I'd rather we show extra care towards Fedora developers who are still using Fedora, if there are any. If we make Fedora development overly complicated compared to working in a completely different environment or distro, people may migrate (I'd argue that's not good for Fedora). On pátek 25. července 2025 20:01:14, středoevropský letní čas Stephen Gallagher wrote: > If you are modifying contents of /usr in-place on your running host > system... you are taking your life in your hands. That's pretty much > the largest foot-gun you can have. If you aren't debugging in a > container or a virtual machine, but instead polluting your own primary > operating environment, you have only yourself to blame when something > goes horribly wrong. Yes, I agree that GNU/Linux power users should rule with with great power and understand their responsibility. The key is to understand what we are doing and sometimes take calculated risks (but how, when I don't know what I'm actually modifying). In the worst-case scenario, tools such as `rpm -V` and `dnf diff` can be invaluable (propose the local fix upstream). > If you're talking about doing something like `make install` which is > putting work-in-progress content into /usr... No please, never do `sudo make install` or other craziness like `curl example.com/foo | bash` commands (key point of RPM packaging workshops we do). > that's unlikely to be meaningfully affected here; it will not modify the > existing files, it will create new ones with new inodes and put them in > the same > location. So that shouldn't be affected by this. However, if you're > actually hand-editing a file on disk that is now hardlinked in > multiple places, yeah: you're in for some pain. But on the flip side, > how likely is it that two packages that contain identical files > wouldn't be able to tolerate whatever change you are making anyway? I don't know. What is the expected storage saving for a minimal installation, default workstation, etc.? Any estimations or research available on this? I can see a perfect use-case for IoT here, where every byte matters. However, I also think this deserves an option and a way to revert to a normal distribution. But if the hard-link "hit" ratio is low, the risk will also be low. Pavel -- _______________________________________________ devel mailing list -- devel@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe send an email to devel-leave@xxxxxxxxxxxxxxxxxxxxxxx Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@xxxxxxxxxxxxxxxxxxxxxxx Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
