What | Removed | Added |
---|---|---|
Status | RESOLVED | REOPENED |
Resolution | FIXED | --- |
The bug is back and this time I had the opportunity to try several installation and I think I understood what is going on. The installation I did was with tumbleweed snapshot 20240208. I have a local copy of it so I can do several installation test without worrying about new packages being available in tumbleweed. During a installation, after the 'man' package is installed the error messages about '/var/cache/man/nnnnn not found' starts to appear in /var/log/zypp/history. Each error message is printed after the corresponding package is installed. The error is consistent with the fact that /var/cache/man does not exist while the packages are installed, because it gets created at the end of the installation process when the transpost scriptlet of the package 'man' get executed. During the installation, while the error messages were written in /var/log/zypp/history, I decided to run: /usr/bin/systemd-tmpfiles --create /usr/lib/tmpfiles.d/man-db.conf and after that /var/cache/man was created and in /var/log/zypp/history there were no more error messages. It looks to me like the macro %transfiletriggerin defined in the package 'man' is executed not at the end of the transaction, but at the end of each package installation, like if it was a %filetriggerin macro. If I have to guess, there is a bug in the code that decides when to run the filetrigger macro.