production updater, including recovery copies and startup. It refuses mutation unless the directory name starts with `coa-schema-fixture-real` and contains `.coa-update-fixture`. Prepare a disposable copy from a verified recovery point, with separate ports, metadata and no client path. Never point it at a working installation.
`migration_plan <copy-root> <signed-manifest>` reads database history without creating a migration ledger. On a separate marked `update-database-diagnosis-*` copy, `--apply <migration-directory>` exercises the production database phase. `--sql-only <migration-directory>` verifies and applies the signed SQL list for both realms and compares account/character identities. SQL-only success is not full package/startup qualification and does not verify the target schema contract.
Record package signatures, exact source recovery point, actual result, repeat-run behavior and unavailable stages. Do not label a route universally tested based on a version list or SQL-only success.
@@ -5,28 +5,12 @@ installed worldserver must contain Warcraft Reborn support; an old build is
refused before any databases are created. Stop the client before switching to
have its saved realm updated immediately.
The profile uses `CoA.ClassModel = "wcr"`, game-mode mask `0`, live realm type,
and realm ID `3`. Accounts remain in the shared `acore_auth` database. World data
is initially copied from CoA into `acore_world_reborn`; the structure of
`acore_characters` is cloned into `acore_characters_reborn` without copying
players. Existing database names or realm ID 3 block creation.
Accounts are shared with your other local realms. Warcraft Reborn starts with separate characters; switching profiles does not copy existing characters or progress.
Settings are stored in `Settings/realm-profiles/reborn.json`. Creating a profile
from Wildcard uses the saved CoA settings as its base, preventing inheritance of
Wildcard gameplay flags. Both bot systems are disabled and cannot be enabled
through this profile; their class mechanics have not been validated. Other
optional modules are marked experimental.
Bots are unavailable on Warcraft Reborn. Other optional modules are experimental. Portable character transfers into Warcraft Reborn are not supported. **Portable characters remain experimental.**
Backups, updates and database checks include every created realm. Restore keeps
using the realm recorded in the recovery point. The Windows launcher is upgraded
from the Manager's previous two-realm patch and routes Warcraft Reborn to its
own databases. Docker uses the same realm schema mapping.
Backups and updates include every created realm. Check the realm recorded in a recovery point before restoring it.
Simultaneous startup continues to support CoA + Wildcard only. Disable that
option before switching to Warcraft Reborn; it cannot be enabled while Warcraft
Reborn is selected.
Simultaneous startup supports CoA + Wildcard only. Disable **Start CoA and Wildcard together** in Settings before switching to Warcraft Reborn.
Automated checks cover profile configuration and restoration, old state files,
binary support detection, database routing, all-realm enumeration, client realm
selection, and upgrading/reversing the launcher patch. These checks do not
constitute a gameplay validation of Warcraft Reborn or optional modules.
-`cargo test --workspace`: 178 core tests passed; command layer and examples compiled.
-`npm run build`, translation key checks and changelog validation passed.
- Native disposable fixture on offset ports: fresh Wildcard creation and startup, CoA startup after switching back,
separate probe values and restored CoA configuration. A second fresh creation verified the arena-season seed fix.
- Full backup verified both worlds, both character databases, shared auth and configurations.
- Native migration probe applied once to each realm and once to auth; repeating it did nothing.
- Restoring Wildcard characters removed a fixture-only change and preserved the CoA probe value.
- Core gameplay wrapper: `VERIFY ALL: PASSED`, gameplay stage only, real clock. First login (7 assertions),
level rolls/reroll (31), skill card rolls (13), stat path (12). The level-roll and skill-card scenarios passed on isolated
reruns after failing in the shared batch; the wrapper records two batch-sensitive cases.
- Browser demo verified active profile, hidden Bots navigation and module compatibility labels. Screenshots show
mock IPC; native checks above exercised the actual backend and server.
The gameplay harness required fixture-only synchronization of already applied migration history by verified SQL
hashes and its normal Windows module config path. Owned server files and databases were not modified.
[All guides](README.md)
Reference in new issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.