$(SHELL) config.status broken
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
GLib |
Fix Released
|
Medium
|
|||
GTK+ |
Fix Released
|
Medium
|
|||
Gnulib |
Fix Released
|
Undecided
|
|||
gettext |
Fix Released
|
Undecided
|
|||
intltool |
Fix Released
|
Medium
|
|||
autoconf (Ubuntu) |
Invalid
|
High
|
Unassigned | ||
dash (Ubuntu) |
Invalid
|
High
|
Unassigned | ||
gettext (Ubuntu) |
Fix Released
|
Undecided
|
Unassigned | ||
glib2.0 (Ubuntu) |
Fix Released
|
Low
|
Ubuntu Desktop Bugs | ||
gnulib (Ubuntu) |
Fix Released
|
Undecided
|
Unassigned | ||
gtk+2.0 (Ubuntu) |
Fix Released
|
Low
|
Ubuntu Desktop Bugs | ||
intltool (Ubuntu) |
Fix Released
|
Low
|
Unassigned |
Bug Description
Hi,
My "libtool" script in a random project (gtk+) was full of "^A" chars (\001); examining the build log, I could see that some Makefile.in.in would call: $(SHELL) ./config.status
I confirmed that after ./configure, libtool was ok (no ^A), but after "/bin/sh config.status" it was broken (with ^A). /bin/dash gives the same result, /bin/bash and /bin/zsh work fine.
I'm not sure that this is a dash regression, it could also be a new autoconf.
NB: I think the Makefile.in.in shouldn't have to use $(SHELL), but there are so many weird usages that I don't think I care to fight, also this is in at least gettext, intltool, glib, and copied in some custom Makefile.in.in (such as gtk+/po-
Bye
Related branches
Changed in dash: | |
importance: | Undecided → High |
Changed in autoconf: | |
importance: | Undecided → High |
Changed in gettext: | |
status: | New → Fix Committed |
Changed in gnulib: | |
status: | New → Fix Committed |
status: | Fix Committed → In Progress |
Changed in glib: | |
status: | Unknown → Fix Released |
Changed in gtk: | |
status: | Unknown → Fix Released |
Changed in intltool: | |
status: | Unknown → Fix Released |
Changed in glib: | |
importance: | Unknown → Medium |
Changed in intltool: | |
importance: | Unknown → Medium |
Changed in gtk: | |
importance: | Unknown → Medium |
Please provide a tarball containing the problem package, and instructions to replicate.
Clearly you're the only person seeing this, and so it must be something about the program you're using.