diff --git a/isso/compat.py b/isso/compat.py
index 888afe0..772d561 100644
--- a/isso/compat.py
+++ b/isso/compat.py
@@ -4,16 +4,17 @@ import sys
 PY2K = sys.version_info[0] == 2
 
 if not PY2K:
-    # iterkeys = lambda d: iter(d.keys())
-    # iteritems = lambda d: iter(d.items())
+
+    map, zip, filter = map, zip, filter
 
     text_type = str
     string_types = (str, )
 
     buffer = memoryview
 else:
-    # iterkeys = lambda d: d.iterkeys()
-    # iteritems = lambda d: d.iteritems()
+
+    from itertools import imap, izip, ifilter
+    map, zip, filter = imap, izip, ifilter
 
     text_type = unicode
     string_types = (str, unicode)
diff --git a/isso/core.py b/isso/core.py
index f4721a8..e1636e1 100644
--- a/isso/core.py
+++ b/isso/core.py
@@ -117,6 +117,9 @@ class Mixin(object):
 
 
 def threaded(func):
+    """
+    Decorator to execute each :param func: call in a separate thread.
+    """
 
     def dec(self, *args, **kwargs):
         thread.start_new_thread(func, (self, ) + args, kwargs)
diff --git a/isso/utils/__init__.py b/isso/utils/__init__.py
index cdcf6ad..12216e7 100644
--- a/isso/utils/__init__.py
+++ b/isso/utils/__init__.py
@@ -15,10 +15,10 @@ def anonymize(remote_addr):
     Anonymize IPv4 and IPv6 :param remote_addr: to /24 (zero'd)
     and /48 (zero'd).
 
-    >>> anonymize(u'12.34.56.78')
-    u'12.34.56.0'
-    >>> anonymize(u'1234:5678:90ab:cdef:fedc:ba09:8765:4321')
-    u'1234:5678:90ab:0000:0000:0000:0000:0000'
+    >>> anonymize(u'12.34.56.78')  # doctest: +IGNORE_UNICODE
+    '12.34.56.0'
+    >>> anonymize(u'1234:5678:90ab:cdef:fedc:ba09:8765:4321') # doctest: +IGNORE_UNICODE
+    '1234:5678:90ab:0000:0000:0000:0000:0000'
     """
     try:
         ipv4 = ipaddress.IPv4Address(remote_addr)
diff --git a/isso/utils/http.py b/isso/utils/http.py
index 15cdb86..1237cc7 100644
--- a/isso/utils/http.py
+++ b/isso/utils/http.py
@@ -9,74 +9,17 @@ try:
 except ImportError:
     import http.client as httplib
 
-import html5lib
-
 from isso.utils import parse
 
 
-def urlexists(host, path):
+def curl(method, host, path, timeout=3):
 
     host, port, ssl = parse.host(host)
     http = httplib.HTTPSConnection if ssl else httplib.HTTPConnection
 
-    with closing(http(host, port, timeout=3)) as con:
+    with closing(http(host, port, timeout=timeout)) as con:
         try:
-            con.request('HEAD', path)
+            con.request(method, path)
         except (httplib.HTTPException, socket.error):
-            return False
-        return con.getresponse().status == 200
-
-
-def heading(host, path):
-    """Connect to `host`, GET path and start from #isso-thread to search for
-    a possible heading (h1). Returns `None` if nothing found."""
-
-    host, port, ssl = parse.host(host)
-    http = httplib.HTTPSConnection if ssl else httplib.HTTPConnection
-
-    with closing(http(host, port, timeout=15)) as con:
-        con.request('GET', path)
-        html = html5lib.parse(con.getresponse().read(), treebuilder="dom")
-
-    assert html.lastChild.nodeName == "html"
-    html = html.lastChild
-
-    # aka getElementById
-    el = list(filter(lambda i: i.attributes["id"].value == "isso-thread",
-              filter(lambda i: i.attributes.has_key("id"), html.getElementsByTagName("div"))))
-
-    if not el:
-        return "Untitled"
-
-    el = el[0]
-    visited = []
-
-    def recurse(node):
-        for child in node.childNodes:
-            if child.nodeType != child.ELEMENT_NODE:
-                continue
-            if child.nodeName.upper() == "H1":
-                return child
-            if child not in visited:
-                return recurse(child)
-
-    def gettext(rv):
-        for child in rv.childNodes:
-            if child.nodeType == child.TEXT_NODE:
-                yield child.nodeValue
-            if child.nodeType == child.ELEMENT_NODE:
-                for item in gettext(child):
-                    yield item
-
-    while el is not None:  # el.parentNode is None in the very end
-
-        visited.append(el)
-        rv = recurse(el)
-
-        if rv:
-            return ''.join(gettext(rv)).strip()
-
-        el = el.parentNode
-
-    return "Untitled."
-
+            return None
+        return con.getresponse()
diff --git a/isso/utils/parse.py b/isso/utils/parse.py
index 42f79f9..6af6425 100644
--- a/isso/utils/parse.py
+++ b/isso/utils/parse.py
@@ -4,11 +4,21 @@ from __future__ import print_function
 import re
 import datetime
 
+from itertools import chain
+
 try:
     from urlparse import urlparse
 except ImportError:
     from urllib.parse import urlparse
 
