diff --git a/README.md b/README.md
index ddb3ed8692..c31b122303 100644
--- a/README.md
+++ b/README.md
@@ -103,6 +103,13 @@ To execute `run_full.sh` on multiple machines in parallel create a machine addre
 ./run_multiple_machine.sh main machines.txt <project_path_on_machine>
 ```
 
+#### If `make small-test` fails
+
+If `make small-test` fails in a branch that you wish to merge into the main
+branch, run `make small-test` in the main branch (which should succeed) and also
+in your branch, and investigate the differences.
+
+
 ### Load the stored cache
 
 To decompress the cache run `make decompress-cache`. This is done automatically in `run_full.sh` if `cache/` does not exist.
diff --git a/run.sh b/run.sh
index 6fcf56e7bc..1da71605a8 100755
--- a/run.sh
+++ b/run.sh
@@ -138,7 +138,7 @@ echo "Number of machines: $num_machines"
 echo "Output directory: $OUT_DIR"
 echo "Options: $comparator_flags"
 
-./gradlew -q assemble -g ../.gradle/
+JAVA_HOME="$JAVA11_HOME" ./gradlew -q assemble -g ../.gradle/
 
 mkdir -p "$OUT_DIR"
 
@@ -168,7 +168,7 @@ python3 src/python/test_repo_heads.py \
     --output_path "$OUT_DIR/repos_head_passes.csv" \
     --cache_dir "$CACHE_DIR"
 
-java -cp build/libs/astmergeevaluation-all.jar \
+JAVA_HOME="${JAVA17_HOME}" "${JAVA17_HOME}"/bin/java -cp build/libs/astmergeevaluation-all.jar \
     astmergeevaluation.FindMergeCommits \
     "$OUT_DIR/repos_head_passes.csv" \
     "$OUT_DIR/merges"
diff --git a/src/python/cache_utils.py b/src/python/cache_utils.py
index ff76b4dd09..1532849405 100755
--- a/src/python/cache_utils.py
+++ b/src/python/cache_utils.py
@@ -86,7 +86,7 @@ def lookup_in_cache(
         lock.release()
         return cache_data
     if set_run:
-        logger.debug(f"lookup_in_cache: Setting {cache_key} to None")
+        logger.debug(f"lookup_in_cache: Setting {cache_key} to None for {repo_slug}")
         set_in_cache(cache_key, None, repo_slug, cache_directory, acquire_lock=False)
     lock.release()
     return None
diff --git a/src/python/merge_runtime_measure.py b/src/python/merge_runtime_measure.py
index b6977dd1a3..3183e37a7a 100755
--- a/src/python/merge_runtime_measure.py
+++ b/src/python/merge_runtime_measure.py
@@ -81,7 +81,7 @@ def main():
                     else:
                         logger.info(
                             f"merge_timer: Running {merge_tool.name} "
-                            f"on {repo_slug} {left_hash} {right_hash}"
+                            f"on {idx} {repo_slug} {left_hash} {right_hash}"
                         )
                         run_times = []
                         for _ in range(args.n_timings):
diff --git a/src/python/test_repo_heads.py b/src/python/test_repo_heads.py
index da31c897aa..61f4129e5e 100755
--- a/src/python/test_repo_heads.py
+++ b/src/python/test_repo_heads.py
@@ -136,9 +136,9 @@ def head_passes_tests(args: Tuple[pd.Series, Path]) -> pd.Series:
     logger.info("test_repo_heads: Finished Building Output")
 
     logger.success(
-        "test_repo_heads: Number of repos whose head passes tests:"
+        "test_repo_heads: Number of repos whose head passes tests: "
         + str(len(filtered_df))
-        + "out of"
+        + " out of "
         + str(len(df))
     )
     if len(filtered_df) == 0:
diff --git a/src/python/write_head_hashes.py b/src/python/write_head_hashes.py
index 732de1ec6c..a0243aa1ff 100755
--- a/src/python/write_head_hashes.py
+++ b/src/python/write_head_hashes.py
@@ -77,13 +77,17 @@ def get_latest_hash(args):
 
     # If file exists ignore this step
     if os.path.isfile(arguments.output_path):
-        logger.info("write_head_hashes: File already exists, skipping")
+        logger.info(
+            f"write_head_hashes: File already exists, skipping {arguments.output_path}"
+        )
         sys.exit(0)
 
     df = pd.read_csv(arguments.repos_csv, index_col="idx")
     df["repository"] = df["repository"].str.lower()
 
-    logger.info("write_head_hashes: Started cloning repos and collecting head hashes")
+    logger.info(
+        "write_head_hashes: Started cloning repos and collecting head hashes for {arguments.output_path}"
+    )
     with multiprocessing.Pool(processes=num_processes()) as pool:
         with Progress(
             SpinnerColumn(),
diff --git a/src/scripts/merge_tools/git_hires_merge.sh b/src/scripts/merge_tools/git_hires_merge.sh
index 19e47d2c88..a91d69cda1 100755
--- a/src/scripts/merge_tools/git_hires_merge.sh
+++ b/src/scripts/merge_tools/git_hires_merge.sh
@@ -9,7 +9,7 @@ branch2=$3
 # Print the current PATH
 echo "PATH: $PATH"
 
-cd "$clone_dir" || exit 1
+cd "$clone_dir" || (echo "$0: cannot cd to $clone_dir" ; exit 1)
 
 git checkout "$branch1" --force
 
diff --git a/src/scripts/merge_tools/gitmerge.sh b/src/scripts/merge_tools/gitmerge.sh
index 3aacd86864..0fdbeddd10 100755
--- a/src/scripts/merge_tools/gitmerge.sh
+++ b/src/scripts/merge_tools/gitmerge.sh
@@ -20,7 +20,7 @@ branch2=$3
 strategy=$4
 
 # perform merge
-cd "$clone_dir" || exit 1
+cd "$clone_dir" || (echo "$0: cannot cd to $clone_dir" ; exit 1)
 
 git checkout "$branch1" --force
 git config merge.conflictstyle zdiff3
diff --git a/src/scripts/merge_tools/gitmerge_resolve.sh b/src/scripts/merge_tools/gitmerge_resolve.sh
index 859faf70e2..5c5965d0bc 100755
--- a/src/scripts/merge_tools/gitmerge_resolve.sh
+++ b/src/scripts/merge_tools/gitmerge_resolve.sh
@@ -12,7 +12,7 @@ status=$?
