Re: [Nmh-workers] [nmh-1.2] a serious bug in the nmh configure script

[Joel Reicher]

> > >Perhaps someone with a better knowledge on the configuration script
> > >used by nmh can provide some feedback on this matter.
> > 
> > If you edit aclocal.m4 to change the == to a single =, rerun autoconf
> > and configure, does that fix the problem?
> 
> No, it does not fix the problem.  I have installed autoconf, but it
> requires m4 and perl too.  After changing aclocal.m4 as suggested
> and run autoreconf the "configure" script "works" but slocal.c fails
> when looking for the same header file:

Something like this is what's required.

--- configure.in.orig   2005-12-15 11:45:36.000000000 +1100
+++ configure.in        2005-12-23 09:07:29.000000000 +1100
@@ -603,7 +603,7 @@
 dnl CHECK FOR NDBM.H
 dnl ----------------
 
-AC_CHECK_HEADERS(db1/ndbm.h gdbm/ndbm.h db.h, break, )
+AC_CHECK_HEADERS(db1/ndbm.h ndbm.h gdbm/ndbm.h db.h, break, )
 
 dnl --------------
 dnl CHECK FOR NDBM
--- aclocal.m4.orig     2005-12-15 11:45:36.000000000 +1100
+++ aclocal.m4  2005-12-23 09:14:06.000000000 +1100
@@ -44,7 +44,7 @@
 dnl the right header files.)
 AC_DEFUN(NMH_CHECK_DBM,
 [
-if test "x$1" == "x"; then
+if test "x$1" = "x"; then
    nmh_libs=
    dnl this is just for the benefit of AC_CACHE_CHECK's message
    nmh_testname=libc
@@ -59,6 +59,9 @@
 #ifdef HAVE_DB1_NDBM_H
 #include <db1/ndbm.h>
 #else
+#ifdef HAVE_NDBM_H
+#include <ndbm.h>
+#else
 #ifdef HAVE_GDBM_NDBM_H
 #include <gdbm/ndbm.h>
 #else
@@ -70,6 +73,7 @@
 #endif
 #endif
 #endif
+#endif
 ]],
 [[dbm_open("",0,0);]]),[nmh_cv_check_dbm_$1=yes],[
 nmh_cv_check_dbm_$1=no])



But I'm not sure how well this works for other platforms and situations.
That's why I wrote what I did in my previous emails.

Cheers,

        - Joel