Repository navigation
Behavior differences with ranger #2706
|
While investigating permission related issues, I looked into ranger and where lf behaves differently.
Any opinion if these differences should be addressed? I tend towards yes except for the special bits on copy (they can be left out) As for the mtime issue, I would tend towards addressing that by changing the preserve option into bool and make it preserve mtime as well. The separation between mode and timestamps does not make a lot of sense to me |
Replies: 2 comments 3 replies
To get this out of the way: As already discussed in #2678, a final
I don't have strong feelings about this. If you want to change the default value for
I am against that. Imagine The fact that seemingly nobody ever noticed or complained about the things you mention tells me that things like special bits and xattrs are rather niche use cases. There are so many differences between There is also the point to be made that Besides, |
To get this out of the way:
lfdoesn't intentionally strip these attributes when moving files (currently). It first triesos.Rename, which doesn't change the file's inode and therefore retains permissions, ACLs, xattrs, etc. If that fails because the move crosses filesystems, it falls back tocopyAll+os.RemoveAlland now relies onlf's normal copy a…