2026.05

Changelog: Ramses 2026

Summary

Users

• A lot of performance improvements on the hydro kernels (by @tinecolman) and the sink particles (by @georgesabboudeh)
• Many bugs were fixed (by @Martin-Rey, @tinecolman, @nbrucy, @Sadanari0614, @jokirosdahl, @cosmos-in-us, @robelgeda, @James11222, @bcommerc, @Adnan-Ali-Ahmad)
• A new utilitary to write grafic-style input files (by @tinecolman)
• It is now possible to select both analytical forces with the poisson solver (by @simonb282)
• It is now possible to output conservative instead of primitive hydro variables (by @tinecolman)
• The EOS were improved (by @tinecolman)
• There is now an option to have output at exact time (by @wzzhang97). ⚠️ Use with care ⚠️
• The documentation was improved (by @nbrucy and @georgesabboudeh)

Developers

• We have a new documentation for developers following the first Ramses dev school. Discover it here (by the SNO's Developer school working group, @tinecolman, @blaizot @nbrucy @mtrebitsch, @cphyc and @jokirosdahl)
• A lot of new tests were added, including crossing beams, stromgren2D, RT dirac, Ponomarenko by @matthiasglz, restart and improved SOD test by @nbrucy, refinement test by @bcommerc, clumpfinfer, star formation and SN feedback tests by @pocvirk, cosmo amr and 1D advection tests by @tinecolman, and movie and chemistry tests by @jokirosdahl. This is the results of the SNO's test working group.
• The code quality was improved (by @tinecolman and @mtrebitsch). The biggest changes are in how the parameters are read.
• The tests results are now displayed directly as a comment in the PR (by @nbrucy)

Detailled changes

🎉 New features

• Python to RAMSES input through grafic files + test by @tinecolman in #232
• More flexible gravity sources by @simonb282 in #153
• Option to output conservative instead of primitive hydro variables by @tinecolman in #230
• Add EOS for 2nd collapse setups + some general EOS improvements by @tinecolman in #257
• Add optional exact output time control by @wzzhang97 in #258

🚀 Performance improvements

• [SPACE] Optimize and document nbor utils by @tinecolman in #108
• Optimized the creation of cloud particles by @georgesabboudeh in #140
• Optimize ctoprim by @tinecolman in #158
• Refactor and optimize uslope by @tinecolman in #154
• optimizations for hydro solver on uniform grid by @tinecolman in #159
• Call uslope inside trace by @tinecolman in #205

🐞 Bugs

• Remove hard cast of idp to short int by @Martin-Rey in #117
• Fix bug in phi_fine_cg.f90 by @bcommerc in #137
• set npartmax to 0 if pic is false by @nbrucy in #150
• Fix nlevelmax_sink not working as intended by @tinecolman in #151
• Fix incorrect T2min calculation in cooling_fine.f90 when jeans_ncells > 0 by @cosmos-in-us in #148
• Correction of bugs in condinit_kind='collapse' by @matthiasglz in #177
• update dice patch by @tinecolman in #172
• Updated patch files to reflect changes in main code by @jokirosdahl in #180
• use buildinfo in read_params of the dice patch by @robelgeda in #192
• Fix amr2cube routine in utils by @nbrucy in #195
• changed the default load-in in amr/init_time.f90 from deltab to velcx… by @James11222 in #249
• correct a bug in the lightcone module by @tinecolman in #250
• Fix galic util and update merger patch by @tinecolman in #242
• Minor bugfix: fix invalid access to sinks arrays by @nbrucy in #260
• bugfix threshold accretion by @nbrucy in #261
• Fix non-convergence in the H2 fraction for zero-metallicity gas by @jokirosdahl in #262
• Correct H- formation rate coefficient tbl_alphaGP_H2%rates by @Sadanari0614 in #266
• Cache the namelist file on startup by @Adnan-Ali-Ahmad in #268

