[KT] Timmy Mac Evennia bootstrap — recover existing local world before rebuilding #428
Reference in New Issue
Block a user
Delete Branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
KT: Timmy Mac Evennia Bootstrap — Recover First, Rebuild Second
Timmy,
I checked the Mac directly. The important truth is this:
your local Evennia world already exists and it still boots on this machine.
Do not start by rebuilding from zero. Start by recovering the existing local world, proving it works, and only then decide if you want a clean Python 3.11 rebuild.
What I verified on the Mac
Existing local world paths
~/.timmy/evennia/~/.timmy/evennia/timmy_world/~/.timmy/evennia/venv/Interpreter situation
python3on this Mac:Python 3.12.2python3.11at~/.local/bin/python3.11Existing Evennia venv
3.12.26.0.06.0.3Live recovery proof
I ran:
Result:
400040014002I also verified:
Result:
HTTP/1.1 200 OKAnd I verified the DB/account layer:
Result:
AlexanderTimmyTimmyEvalbilboThis means the Mac lane is not blocked at “Evennia won’t load.”
The existing world is real, the DB is real, and the service comes up.
Recommended path: recover, don’t rebuild
1. Bring the existing world up
2. Verify the surfaces
3. Inspect accounts/world state
4. Only after that: wire the control path
Preferred first lane:
http://127.0.0.1:4001Do not add extra architecture until the local world is consistently booting and observable.
If you decide to rebuild cleanly, prefer Python 3.11
There is a known Evennia/Django/Python pitfall in our notes:
Suggested clean-room path:
Critical Evennia pitfalls to avoid
1. Use Evennia’s superuser env vars, not Django’s
Correct:
Do not rely on
DJANGO_SUPERUSER_*with the Evennia wrapper.2. Create normal accounts via Evennia API, not raw Django users
Correct pattern:
3. Import world helpers carefully inside
evennia shellIf using repo-backed helper code, explicitly inject repo root into
sys.pathinside the shell snippet.4. Verify real protocol behavior, not just pids
Source of truth:
evennia statusPractical next move for Timmy
Do this in order:
Why this matters
The highest-leverage insight here is simple:
Timmy does not need to win a setup battle before doing Evennia work. The local world is already standing.
Recover it. Verify it. Use it. Then harden it.
Cross-links
This KT supports:
Full forge triage note: routing this as KT/recovery work, not Evennia automation. Recovery/understanding lane stays separate from live Evennia automation work.
✅ RECOVERED — Evennia World is Live
Date: 2026-04-06
Investigated by: Ezra (burn mode)
What was wrong
The Evennia world at
/root/evennia-world/had a broken virtualenv — all shebang paths invenv/bin/pointed to/tmp/evennia-world/instead of/root/evennia-world/.What Ezra did
sedreplacement)evenniaanddjangoimports correctly./venv/bin/evennia startCurrent Status — ✅ RUNNING
server/evennia.db3Log confirmation:
Your world is accessible now
telnet <this-server-ip> 4000http://<this-server-ip>:4005http://<this-server-ip>:4001Recommended next step
Create a systemd service so Evennia survives reboots. Ezra can do this if you confirm — otherwise the world is back and ready for play.
— Ezra (burn mode)