Skip to content

Improve Star Wars Trilogy lightgun and multi-mouse input - #3

Open
Zer0one wants to merge 2 commits into
libretro:masterfrom
Zer0one:supermodel-modern
Open

Improve Star Wars Trilogy lightgun and multi-mouse input#3
Zer0one wants to merge 2 commits into
libretro:masterfrom
Zer0one:supermodel-modern

Conversation

@Zer0one

@Zer0one Zer0one commented Aug 17, 2026

Copy link
Copy Markdown
Contributor

Summary

Follow-up to #2. This PR improves the dedicated Star Wars Trilogy Arcade
control profile without changing the profiles used by ordinary gun games.

  • exposes the same input-source choices available to gun profiles: Standard,
    Lightgun Only, Mouse + Analog Stick, Mouse Only, and Analog Stick Only;
  • translates RetroLightgun absolute coordinates, RetroMouse relative motion,
    and the RetroPad left stick into the game's absolute self-centering yoke;
  • keeps source ownership deterministic in Standard mode instead of mixing
    simultaneous values;
  • maps the cabinet's duplicated Trigger/Event pair to the second Mouse or
    Lightgun port, leaving each device's extra buttons available for common
    Start/Coin inputs;
  • applies the optional Upright cabinet X-axis correction after every input
    source converges on the virtual yoke;
  • updates the control-profile documentation and runtime input descriptors.

Validation

The exact current head (dbd7b7d) passed the complete four-platform
Libretro Core CI matrix:

  • Linux x86_64;
  • Windows x86_64;
  • macOS x86_64;
  • macOS ARM64.

Real-ROM testing covered Star Wars Trilogy Arcade on RetroArch/macOS ARM64 and
Batocera 43.1 x86_64. The five source modes, mouse and analog yoke motion,
Lightgun behavior, Upright X-axis correction, separate Mouse 1/Mouse 2
Trigger/Event inputs, and Coin Chute 1/2 through two physical mice were
verified. The Batocera test used only frontend-side device routing; the core
continues to consume the standard Libretro Mouse and Lightgun APIs.

AI-assisted development disclosure

As with the post-sgiannop integration documented in #2, this follow-up was
developed as an AI-assisted study using OpenAI GPT-5.6 Sol through Codex. The
human maintainer defined the behavior, supplied hardware and real-ROM testing,
evaluated the results, and approved the implementation. The AI system performed
source analysis, implementation, debugging, documentation, and automated
validation. Independent human review is still required; passing CI and runtime
tests should not be treated as proof of correctness.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant