In the issue #31 it was clarified that there was a bug regarding different bit sizes of machine numbers on different systems/platforms (see also list_numpy_dtypes).
In list_numpy_dtypes I used my github action i386_debian_latest to run on a 32 bit system (at least using 32 bit libraries, the physical CPU is probable stil 64 bit). I expect that the code before the pull request #32 breaks on such a system.
Do you think cfunits could profit by this action? (I will not really maintain this action, but you/we could use it here for cfunits until it breaks and I'm not fixing it.)
daniel-mohr
commented
2 years ago
In the issue #31 it was clarified that there was a bug regarding different bit sizes of machine numbers on different systems/platforms (see also list_numpy_dtypes).
In list_numpy_dtypes I used my github action i386_debian_latest to run on a 32 bit system (at least using 32 bit libraries, the physical CPU is probable stil 64 bit). I expect that the code before the pull request #32 breaks on such a system.
Do you think cfunits could profit by this action? (I will not really maintain this action, but you/we could use it here for cfunits until it breaks and I'm not fixing it.)