Decision log

What was decided and why. Read this before reopening an argument with yourself.

Settled

Decision	Reason
Not voxel	Minecraft's appeal is persistence and self-set goals, not blocks.
No wipes or runs	The whole reason the genre didn't fit. Most indie survival is run-based because it's cheaper.
Fixed terrain, procedural loot	Route planning is only a decision if distances are knowable. Random terrain makes it a reaction.
Vessel is fictional, not a real vehicle	No drivetrain knowledge required. Invented parts, authored failure rules.
Spoilage and cold chain as the core	Drawn from actual dietary work. Nothing in the genre models it.
Cooler degrades, not random failure	Visible decline reads as dread. A surprise roll reads as unfair.
Fuel competes with cargo for hold space	Turns route length into a cost you pay in the thing you drove out to get.
Jerry cans sold at Home	If reaching the best location depends on a loot roll, players get stranded by luck. That's the version people hate.
Upgrades carry between runs	Opens the far map by earning it instead of by tuning a constant.
Prototype in C#, text only	Inventory, timers, quantities, state. Same software already written before. Not an engine problem.

Corrected after checking the math

All 1,956 routes were enumerated. Two things in the first draft did nothing.

Was	Now	Why
Cold life 90–120h	10–18h	No route exceeds 24h, so spoilage never fired and the cooler was decorative.
Unlimited stock per location	6–8 bulk each	One good stop filled the hold. No second stop was ever worth making.
Tank 22 with F reachable	Tank 22, F needs an upgrade	Home to F and back is 24h. It was never reachable. Now that's intentional.

Don't raise cold life without rechecking it against the distance table, and don't change stock without checking that the score curve across stops still flattens.

Deliberately not built yet

Idea	Why it waits
Site exposure / radiation budget	Good mechanic, same decision at a smaller scale. But it makes four clocks and three is already a lot for an untested prototype.
Uncertainty about whether an item is spoiled	Tests a different question than the one on the front page.
Driving as an activity	Content, not a question. Weeks of work testing nothing uncertain.
Repair minigame	Same. Repair stays a menu choice.
3D, terrain, art, combat, story	All of it is more fun to build than the thing being tested. That's the warning sign, not the reason.

Open

The title
The era and setting
Whether fixed terrain runs exploration dry, and how fast