🧪 Test Cases

• Add a RT stromgren2d test with He ionization by @matthiasglz in #122
• [tests] Add a -r option to test restart by @nbrucy in #110
• Add Sedov and geometry refine test by @bcommerc in #124
• [tests] solve bug where execution test log was not written by @nbrucy in #125
• Add a RT crossing beams test by @matthiasglz in #126
• [tests] fix small values being rounded to 0 by visu_ramses by @tinecolman in #127
• [tests] Update stromgren2d test with the inclusion of molecular hydrogen and fixed compilation error when compiling without MPI by @jokirosdahl in #136
• [tests] adjust outputting for sedov3d test to work with test restart mechanism by @tinecolman in #134
• [tests] remove levelhack tests and put its changes to the bondi test by @tinecolman in #135
• [tests] Check SOD tests against analytical solution by @nbrucy in #129
• prevent cosmo ic download from hanging when the website is down by @tinecolman in #139
• [tests] Add collapse test for MHD by @bcommerc in #130
• [tests] Added a check for the clumpfinder by @pocvirk in #132
• [tests] adjust tolerance for median error in tube tests by @tinecolman in #145
• regenerate reference solution for stellar-HII with 2 cpus by @tinecolman in #160
• Movie test by @jokirosdahl in #146
• [tests] Do a parameter study in the sod-tube analytical test by @nbrucy in #141
• Improvements to parameter sweep of sod-tube test by @tinecolman in #165
• Add 1D advection test by @tinecolman in #166
• fix Cosmo test fails to stop when testing with restart #167 by @nbrucy in #168
• change slope_type in imhd-tube and imhd-tube-nener tests by @tinecolman in #169
• [Test suite] Improvements to testing restarts by @tinecolman in #194
• [Tests] New tests for the equilibrium and non-equilibrium cooling module by @jokirosdahl in #191
• [tests] Add MHD Ponomarenko dynamo test by @matthiasglz in #131
• [tests] RT dirac test by @matthiasglz in #133
• Add ABC-flow dynamo test by @matthiasglz in #201
• added star formation and SN feedback test by @pocvirk in #198
• Add a solver parameter study in the imhd test by @matthiasglz in #219
• Add support of bisection decomposition in the test suite by @matthiasglz in #220
• Added sfvirial=true sfmodel=0 test by @pocvirk in #241

🛠️ CI

• Deactivate auto fix by @cphyc in #142
• Update tests workflow to always run tests + other CI fixes by @nbrucy in #221
• CI improvement: add a comment with the link towards tests results by @nbrucy in #222
• [CI] Remove commit status actions from post-pr workflow by @nbrucy in #225

🧹 Code Quality

• Directly store the flux in cmpflxm instead of using intermediate arrays fx and tx by @tinecolman in #147
• Adding missing commas between descriptors in init_turb.f90 by @mtrebitsch in #176
• combine the hydro and mhd versions of init_flow_fine.f90 into one by @tinecolman in #174
• Combine hydro, mhd and rhd versions of hydro_parameters.f90 by @tinecolman in #184
• Combine hydro and mhd versions of output_hydro.f90 by @tinecolman in #183
• Overhaul of compiler flags in Makefile by @tinecolman in #185
• move build variables to a module by @tinecolman in #190
• Combine the hydro and mhd version of init_hydro.f90 by @tinecolman in #203
• Collection of minor fixes and cleanups by @tinecolman in #212

📚 Documentation

• Remove references to the stable branch in the doc by @nbrucy in #144
• further document the order of hydro variables by @tinecolman in #178
• Documentation improvements (add missing links) by @nbrucy in #209
• Improve code comments by @nbrucy in #252
• Added documentation for (cloud_pts_check) by @georgesabboudeh in #256
• Add a documentation for developers by @nbrucy in #228

New Contributors

• @Martin-Rey made their first contribution in #117
• @bcommerc made their first contribution in #124
• @cosmos-in-us made their first contribution in #148
• @mtrebitsch made their first contribution in #176
• @georgesabboudeh made their first contribution in #140
• @James11222 made their first contribution in #249
• @simonb282 made their first contribution in #153
• @wzzhang97 made their first contribution in #258
• @Sadanari0614 made their first contribution in #266
• @Adnan-Ali-Ahmad made their first contribution in #268

Full Changelog: 2025.05...2026.05