diff --git a/fetch_posts.py b/fetch_posts.py
index f1b101e..557d84e 100755
--- a/fetch_posts.py
+++ b/fetch_posts.py
@@ -188,36 +188,7 @@ class PostFetcher:
 		assert outbox['type'] in {'OrderedCollection', 'OrderedCollectionPage'}
 		return outbox
 
-	async def _finger_actor(self, username, instance):
-		# despite HTTP being a direct violation of the WebFinger spec, assume e.g. Tor instances do not support
-		# HTTPS-over-onion
-		finger_url = f'http://{instance}/.well-known/webfinger?resource=acct:{username}@{instance}'
-		async with self._http.get(finger_url) as resp: finger_result = await resp.json()
-		return (profile_url := self._parse_webfinger_result(username, instance, finger_result))
-
-	def _parse_webfinger_result(self, username, instance, finger_result):
-		"""given webfinger data, return profile URL for handle"""
-		def check_content_type(type, ct): return ct == type or ct.startswith(type+';')
-		check_ap = partial(check_content_type, ACTIVITYPUB_CONTENT_TYPE)
-
-		try:
-			# note: the server might decide to return multiple links
-			# so we need to decide how to prefer one.
-			# i'd put "and yarl.URL(template).host == instance" here,
-			# but some instances have no subdomain for the handle yet use a subdomain for the canonical URL.
-			# Additionally, an instance could theoretically serve profile pages over I2P and the clearnet,
-			# for example.
-			return (profile_url := next(
-				link['href']
-				for link in finger_result['links']
-				if link['rel'] == 'self' and check_ap(link['type'])
-			))
-		except StopIteration:
-			# this should never happen either
-			raise RuntimeError(f'fatal: while fingering {username}@{instance}, failed to find a profile URL')
-
 async def amain():
-	import json5 as json
 	import third_party.utils as utils
 	args = utils.arg_parser_factory(description='Fetch posts from all followed accounts').parse_args()
 	config = utils.load_config(args.cfg)