v0.12.0

New

• Any simulation code's snapshot can be synthesised from an atmosphere file, one HDF5 file of the temperature, density and velocity, which synthesise-spectra --atmosphere reads and eclipse-atmosphere info describes. (#88)
• eclipse can observe a time series of atmosphere or synthesis files as a raster or a sit-and-stare, so that each exposure sees the atmosphere at its own time. A raster can step towards increasing or decreasing x. (#93, #108, #129)
• Every synthesis is one HDF5 synthesis file. Spectra from another code can be written as one with write_synthesis, and observed like ECLIPSE's own. (#108)
• The results are written as an HDF5 file, run/result/<config name>.h5. (#81)
• telescope.psf_across_slit adds the telescope's blur across the slit, which brings in light from either side of it. (#124)
• A configuration is checked in full before anything runs, and a mistake stops the run with a message that names it. (#122, #129)
• ECLIPSE offers to build fiasco's CHIANTI database before the first synthesis needs it. (#113)

Replaced, and still working with a warning

• Giving MURaM's own files to synthesise-spectra, and its dynamic mode, --slit-rest-time. Atmosphere files and atmosphere_series replace them. (#88, #93)
• Synthesis and results pickles. convert_synthesis_pickle and convert_results_pickle rewrite them as HDF5 files. (#81, #108)
• --mean-mol-wt is now --mass-per-electron. (#88)

Requirements

• dill 0.3.6 or later. specutils is no longer needed. (#108, #131)
• Building from source needs setuptools 77 or later, which pip fetches itself. (#131)

Documentation

• The docs are rewritten and arranged around the three stages of a run. They add a first run that needs no downloads, a configuration reference, a page on analysing the results, and an API reference. (#140, #142)

Changes to output

• Output can differ from v0.11.0 for the same input, so results made with earlier versions should be regenerated.