diff --git a/gawa/blog/models.py b/gawa/blog/models.py
index 395ec6d..0bc6f2c 100644
--- a/gawa/blog/models.py
+++ b/gawa/blog/models.py
@@ -228,72 +228,28 @@ class BlogPost(Searchable):
 
         files = [f for f in os.listdir(data_dir) if (
             data_dir.joinpath(f)).is_file()]
-        logger.debug(f"discovered files: {files}")
 
-        # finding lang and title
-        regex = r"^(en|de)-(.*)\.md"
-        # TODO: only find toml files
+        # find the meta file
+        regex = r"^(.*)\.toml"
 
-        # filepath, language codes, slug
-        files = [[f, cls.DEFAULT_LANGS, ""] for f in files]
+        # filepath, slug
+        files = [[f, ""] for f in files]
         for file in files:
             # parse file name
             try:
                 matches = re.match(regex, file[0])
                 if matches is None:
-                    logger.warning(
-                        f"Data file '{file[0]}' does not fit to the filename\
-                        regex")
                     files.remove(file)
                 else:
                     current_lang = matches.group(1)
-                    file[1][current_lang] = True
-                    file[2] = matches.group(2)
-            except Exception as e:
-                logger.error(e)
-                files.remove(file)
+                    file[1] = matches.group(1)
 
-        # PERF:
-        # Optimize for single loop, should be doable and better design
-
-        # collapse diffrent versions
-        for file in files:
-            try:
-                if [_f[2] for _f in files].count(file[2]) >= 2:
-                    logger.debug(f"multiple versions of '{file[2]}'")
-                    versions = [_f for _f in files if _f[2] == file[2]]
-                    lang: dict[str, bool] = file[1]
-                    for version in versions:
-                        for key in version[1]:
-                            lang[key] |= version[1][key]
-                else:
-                    # only a single version of this file
-                    continue
-            except Exception as e:
-                logger.error(
-                    f"Could not combine BlogPosts for '{file[0]}': {e}")
-
-        # TODO: not needed when relying on toml files
-        try:
-            # deduplicate
-            _files = []
-            for f in [[_f[1], _f[2]]
-                      for _f in files]:   # dont care about fname
-                if f not in _files:
-                    _files.append(f)
-            files = _files
-            logger.debug(f"to save: {files}")
-        except Exception as e:
-            logger.error(f"Could not dedup BlogPosts: {e}")
-
-        for file in files:
-            try:
-                obj = BlogPost(langs=file[0], slug=file[1])
+                obj = BlogPost(slug=file[1])
                 obj.sync_file()
                 obj.regenerate()
-                obj.save()
             except Exception as e:
                 logger.error(f"Could not create BlogPost for '{file[1]}': {e}")
+                files.remove(file)
 
     class Meta:
         verbose_name = _("blog post")
diff --git a/gawa/start/models.py b/gawa/start/models.py
index 636e86d..71a8d0e 100644
--- a/gawa/start/models.py
+++ b/gawa/start/models.py
@@ -59,7 +59,7 @@ class Searchable(models.Model):
             obj.regenerate()
 
     def __str__(self):
-        return f"{{<{self.__class__.__name__}>\"{self.slug}\"}}"
+        return f"{{<{self.__class__.__name__}>\"{self.title_en}\"}}"
 
     def regenerate(self):
         """