We discussed a sliding scale (bigger files get a longer time out). But.. there is no meaningful way to choose that sliding scale. So I think the "single number timeout for all files" solution that Arrielle implemented here.
This function exists for a bonkers reason. Long term, I think we should move this function out of ARMI and into the cluster that is causing the problem.
What is the change?
os.system(cmd)
for windows toshutil.copyfile
/shutil.copymode
Why is the change being made?
copy
viaos.system
on Windows was taking an obscenely long time. And really we just needed the update from #1691 for linux functionality.Checklist
doc
folder.pyproject.toml
.