Melee PC asks you to bring a Melee disc image and accept a testing beta. The native port adds PC settings and customization around the game, but the release warns of crashes and missing features. I'd approach it as a project to help test, especially if your usual play sessions depend on online matches or particular controller hardware.
That recommendation has a firm limit: our source setup stopped before we could play. The feature judgments below come from the project's own status table and release notes, with the failed sandbox attempt kept separate.
What happened when we ran it
Our Melee PC install step failed with exit 1 after 48 seconds. We ran commit e833835 on October 3, 2026, with a recorded configuration of 3 CPUs, 8 GB RAM and image lab-cpp:1. The image label doesn't establish the actual compiler, runtime or operating-system versions.
The final output came from SDL's CMake configuration. It linked to SDL's guidance for configurations without X11 or Wayland support, then ended with configuration incomplete. It also listed dummy and v4l2 camera drivers and pointed to CMakeOutput.log and CMakeError.log. The supplied tail doesn't identify a particular missing package, so prescribing one would go beyond the evidence.
We have no completed build or test result from this attempt. It establishes that our recorded source setup failed at configuration, without establishing that the game itself is broken. The checkout contained 2,919 files, approximately 703,401 source lines and 36.6 MB. The scan found two CI workflow files and a tests directory, with no Dockerfile detected. None of that proves playable operation.
Release downloads are a separate route. The v0.4.2 release, published October 21, came after our October 3 attempt. We didn't evaluate those packages and can't transfer the earlier configuration result to them.
Your disc supplies the game data
The README describes a native port built from decompiled Melee code, Aurora's compatibility layer and SDL3. USA revision 2, NTSC-U 1.02 with disc identifier GALE01, is the supported disc. No game data ships with the project.
PAL support is experimental. Its status row describes USA game code running against PAL data, with trophy tables stubbed out and no reference hash for disc verification. If PAL behavior is your main reason for trying the port, those limits matter more than a screenshot of a running match.
The port code is GPL-3.0-or-later, but the decompiled game code is not licensed. The README explicitly says the repository as a whole isn't distributable under the GPL. For a developer looking for reusable open-source game code, the distinction changes what this repository can offer. Check the relevant component rather than treating the repository as one uniformly licensed package.
Customization has specific boundaries
The documented F1 overlay pauses the game beneath its settings. Gamepad remapping and Dolphin-format texture packs give a player reasons to try the native port. Replacement music is supported too, although the status table says tracks decode wholly into RAM and ignore their own loop points.
Widescreen covers VS, Sudden Death and Training. Menus, results and cutscenes keep the original aspect, and the wide HUD has a separate toggle. Someone expecting every screen to fill a wide monitor would be disappointed. These are documented boundaries, rather than defects discovered during our run.
The official GameCube adapter path is another reason to read past the feature list. It is implemented but not yet confirmed against a physical adapter. I wouldn't choose this beta for a controller-dependent session without checking the actual adapter first.
Implemented online play still needs acceptance checks
The README lists LAN, internet friend codes, Unranked and ranked sets while marking online play partial. Broader two-NAT and live ranked acceptance remain pending. Its netplay instructions require matching builds and game images, so coordinating a session includes coordinating those inputs.
The platform matrix distinguishes Android match reports from device rollback gameplay, which it still calls unproven. That distinction is useful: reaching a match and verifying rollback behavior answer different questions. Our failed setup adds no independent networking evidence. If dependable online sessions are the requirement, I would evaluate Slippi Launcher first rather than adopt this beta on the strength of implemented menu entries.
Platform labels don't promise verified hardware
The README marks macOS support partial and says Intel hardware remains untested. The Windows Direct3D 11 fallback is also partial, with no working D3D11 device observed and verification outstanding on the older Intel hardware it targets. Linux requires Vulkan 1.1. Check your platform's row before choosing a package.
The browser route has its own restrictions: no online play and no gamepad remapping. An Android graphics report for v0.4.2 and controller-layout reports appear in the recent issue feed. They are user reports, not our reproduced findings or proof of a fix. GitHub records the last repository push as October 20, 2026; issue activity continued into November. That supports neither an abandonment claim nor a promise of prompt support.
Dolphin and Slippi answer different play needs
Dolphin runs GameCube and Wii games through emulation. Evaluate it if you want other games from those libraries alongside Melee. Slippi Launcher handles Slippi Online and replay launching and analysis. Its documented purpose matches a reader seeking that workflow. We have no comparative performance or reliability measurements for either alternative.
For a beta trial, choose the supported disc and read the platform-specific known issues for the exact release. Keep existing saves separate until you've checked a complete play-and-save cycle. If you hit a problem, attach the requested log and identify the disc and platform; those details give a report more value than simply saying the port crashed.

