I get following error when using ranger's help (man page displayed on '?'): bzip2: I/O or other error, bailing out. Possible reason follows. bzip2: Broken pipe Input file = /usr/share/man/man1/ranger.1.bz2, output file = (stdout) bzip2: I/O or other error, bailing out. Possible reason follows. bzip2: Broken pipe Input file = /usr/share/man/man1/ranger.1.bz2, output file = (stdout) Reproducible: Always Steps to Reproduce: 1. Run ranger 2. Press '?' to open man page 3. Quit man page 4. Quit ranger
Do normal man pages work for you without this error? I see this bug with normal man usage (just from recently).
(In reply to comment #1) > Do normal man pages work for you without this error? > I see this bug with normal man usage (just from recently). Yes. I have no errors when doing "man ranger" from the shell.
And what is your bash version? I tracked down my problem to update bash from 4.1_p9 to 4.2_p20.
(In reply to comment #3) > And what is your bash version? 4.2_p24-r1 > I tracked down my problem to update bash from 4.1_p9 to 4.2_p20. But problem still occurs for me.
if you try to downgrade bash for a while does it improve? if not i will create separate bug, since it would be different issue.
(In reply to comment #5) > if you try to downgrade bash for a while does it improve? Hm, to bash-4.1_p9 or earlier, or later? The problem still occurs with 4.1_p9.
aha,that means these two are unrelated, because my issue is fixed when downgrade to 4.1_p9.
I'm been using ranger-1.6.1 ~amd64 and I can't reproduce the issue. Are we sure it's still current?
Me either. :-)
I get the error on 1.6.1 stable.
Also on ~amd64 1.7.1
I don't know if this will help, but this is the relevant code upstream: https://github.com/hut/ranger/blob/a65efba9c1db5cda2284e95d364e47bf652bf160/ranger/core/actions.py#L780 This part hasn't changed since 2013. As we can see, it's simply "process = self.run(['man', 'ranger'])" being called. I can't figure out for sure where `self.run()` comes from, but I think it's from https://github.com/hut/ranger/blob/bababd8cd4259e90f20ab7dccde3de27cde8dcf6/ranger/core/runner.py#L88 ... Oh, I think I have it. If I re-emerge ranger with python 2.7 use flags (I've always used ranger through python 3), I can reproduce the bug. Can we confirm that reproducing the bug is tied to python 2?
I've opened a ticket at https://github.com/hut/ranger/issues/325
It would seem that this bug is not related to ranger: python2 -c 'from subprocess import *; Popen(["man", "ls"]).wait()' reproduces the bug for me (but not with python 3).
Also, this bug seems to be specific to gentoo: I can't reproduce it on an Ubuntu 14.04 environment (2.7.6) I have lying around. Ranger's author says he can't reproduce it on Arch (2.7.10).
Might this be related? man prints warnings https://github.com/kovidgoyal/kitty/issues/277
It is very probable. Let's close this bug.
For me * Fixed by re-emerging kitty. Before re-emerging kitty, * Not related to ranger. * Happened using $ man foobar from the command line. * I use sys-apps/man-db. I upgraded man-db and all dependency, but it didn't help.
