On Wed, Aug 02, 2023 at 10:18:16AM +0200, Omar Polo wrote:
> On 2023/08/02 09:56:07 +0200, Omar Polo <[email protected]> wrote:
> > On 2023/07/31 19:37:44 -0400, Thomas Frohwein <[email protected]> 
> > wrote:
> > > I am attaching a tarball with the above changes. ok?
> > 
> > you could consider changing the hardcoded /usr/local/ with ${PREFIX}
> > to please portcheck.
> 
> err, i meant LOCALBASE :-)

Thanks, I committed it with /usr/local/ still. There have been
discussions around this topic and the value (or lack thereof) on
changing the hardcode paths.

As we seem to be in a state of limbo about this, let me offer a step in
the direction of relaxing about /usr/local/ by calming portcheck:

Index: portcheck
===================================================================
RCS file: /cvs/ports/infrastructure/bin/portcheck,v
retrieving revision 1.144
diff -u -p -r1.144 portcheck
--- portcheck   24 Apr 2023 11:44:20 -0000      1.144
+++ portcheck   2 Aug 2023 22:13:06 -0000
@@ -1715,13 +1715,12 @@ check_pkg_dir() {
 }
 
 # Checks made:
-#   * There are no hardcoded /usr/local or /var paths in file.
+#   * There are no hardcoded /var paths in file.
 #     /var/log, /var/run and /var/tmp are perfectly fine, though.
 check_hardcoded() {
        $debugging && echo "CALLED: check_hardcoded($*)" >&2
 
        perl -n -e 'BEGIN { $ec=0; }
-                   if (m,/usr/local\b,o) { $ec=1; close ARGV; }
                    if (m,/var((?:/+[^/\s]+)*)(?:\s.*)?$,o) {
                        unless ($1 =~ m,^/+(?:log|run|tmp),o) {
                            $ec=1; close ARGV;
@@ -1729,7 +1728,7 @@ check_hardcoded() {
                    }
                    END { $? = $ec; }' \
                "$1" || err "hardcoded paths detected in $1, consider using" \
-                   "SUBST_VARS and TRUEPREFIX/LOCALBASE/LOCALSTATEDIR/VARBASE"
+                   "SUBST_VARS and LOCALSTATEDIR/VARBASE"
 }
 
 # Checks made:

Reply via email to