Bug 56656
| Summary: | rpmb is producing packages with messed-up man page links | ||
|---|---|---|---|
| Product: | [Retired] Red Hat Raw Hide | Reporter: | Jonathan Kamens <h1k6zn2m> |
| Component: | rpm-build | Assignee: | Jeff Johnson <jbj> |
| Status: | CLOSED DUPLICATE | QA Contact: | |
| Severity: | medium | Docs Contact: | |
| Priority: | medium | ||
| Version: | 1.0 | ||
| Target Milestone: | --- | ||
| Target Release: | --- | ||
| Hardware: | i386 | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Fixed In Version: | Doc Type: | Bug Fix | |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2001-11-23 18:09:44 UTC | Type: | --- |
| Regression: | --- | Mount Type: | --- |
| Documentation: | --- | CRM: | |
| Verified Versions: | Category: | --- | |
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
| Cloudforms Team: | --- | Target Upstream Version: | |
| Embargoed: | |||
| Bug Depends On: | |||
| Bug Blocks: | 55960, 56653, 56654, 56655 | ||
|
Description
Jonathan Kamens
2001-11-23 14:38:10 UTC
Report against the packages, not rpm. It's up to the package owners to get the bits into the package correctly, not up to rpm. It irritates me when I file a bug and the owner of the bug assumes that I'm
wrong rather than taking the time to check.
There is no bug in the other packages about which I filed this bug. The bug is
in rpm-build, in particular, in this line in /usr/lib/rpm/brp-compress:
l=`ls -l $f | awk '{ print $11 }' | sed -e 's/\.gz$//' -e 's/\.bz2$//' -e
's/\.Z$//'`
The output format of "ls" changed recently (perhaps as the result of something
related to my locale settings combined with a change in how ls handles various
locale settings?), such that $11 is no longer the correct field number for
retrieving the file name to which a link points:
jik:/tmp!39> ls -l /usr/share/man/man1/zsoelim.1.gz
lrwxrwxrwx 1 root root 11 10-12 09:55
/usr/share/man/man1/zsoelim.1.gz -> soelim.1.gz
Rather than simply changing the script to rely on a different field number,
which is prone to work on some systems and not others and which in any case may
break again in the future, here's a better fix:
--- /usr/lib/rpm/brp-compress Thu Sep 13 17:03:15 2001
+++ /tmp/brp-compress Fri Nov 23 12:18:31 2001
@@ -48,7 +48,7 @@
for f in `find $d -type l`
do
- l=`ls -l $f | awk '{ print $11 }' | sed -e 's/\.gz$//' -e 's/\.bz2$//' -e
's/\.Z$//'`
+ l=`ls -l $f | sed -e 's/.* -> //' -e 's/\.gz$//' -e 's/\.bz2$//' -e
's/\.Z$//'`
rm -f $f
b=`echo $f | sed -e 's/\.gz$//' -e 's/\.bz2$//' -e 's/\.Z$//'`
ln -sf $l$COMPRESS_EXT $b$COMPRESS_EXT
It irritates me when a user makes assumptions about my behavior and then presumes to criticize me for problems outside my control. I carry the brp-compress script for the os-devel group. I do not, and can not, change those bits without consensus and approval. Yes the script is feeble and broken. I've known that for quite awhile. And thanks for the patch :-) I blame it all on jbj. In fact, I am starting to think a good 50 or so XFree86 bugs are actually the result of bugs introduced by RPM. ;o) |