Make an FPS: the weapon first, then everything else

The temptation in a shooter is an arsenal. Ten weapons that all feel the same is worse than one that feels excellent, and the second is what teaches you how to make the other nine.

What makes a weapon feel like something is a stack of small things firing together: the muzzle flash, the recoil kick and its recovery, the shell, the sound, the crosshair bloom, the hit marker, and what the enemy does when hit. Get those synchronised and a plain rifle is satisfying. Miss the feedback on impact and the best model in the world feels like a water pistol.

Director 1 writes that stack as GDScript in your project. You describe the weapon, fire it, and say what is missing. Usually the answer is the impact, because impact is the part people forget to ask for.

A shooting range proves the weapon works. It does not prove the game does. That needs something that notices you, takes cover, flanks badly enough to be beatable and reacts visibly to being hit.

That behaviour is code, and it gets written for you. What is not code is the encounter: where they come from, how many, what the space gives the player to work with. That is level design and you build it by hand.

The honest measure of an encounter is playing it four times and seeing whether you solve it the same way every time. That is a judgement call and it belongs to you, not to the tool.

Shooter levels are about cover, sightlines and the distance between them. A room where every angle is exposed is a bad room and you can only see that from inside it, at eye height, with the actual weapon range in mind.

So the builder is a direct editor. You place geometry, terrain, buildings and light, then drop into the level and play it. Change something, play it again. That loop is short enough to be worth doing constantly.

Because the game runs locally on the copy of Godot bundled with the app, playing is instant and the framerate you see is the framerate a player would see. In a genre where aiming is the whole interaction, that is not a detail.

Stage Engine is a desktop app for Mac and Windows. You download it and install it. The engine underneath is open source Godot and your project is a real Godot project.

Building needs an internet connection because Director 1 and asset generation come over the network. Playing what you built does not.

The download is free with a free plan. Indie is $10 a month and Pro is $50, and both buy more usage rather than a different engine.

Most of what people call weapon feel is audio and timing rather than visuals. The report of the shot, the tail on it, the distinct sound of a hit landing versus a miss, and all of it arriving on the same frame.

Get the impact sound wrong and a perfectly good weapon feels weak. Get it right and a very simple one feels heavy. This is the cheapest improvement available in a shooter and the most often left until last.

Ask for it explicitly. Hit confirmation is the kind of thing that is easy to describe and easy to forget to describe, and the difference in how the game feels is not subtle.

Shooting for its own sake holds attention for about ten minutes. After that a shooter needs a reason: reach somewhere, hold somewhere, find something, survive a duration.

Objectives are code and they are quick to get working, which makes them easy to try several of. The one that suits your level is often not the one you assumed.

The useful test is whether the objective changes how you move through the space. If you play the same way with it as without it, it is decoration rather than design.