+# Using this file in a test
+# =========================
+#
+# The typical skeleton of a test looks like this:
+#
+# #!/bin/sh
+# : ${srcdir=.}
+# . "$srcdir/init.sh" --set-path=.
+# Execute some commands.
+# Note that these commands are executed in a subdirectory, therefore you
+# need to prepend "../" to relative filenames in the build directory.
+# Set the exit code 0 for success, 77 for skipped, or 1 or other for failure.
+# with the corresponding exit code.
+# Exit $?
+
+# Executing a test that uses this file
+# ====================================
+#
+# Running a single test:
+# $ make check TESTS=test-foo.sh
+#
+# Running a single test, with verbose output:
+# $ make check TESTS=test-foo.sh VERBOSE=yes
+#
+# Running a single test, with single-stepping:
+# 1. Go into a sub-shell:
+# $ bash
+# 2. Set relevant environment variables from TESTS_ENVIRONMENT in the
+# Makefile:
+# $ export srcdir=../../tests # this is an example
+# 3. Execute the commands from the test, copy&pasting them one by one:
+# $ . "$srcdir/init.sh" --set-path=.
+# ...
+# 4. Finally
+# $ exit
+