You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
As EUMM has logic based on the mtime of the Makefile.PL, issues can be encountered when a machine's clock is further in the past than when the Makefile.PL was generated. One way to fix this scenario is to normalize the mtime (with e.g. touch) of Makefile.PL when extracting if it's in "the future". It may or may not be reasonable to do this for all extracted files. There's also the -m option for tar, but that may not work on all tars and I don't think it's a good idea to normalize mtimes of files that will be installed (plus I bet there are tests depending on the mtime of various files).
For context: https://gist.github.com/Grinnz/c0554c9a373de99cb79a6d19d9d83f2e
As EUMM has logic based on the mtime of the Makefile.PL, issues can be encountered when a machine's clock is further in the past than when the Makefile.PL was generated. One way to fix this scenario is to normalize the mtime (with e.g. touch) of Makefile.PL when extracting if it's in "the future". It may or may not be reasonable to do this for all extracted files. There's also the -m option for tar, but that may not work on all tars and I don't think it's a good idea to normalize mtimes of files that will be installed (plus I bet there are tests depending on the mtime of various files).
I believe this logic would fit in the init_tools function of Menlo::CLI::Compat, https://github.com/miyagawa/cpanminus/blob/devel/Menlo-Legacy/lib/Menlo/CLI/Compat.pm#L2745
The text was updated successfully, but these errors were encountered: