FerroCella: Difference between revisions
Jump to navigation
Jump to search
AdminIsidore (talk | contribs) No edit summary |
AdminIsidore (talk | contribs) |
||
| Line 31: | Line 31: | ||
and fluid (Navier-Stokes) solvers within the simulation/core.py MultiphysicsFerrocella class. | and fluid (Navier-Stokes) solvers within the simulation/core.py MultiphysicsFerrocella class. | ||
The CPU-based approach on local droplets has been proven non-viable; stick with the Colab GPU architecture. | The CPU-based approach on local droplets has been proven non-viable; stick with the Colab GPU architecture. | ||
<!-- === Branch: display-v1 === --> | |||
<!-- METADATA: This branch is forked from main at the completion of Phase 1. --> | |||
<!-- STATUS: The core physics engine (simulation/core.py) and the AetherOS API server (run_realtime_server.py) are functionally complete. The server exposes a `/api/get_state` endpoint that returns physically accurate, unnormalized floating-point data for the 1000x1000 grid, along with the corresponding SEXTET values. --> | |||
<!-- OBJECTIVE: Your mission on this branch is to create a visually appealing, but not necessarily physically accurate, web-based viewer. The primary user of the physically accurate data is the AetherOS AI; your user is a human observer. --> | |||
<!-- RECOMMENDED ACTION: Create a new, separate server script (e.g., `run_dashboard_server.py`). This server will act as a *client* to the main AetherOS API. It will fetch the raw data, perform aggressive contrast stretching and normalization (e.g., mapping the 1st and 99th percentiles to black and white), colorize it, and stream the resulting "pretty images" to a web dashboard using SocketIO. Consider using a library like `three.js` or `p5.js` on the frontend for more advanced rendering effects. This approach completely decouples the scientific simulation from the public-facing visualization. --> | |||
{{AetherOS_Navigation}} | {{AetherOS_Navigation}} | ||