Skip to content

Releases: tin7/aster-view

Aster View 0.4.1

Choose a tag to compare

@tin7 tin7 released this 26 Oct 17:01

Install

  • Python package (Python ≥ 3.10): pip install git+https://github.com/tin7/aster-view, or without git: pip install https://github.com/tin7/aster-view/archive/refs/tags/v0.4.1.zip
  • VS Code extension: download aster-view-0.4.1.vsix below, then Extensions: Install from VSIX… and reload the window.

Fixed

  • The VS Code extension now finds aster-view installed in a virtual environment under ~/venvs (as the README recommends), ~/.venvs, ~/.virtualenvs, ~/venv or ~/.venv, without setting its path by hand.

Changed

  • README: what to do when the installation fails on Ubuntu/Debian.
  • Gallery: a garter spring crushed between two tubes.

Example data

  • crush_completo_damping_a05.med (1.96 GB, uploaded after this release was created): the result file of the garter spring animation in the README. It has 52 time steps with DEPL, SIEF_ELGA, VARI_ELGA and CONT_NOEU. Open it with aster-view crush_completo_damping_a05.med --field SIEF_ELGA --comp VMIS.

Aster View 0.4.0

Choose a tag to compare

@tin7 tin7 released this 26 Oct 12:00

First public release.

Install

  • Python package (Python ≥ 3.10): pip install git+https://github.com/tin7/aster-view
  • VS Code extension: download aster-view-0.4.0.vsix below, then in VS Code run Extensions: Install from VSIX… and reload the window.
  • MCP server for AI assistants: see the README.

What it does

  • Interactive viewer for code_aster .med results, in the browser or a VS Code tab. ParaView-like layout: pipeline (deformed shape, threshold, cut, axial window), properties, light and dark themes.
  • Fields and time steps, cylindrical components (σr, σθ…), Gauss-point and ELNO values per element, values on click, draggable color bar.
  • 3D and 2D models (D_PLAN, C_PLAN, AXIS), and files with several meshes.
  • Journal figures (PDF + 600 dpi PNG + recipe) and journal-style line and history plots.
  • Read-only MCP server aster-med: statistics, profiles, histories, images, figures.
  • English, Spanish and French.

Tested on Linux with Python 3.10–3.12. Verification and known limitations are listed in the README.