diff --git a/osm_fieldwork/CSVDump.py b/osm_fieldwork/CSVDump.py
index c9e29277..83cc256f 100755
--- a/osm_fieldwork/CSVDump.py
+++ b/osm_fieldwork/CSVDump.py
@@ -1,6 +1,6 @@
 #!/usr/bin/python3
 
-# Copyright (c) 2020, 2021, 2022, 2023 Humanitarian OpenStreetMap Team
+# Copyright (c) 2020, 2021, 2022, 2023, 2024 Humanitarian OpenStreetMap Team
 #
 # This file is part of OSM-Fieldwork.
 #
@@ -24,6 +24,7 @@
 import os
 import re
 import sys
+from datetime import datetime
 
 import pandas as pd
 from geojson import Feature, FeatureCollection, Point, dump
@@ -125,7 +126,8 @@ def writeOSM(
 
     def finishOSM(self):
         """Write the OSM XML file footer and close it."""
-        self.osm.footer()
+        # This is now handled by a destructor in the OsmFile class
+        # self.osm.footer()
 
     def createGeoJson(
         self,
@@ -149,6 +151,9 @@ def finishGeoJson(self):
         """Write the GeoJson FeatureCollection to the output file and close it."""
         features = list()
         for item in self.features:
+            if len(item["attrs"]["lon"]) == 0 or len(item["attrs"]["lat"]) == 0:
+                log.warning("Bad location data in entry! %r", item["attrs"])
+                continue
             poi = Point((float(item["attrs"]["lon"]), float(item["attrs"]["lat"])))
             if "private" in item:
                 props = {**item["tags"], **item["private"]}
@@ -293,8 +298,9 @@ def createEntry(
 
                 if value is not None and value != "no" and value != "unknown":
                     if key == "track" or key == "geoline":
-                        refs.append(tag)
-                        log.debug("Adding reference %s" % tag)
+                        # refs.append(tags)
+                        # log.debug("Adding reference %s" % tags)
+                        refs = value.split(";")
                     elif len(value) > 0:
                         if self.privateData(key):
                             priv[key] = value
@@ -303,7 +309,7 @@ def createEntry(
             if len(tags) > 0:
                 feature["attrs"] = attrs
                 feature["tags"] = tags
-            if len(refs) > 0:
+            if len(refs) > 1:
                 feature["refs"] = refs
             if len(priv) > 0:
                 feature["private"] = priv
@@ -344,12 +350,27 @@ def main():
     log.debug("Parsing csv files %r" % args.infile)
     data = csvin.parse(args.infile)
     # This OSM XML file only has OSM appropriate tags and values
+    nodeid = -1000
     for entry in data:
         feature = csvin.createEntry(entry)
-        # Sometimes bad entries, usually from debugging XForm design, sneak in
         if len(feature) == 0:
             continue
-        if len(feature) > 0:
+        if "refs" in feature:
+            refs = list()
+            for ref in feature["refs"]:
+                now = datetime.now().strftime("%Y-%m-%dT%TZ")
+                if len(ref) == 0:
+                    continue
+                coords = ref.split(" ")
+                print(coords)
+                node = {"attrs": {"id": nodeid, "version": 1, "timestamp": now, "lat": coords[0], "lon": coords[1]}, "tags": dict()}
+                csvin.writeOSM(node)
+                refs.append(nodeid)
+                nodeid -= 1
+            feature["refs"] = refs
+            csvin.writeOSM(feature)
+        else:
+            # Sometimes bad entries, usually from debugging XForm design, sneak in
             if "lat" not in feature["attrs"]:
                 log.warning("Bad record! %r" % feature)
                 continue
diff --git a/osm_fieldwork/odk2osm.py b/osm_fieldwork/odk2osm.py
index 9fc3c9cf..19d1bf3d 100755
--- a/osm_fieldwork/odk2osm.py
+++ b/osm_fieldwork/odk2osm.py
@@ -18,6 +18,7 @@
 #
 
 import argparse
+import csv
 import logging
 import os
 import re
@@ -102,14 +103,21 @@ def main():
                             continue
                         if jj is None:
                             continue
-                        # print(f"tag: {i} == {j}")
+                        print(f"tag2: {i} == {j}")
                         if type(jj) == OrderedDict or type(jj) == dict:
                             for iii, jjj in jj.items():
                                 if jjj is not None:
-                                    tags[iii] = jjj
-                                    # print(iii, jjj)
+                                    pat = re.compile("[0-9.]* [0-9.-]* [0-9.]* [0-9.]*")
+                                    if pat.match(str(jjj)):
+                                        gps = jjj.split(" ")
+                                        tags["lat"] = gps[0]
+                                        tags["lon"] = gps[1]
+                                        continue
+                                    else:
+                                        tags[iii] = jjj
+                                        # print(f"FOO {iii}, {jjj}")
                         else:
-                            # print(ii, jj)
+                            # print(f"WHERE {ii}, {jj}")
                             fields.append(ii)
                             tags[ii] = jj
                 else:
@@ -121,18 +129,19 @@ def main():
     tmp = xml.replace(" ", "").split("_")
     now = datetime.now()
     timestamp = f"_{now.year}_{now.hour}_{now.minute}"
+
     outfile = tmp[0] + timestamp + ".csv"
 
-    # with open(outfile, "w", newline="") as csvfile:
-    #     fields = list()
-    #     for row in rows:
-    #         for key in row.keys():
-    #             if key not in fields:
-    #                 fields.append(key)
-    #     csv = csv.DictWriter(csvfile, dialect="excel", fieldnames=fields)
-    #     csv.writeheader()
-    #     for row in rows:
-    #         csv.writerow(row)
+    with open(outfile, "w", newline="") as csvfile:
+        fields = list()
+        for row in rows:
+            for key in row.keys():
+                if key not in fields:
+                    fields.append(key)
+        out = csv.DictWriter(csvfile, dialect="excel", fieldnames=fields)
+        out.writeheader()
+        for row in rows:
+            out.writerow(row)
 
     print("Wrote: %s" % outfile)