Open ssun30 opened 1 week ago
Just FYI that the conda-build workflow doesn't work on RMG-Py, see #2641 which tried to fix it by making RMS optional (now in #2694) and #2636 which attempted to fix it as-is
You can probably just port whatever changes you make in RMG-database into this workflow but don't actually run it or anything, just bring the changes over so we don't forget
WARNING:root:Initial mole fractions do not sum to one; normalizing. WARNING:root:Initial mole fractions do not sum to one; normalizing. WARNING:root:Initial mole fractions do not sum to one; normalizing. ⚠️ One or more regression tests failed. Please download the failed results and run the tests locally or check the log to see why.
beep boop this comment was written by a bot :robot:
WARNING:root:Initial mole fractions do not sum to one; normalizing. WARNING:root:Initial mole fractions do not sum to one; normalizing. WARNING:root:Initial mole fractions do not sum to one; normalizing. ⚠️ One or more regression tests failed. Please download the failed results and run the tests locally or check the log to see why.
beep boop this comment was written by a bot :robot:
This is the mirror PR to this.
Motivation or Problem
The old Conda build workflow to set up Miniconda no longer works due to deprecation of the
default
channel. Now only theconda-forge
channel will be used. Furthermore, the new workflow directly uses Miniforge3 which has built-in mamba solver.Testing
Currently, this workflow only triggers upon releasing a stable version of RMG-Py. So I've set this PR to WIP until I've tested it to prevent overwriting the current stable RMG-Py package on the rmg channel.