diff --git a/maloja/database/__init__.py b/maloja/database/__init__.py
index fbe57a9..108e3b6 100644
--- a/maloja/database/__init__.py
+++ b/maloja/database/__init__.py
@@ -1,6 +1,8 @@
 # server
 from bottle import request, response, FormsDict
 
+from ..pkg_global import conf
+
 
 # decorator that makes sure this function is only run in normal operation,
 # not when we run a task that needs to access the database
@@ -932,6 +934,9 @@ def get_predefined_rulesets(dbconn=None):
 
 
 def start_db():
+
+	conf.AUX_MODE = True # that is, without a doubt, the worst python code you have ever seen
+
 	# Upgrade database
 	from .. import upgrade
 	upgrade.upgrade_db(sqldb.add_scrobbles)
@@ -950,6 +955,8 @@ def start_db():
 
 	dbstatus['healthy'] = True
 
+	conf.AUX_MODE = False # but you have seen it
+
 	# inform time module about begin of scrobbling
 	try:
 		firstscrobble = sqldb.get_scrobbles(limit=1)[0]
diff --git a/maloja/proccontrol/tasks/import_scrobbles.py b/maloja/proccontrol/tasks/import_scrobbles.py
index 32b14f5..546febd 100644
--- a/maloja/proccontrol/tasks/import_scrobbles.py
+++ b/maloja/proccontrol/tasks/import_scrobbles.py
@@ -34,9 +34,12 @@ def import_scrobbles(inputf):
 	filename = os.path.basename(inputf)
 	importfunc = None
 
+	if re.match(r"recenttracks-.*\.csv", filename):
+		typeid, typedesc = "lastfm", "Last.fm (ghan CSV)"
+		importfunc = parse_lastfm_ghan_csv
 
-	if re.match(r".*\.csv", filename):
-		typeid,typedesc = "lastfm", "Last.fm (benjaminbenben export)"
+	elif re.match(r".*\.csv", filename):
+		typeid,typedesc = "lastfm", "Last.fm (benjaminbenben CSV)"
 		importfunc = parse_lastfm
 
 	elif re.match(r"Streaming_History_Audio.+\.json", filename):
@@ -65,8 +68,8 @@ def import_scrobbles(inputf):
 		importfunc = parse_rockbox
 
 	elif re.match(r"recenttracks-.*\.json", filename):
-		typeid, typedesc = "lastfm", "Last.fm (ghan export)"
-		importfunc = parse_lastfm_ghan
+		typeid, typedesc = "lastfm", "Last.fm (ghan JSON)"
+		importfunc = parse_lastfm_ghan_json
 
 	elif re.match(r".*\.json",filename):
 		try:
@@ -83,7 +86,8 @@ def import_scrobbles(inputf):
 		return result
 
 
-	print(f"Parsing {col['yellow'](inputf)} as {col['cyan'](typedesc)} export")
+	print(f"Parsing {col['yellow'](inputf)} as {col['cyan'](typedesc)} export.")
+	print(col['red']("Please double-check if this is correct - if the import fails, the file might have been interpreted as the wrong type."))
 
 	timestamps = set()
 	scrobblebuffer = []
@@ -410,7 +414,7 @@ def parse_lastfm(inputf):
 				continue
 
 
-def parse_lastfm_ghan(inputf):
+def parse_lastfm_ghan_json(inputf):
 	with open(inputf, 'r') as inputfd:
 		data = json.load(inputfd)
 
@@ -432,6 +436,21 @@ def parse_lastfm_ghan(inputf):
 			}, '')
 
 
+def parse_lastfm_ghan_csv(inputf):
+	with open(inputf, 'r') as inputfd:
+		reader = csv.DictReader(inputfd)
+
+		for row in reader:
+			yield ('CONFIDENT_IMPORT', {
+				'track_title': row['track'],
+				'track_artists': row['artist'],
+				'track_length': None,
+				'album_name': row['album'],
+				'scrobble_time': int(row['uts']),
+				'scrobble_duration': None
+			}, '')
+
+
 def parse_listenbrainz(inputf):
 
 	with open(inputf,'r') as inputfd: