GNU bug report logs - #24234
Recursive diff noise for broken symlinks, fifos and sockets

Previous Next

To reply to this bug, email your comments to 24234 AT debbugs.gnu.org.

Toggle the display of automated, internal messages from the tracker.

Message #5 received at submit <at> debbugs.gnu.org (full text, mbox):

From: Marcel Partap <mpartap <at> gmx.net>
To: bug-diffutils <at> gnu.org
Subject: Recursive diff noise for broken symlinks, fifos and sockets
Date: Mon, 15 Aug 2016 16:20:26 +0200

Dear diffutil devs,
when I copy large folder structures between harddisks, I like to check binary equality by running `diff -qr`. However, frequently the output is littered with powerful insights such as:
# File /mnt/source-xfs/repositories/kindlepdfviewer/emu_event is a fifo while file /usr/src/repositories/kindlepdfviewer/emu_event is a fifo
# File /mnt/source-xfs/repositories/android/i927-downloads/lost+found/#16412 is a socket while file /usr/src/repositories/android/i927-downloads/lost+found/#16412 is a socket
This does sound somewhat delusional.
Also, matching failures for broken symlinks:
# diff: /mnt/source-xfs/repositories/par2cmdline/test-driver: No such file or directory
# diff: /usr/src/repositories/par2cmdline/test-driver: No such file or directory
Additionally, recursive directory loops are being pointed out, but only on one side.

What I actually care for is missing and different files, not if there are broken symlinks, dir loops or fifos in the cloned directory hierarchy.
The documentation reads
>         -q, --brief
>                report only when files differ
so I think, it actually should do that. Is there any way at all fifos or sockets can actually differ?
I have no idea how to fix this, but hopefully someone else has : )

#Best Regards/Marcel Partap

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.