 
 if [ "$status" -ne 0 ]; then
   echo "Removing filenames from conflict markers."
-  cd "$clone_dir" || exit 1
+  cd "$clone_dir" || (echo "$0: cannot cd to $clone_dir" ; exit 1)
   readarray -t files < <(grep -l -r '^\(<<<<<<<\||||||||\|>>>>>>>\) .merge_file_')
   for file in "${files[@]}" ; do
     echo "Removing filenames from conflict markers in $file"
diff --git a/src/scripts/merge_tools/intellimerge.sh b/src/scripts/merge_tools/intellimerge.sh
index 8fd4ccaf47..f1d9deaff8 100755
--- a/src/scripts/merge_tools/intellimerge.sh
+++ b/src/scripts/merge_tools/intellimerge.sh
@@ -34,7 +34,7 @@ echo "Temp dir: $temp_out_dir"
 clone_dir_absolutepath=$(realpath "$clone_dir")
 
 # run intellimerge
-cd "$clone_dir" || exit 1
+cd "$clone_dir" || (echo "$0: cannot cd to $clone_dir" ; exit 1)
 
 java -Djava.util.concurrent.ForkJoinPool.common.parallelism=1 -Djava.io.tmpdir="$temp_intellimerge_dir" -jar "$intellimerge_absolutepath" -r "$clone_dir_absolutepath" -b "$branch1" "$branch2" -o $temp_out_dir
 
diff --git a/src/scripts/merge_tools/merge_plumelib.sh b/src/scripts/merge_tools/merge_plumelib.sh
index 4000d080c6..ebfe4745a0 100755
--- a/src/scripts/merge_tools/merge_plumelib.sh
+++ b/src/scripts/merge_tools/merge_plumelib.sh
@@ -10,7 +10,7 @@ merge_strategy=$5 #"--only-adjacent"
 
 echo "$0: Merging $branch1 and $branch2 with git_strategy=$git_strategy and merge_strategy=$merge_strategy"
 
-cd "$clone_dir" || exit 1
+cd "$clone_dir" || (echo "$0: cannot cd to $clone_dir" ; exit 1)
 
 git checkout "$branch1" --force
 
diff --git a/src/scripts/merge_tools/spork.sh b/src/scripts/merge_tools/spork.sh
index 4505403c3d..b652dad3a6 100755
--- a/src/scripts/merge_tools/spork.sh
+++ b/src/scripts/merge_tools/spork.sh
@@ -26,7 +26,7 @@ clone_dir=$1
 branch1=$2
 branch2=$3
 
-cd "$clone_dir" || exit
+cd "$clone_dir" || (echo "$0: cannot cd to $clone_dir" ; exit 1)
 
 # set up spork driver
 git config --local merge.spork.name "spork"
diff --git a/test/check_equal_csv.py b/test/check_equal_csv.py
index 64318e0957..1a47af11e6 100755
--- a/test/check_equal_csv.py
+++ b/test/check_equal_csv.py
@@ -51,7 +51,10 @@ def remove_run_time(df):
             if actual_df[col].equals(goal_df[col]):
                 continue
             # Print the differences.
-            print(os.system(f"diff {goal_folder/goal_file} {actual_file}"))
+            diff_exit_code = os.waitstatus_to_exitcode(
+                os.system(f"diff {goal_folder/goal_file} {actual_file}")
+            )
+            print(f"diff exit code: {diff_exit_code}")
             # Now print details, after diffs so it is not obscured by the diff output.
             different_columns = []
             for col in goal_df.columns: