[Bug-ncurses] ncurses 5.1 (20000909/20000917) build failure

[Matthew Clarke]

Hi.

Just updated ncurses to 5.1 (first to 20000909, then a few days later to
20000917.)

When building outside the source tree, "make install" fails in two places:

The generated file "<build-dir>/man/edit_man.sh", when generating aliases
for each page, looks for the page source in the local (build) directory,
not the source directory.  Possible change to aclocal.m4:

------------------------------------------------------------------------------
--- aclocal.m4.old      Thu Sep 14 18:40:02 2000
+++ aclocal.m4  Sun Sep 17 19:52:58 2000
@@ -1440,7 +1440,7 @@
 CF_EOF
 if test "$cf_manpage_symlinks" = yes ; then
 cat >>man/edit_man.sh <<CF_EOF
-       aliases=\`sed -f \$srcdir/manlinks.sed \$source | sort -u\`
+       aliases=\`sed -f \$srcdir/manlinks.sed \$i | sort -u\`
 CF_EOF
 fi
 if test "$cf_manpage_renames" = no ; then
------------------------------------------------------------------------------

The file "misc/run_tic.sh" is generated, and therefore exists in
"<build-dir>/misc", not $(srcdir).  Possible change to misc/Makefile.in:

Tested:
------------------------------------------------------------------------------
--- misc/Makefile.in.old        Thu Sep 14 18:40:03 2000
+++ misc/Makefile.in    Sun Sep 17 19:54:22 2000
@@ -83,7 +83,7 @@
        datadir=${datadir} \
        ticdir=${ticdir} \
        source=${source} \
-       $(SHELL) $(srcdir)/run_tic.sh
+       $(SHELL) ./run_tic.sh
        @cd $(srcdir)/tabset && \
                $(SHELL) -c 'for i in `echo * | fgrep -v CVS | fgrep -v RCS`; 
do \
                echo installing $$i; \
------------------------------------------------------------------------------

The misc/Makefile.in change is tested, as it requires no other changes.
The aclocal.m4 change is not tested, as my autoconf is 2.13, not a
snapshot.  However, I did make the (hopefully) equivalent change to
configure, and it generated a man/edit_man.sh that did work.

Thanks for ncurses,
Matt.
-- 
>Almost any animal is capable [of] learning a stimulus/response association,
>given enough repetition.
Experimental observation suggests that this isn't true if double-clicking
is involved.                    -- Lionel, Malcolm Ray, in a.s.r.