Mea Melone Install ●

[✓] Python 3.11.9 (venv active) [✓] Node 20.12.0 (electron 28.2) [✓] Core (Rust) version 1.4.2 [✓] GPU detection – NVIDIA RTX 3070 (CUDA 12.2) [✓] Sample dataset load – OK [✓] UI launch – OK If any check fails, the console output contains a short (e.g., ERR_PYENV , ERR_GPU_DRIVER ) that you can look up in the Troubleshooting section (below). 6️⃣ Common Pitfalls & Troubleshooting | Symptom | Likely cause | Fix | |---------|--------------|-----| | mea-melone: command not found | PATH not refreshed | Open a new terminal, or run source ~/.bashrc (or ~/.zshrc ). | | Python packages fail to install ( pip errors) | Missing system libs ( libssl-dev , libffi-dev ) | On Ubuntu: sudo apt-get install build-essential libssl-dev libffi-dev python3-dev | | UI stays on the splash screen (Windows) | Incompatible GPU driver | Update NVIDIA driver to the latest R535 series, then reinstall the optional CUDA component via the installer. | | ImportError: libgomp.so.1: cannot open shared object file (Linux) | Missing OpenMP runtime | sudo apt-get install libgomp1 (Debian/Ubuntu) or sudo dnf install libgomp (Fedora). | | Failed to connect to data source (S3) | Wrong credentials or missing awscli | Run aws configure with a valid access key, or install awscli ( pip install awscli ). | | Plugin installation stalls | Proxy/firewall blocking pypi.org | Export HTTPS_PROXY environment variable or use the offline installer ( mea-melone --install-plugin <path-to-wheel> ). | | Crash on startup (macOS) – “dyld: Library not loaded: @rpath/libffi.8.dylib” | Homebrew mismatch | brew reinstall libffi and then re‑run the installer script. |

| Step | What you do | |------|--------------| | | Pick a directory where all analysis projects will live (default: ~/MEAMeloneProjects ). | | 2️⃣ Data source | Connect to one of the supported back‑ends: local folder, S3 bucket, Google Cloud Storage, or a live MQTT stream from field sensors. | | 3️⃣ GPU enable | If a supported GPU is detected, click Enable GPU – the wizard will write CUDA_PATH and install torch‑cuda (or rocm‑torch ). | | 4️⃣ Plugins | Browse the built‑in plugin marketplace (e.g., NDVI‑Extractor , Spectral‑Unmix , ML‑Anomaly ). Click Install ; the wizard resolves Python dependencies in the virtual env. | | 5️⃣ License | Enter your commercial license key (if you have one). A free‑tier key is auto‑generated for evaluation (valid 30 days). | Configuration file – All settings are saved to $HOME/.config/meamelone/config.yaml . You can edit it manually for advanced tweaks (e.g., custom Python interpreter path). 5️⃣ Verifying the Installation Run the self‑test from the command line: mea melone install

# Show GPU details (if enabled) mea-melone --gpu-info [✓] Python 3