diff --git a/oletools/olevba3.py b/oletools/olevba3.py
index 9fb0e76..b122399 100755
--- a/oletools/olevba3.py
+++ b/oletools/olevba3.py
@@ -244,15 +244,15 @@ except ImportError:
                                + "see http://codespeak.net/lxml " \
                                + "or http://effbot.org/zone/element-index.htm")
 
-import thirdparty.olefile as olefile
-from thirdparty.prettytable import prettytable
-from thirdparty.xglob import xglob, PathNotFoundException
-from thirdparty.pyparsing.pyparsing import \
+import oletools.thirdparty.olefile as olefile
+from oletools.thirdparty.prettytable import prettytable
+from oletools.thirdparty.xglob import xglob, PathNotFoundException
+from oletools.thirdparty.pyparsing.pyparsing import \
         CaselessKeyword, CaselessLiteral, Combine, Forward, Literal, \
         Optional, QuotedString,Regex, Suppress, Word, WordStart, \
         alphanums, alphas, hexnums,nums, opAssoc, srange, \
         infixNotation
-import ppt_parser
+import oletools.ppt_parser as ppt_parser
 
 # monkeypatch email to fix issue #32:
 # allow header lines without ":"
@@ -1774,7 +1774,7 @@ def detect_hex_strings(vba_code):
         value = match.group()
         if value not in found:
             decoded = binascii.unhexlify(value)
-            results.append((value, decoded))
+            results.append((value, decoded.decode('utf-8','replace')))
             found.add(value)
     return results
 
@@ -1799,7 +1799,7 @@ def detect_base64_strings(vba_code):
         if value not in found and value.lower() not in BASE64_WHITELIST:
             try:
                 decoded = base64.b64decode(value)
-                results.append((value, decoded))
+                results.append((value, decoded.decode('utf-8','replace')))
                 found.add(value)
             except (TypeError, ValueError) as exc:
                 log.debug('Failed to base64-decode (%s)' % exc)
@@ -1814,7 +1814,7 @@ def detect_dridex_strings(vba_code):
     :param vba_code: str, VBA source code
     :return: list of str tuples (encoded string, decoded string)
     """
-    from thirdparty.DridexUrlDecoder.DridexUrlDecoder import DridexUrlDecode
+    from oletools.thirdparty.DridexUrlDecoder.DridexUrlDecoder import DridexUrlDecode
 
     results = []
     found = set()
@@ -1959,10 +1959,10 @@ class VBA_Scanner(object):
         """
         # join long lines ending with " _":
         self.code = vba_collapse_long_lines(vba_code)
-        self.code_hex = b''
-        self.code_hex_rev = b''
-        self.code_rev_hex = b''
-        self.code_base64 = b''
+        self.code_hex = ''
+        self.code_hex_rev = ''
+        self.code_rev_hex = ''
+        self.code_base64 = ''
         self.code_dridex = ''
         self.code_vba = ''
         self.strReverse = None
@@ -1995,19 +1995,19 @@ class VBA_Scanner(object):
         if 'strreverse' in self.code.lower(): self.strReverse = True
         # Then append the decoded strings to the VBA code, to detect obfuscated IOCs and keywords:
         for encoded, decoded in self.hex_strings:
-            self.code_hex += b'\n' + decoded
+            self.code_hex += '\n' + decoded
             # if the code contains "StrReverse", also append the hex strings in reverse order:
             if self.strReverse:
                 # StrReverse after hex decoding:
-                self.code_hex_rev += b'\n' + decoded[::-1]
+                self.code_hex_rev += '\n' + decoded[::-1]
                 # StrReverse before hex decoding:
-                self.code_rev_hex += b'\n' + binascii.unhexlify(encoded[::-1])
+                self.code_rev_hex += '\n' + str(binascii.unhexlify(encoded[::-1]))
                 #example: https://malwr.com/analysis/NmFlMGI4YTY1YzYyNDkwNTg1ZTBiZmY5OGI3YjlhYzU/
         #TODO: also append the full code reversed if StrReverse? (risk of false positives?)
         # Detect Base64-encoded strings
         self.base64_strings = detect_base64_strings(self.code)
         for encoded, decoded in self.base64_strings:
-            self.code_base64 += b'\n' + decoded
+            self.code_base64 += '\n' + decoded
         # Detect Dridex-encoded strings
         self.dridex_strings = detect_dridex_strings(self.code)
         for encoded, decoded in self.dridex_strings:
@@ -2026,10 +2026,10 @@ class VBA_Scanner(object):
 
         for code, obfuscation in (
                 (self.code, None),
-                (self.code_hex.decode('utf-8','replace'), 'Hex'),
+                (self.code_hex, 'Hex'),
                 (self.code_hex_rev, 'Hex+StrReverse'),
                 (self.code_rev_hex, 'StrReverse+Hex'),
-                (self.code_base64.decode('utf-8', 'replace'), 'Base64'),
+                (self.code_base64, 'Base64'),
                 (self.code_dridex, 'Dridex'),
                 (self.code_vba, 'VBA expression'),
         ):
diff --git a/oletools/ppt_parser.py b/oletools/ppt_parser.py
index e89a037..a36867d 100644
--- a/oletools/ppt_parser.py
+++ b/oletools/ppt_parser.py
@@ -37,7 +37,7 @@ import struct
 import traceback
 import os
 
-import thirdparty.olefile as olefile
+import oletools.thirdparty.olefile as olefile
 import zlib
 
 
diff --git a/setup.py b/setup.py
index 50ca086..8062eee 100755
--- a/setup.py
+++ b/setup.py
@@ -287,11 +287,11 @@ entry_points = {
 def main():
     # TODO: remove this test once all tools are ported to Python 3
     # TODO: warning about Python 2.6
-    if sys.version >= '3.0':
-        s = "Sorry, %s %s requires Python 2.x."
-        print(s % (name, version))
-        sys.exit(1)
-##    if sys.version < required_python_version:
+#     if sys.version >= '3.0':
+#         s = "Sorry, %s %s requires Python 2.x."
+#         print(s % (name, version))
+#         sys.exit(1)
+# ##    if sys.version < required_python_version:
 ##        s = "I'm sorry, but %s %s requires Python %s or later."
 ##        print(s % (name, version, required_python_version))
 ##        sys.exit(1)