+import html5lib
+
+from isso.compat import map, filter, PY2K
+
+if PY2K:  # http://bugs.python.org/issue12984
+    from xml.dom.minidom import NamedNodeMap
+    NamedNodeMap.__contains__ = lambda self, key: self.has_key(key)
+
 
 def timedelta(value):
     """
@@ -63,3 +73,78 @@ def host(name):
     if rv.scheme == 'https' and rv.port is None:
         return (rv.netloc, 443, True)
     return (rv.netloc.rsplit(':')[0], rv.port or 80, rv.scheme == 'https')
+
+
+def title(data, default=u"Untitled."):
+    """
+    Extract <h1> title from web page. The title is *probably* the text node,
+    which is the nearest H1 node in context to an element with the `isso-thread` id.
+
+    >>> title("asdf")  # doctest: +IGNORE_UNICODE
+    u'Untitled.'
+    >>> title('''
+    ... <html>
+    ... <head>
+    ...     <title>Foo!</title>
+    ... </head>
+    ... <body>
+    ...     <header>
+    ...         <h1>generic website title.</h1>
+    ...         <h2>subtile title.</h2>
+    ...     </header>
+    ...     <article>
+    ...         <header>
+    ...             <h1>Can you find me?</h1>
+    ...         </header>
+    ...         <section id="isso-thread">
+    ...         </section>
+    ...     </article>
+    ... </body>
+    ... </html>''')  # doctest: +IGNORE_UNICODE
+    u'Can you find me?'
+    """
+
+    html = html5lib.parse(data, treebuilder="dom")
+
+    assert html.lastChild.nodeName == "html"
+    html = html.lastChild
+
+    # aka getElementById
+    el = list(filter(lambda i: i.attributes["id"].value == "isso-thread",
+              filter(lambda i: "id" in i.attributes,
+                     chain(*map(html.getElementsByTagName, ("div", "section"))))))
+
+    if not el:
+        return default
+
+    el = el[0]
+    visited = []
+
+    def recurse(node):
+        for child in node.childNodes:
+            if child.nodeType != child.ELEMENT_NODE:
+                continue
+            if child.nodeName.upper() == "H1":
+                return child
+            if child not in visited:
+                return recurse(child)
+
+    def gettext(rv):
+        for child in rv.childNodes:
+            if child.nodeType == child.TEXT_NODE:
+                yield child.nodeValue
+            if child.nodeType == child.ELEMENT_NODE:
+                for item in gettext(child):
+                    yield item
+
+    while el is not None:  # el.parentNode is None in the very end
+
+        visited.append(el)
+        rv = recurse(el)
+
+        if rv:
+            return ''.join(gettext(rv)).strip()
+
+        el = el.parentNode
+
+    return default
diff --git a/isso/views/comment.py b/isso/views/comment.py
index d80bb62..d8187ad 100644
--- a/isso/views/comment.py
+++ b/isso/views/comment.py
@@ -6,7 +6,6 @@ import time
 import hashlib
 import logging
 import sqlite3
-import logging
 
 from itsdangerous import SignatureExpired, BadSignature
 
@@ -16,7 +15,7 @@ from werkzeug.exceptions import abort, BadRequest
 from isso.compat import text_type as str
 
 from isso import utils, notify, db
-from isso.utils import http
+from isso.utils import http, parse
 from isso.crypto import pbkdf2
 
 logger = logging.getLogger("isso")
@@ -50,9 +49,6 @@ class requires:
 @requires(str, 'uri')
 def new(app, environ, request, uri):
 
-    if uri not in app.db.threads and not http.urlexists(app.conf.get('general', 'host'), uri):
-        return Response('URI does not exist', 404)
-
     try:
         data = json.loads(request.get_data().decode('utf-8'))
     except ValueError:
@@ -74,11 +70,23 @@ def new(app, environ, request, uri):
     data['mode'] = (app.conf.getboolean('moderation', 'enabled') and 2) or 1
     data['remote_addr'] = utils.anonymize(str(request.remote_addr))
 
+    # extract site's <h1> title
+    if uri not in app.db.threads:
+        for host in app.conf.getiter('general', 'host'):
+            resp = http.curl('HEAD', host, uri)
+            if resp and resp.status == 200:
+                title = parse.title(resp.read())
+                break
+        else:
+            return Response('URI does not exist', 404)
+    else:
+        title = app.db.threads[uri].title
+
     with app.lock:
         if uri not in app.db.threads:
-            app.db.threads.new(uri, http.heading(app.conf.get('general', 'host'), uri))
-            logger.info('new thread: %s -> %s', uri, app.db.threads[uri].title)
-    title = app.db.threads[uri].title
+            app.db.threads.new(uri, title)
+
+    logger.info('new thread: %s -> %s', uri, title)
 
     try:
         with app.lock:
diff --git a/specs/test_comment.py b/specs/test_comment.py
index 6063ac3..1bc8406 100644
--- a/specs/test_comment.py
+++ b/specs/test_comment.py
@@ -18,8 +18,14 @@ from isso import Isso, notify, views, core
 from isso.utils import http
 from isso.views import comment
 
-http.heading = lambda *args: "Untitled."
-http.urlexists = lambda *args: True
+class Dummy:
+
+    status = 200
+
+    def read(self):
+        return ''
+
+http.curl = lambda method, host, path: Dummy()
 
 loads = lambda data: json.loads(data.decode('utf-8'))
 
diff --git a/specs/test_vote.py b/specs/test_vote.py
index 2a617af..95c4d17 100644
--- a/specs/test_vote.py
+++ b/specs/test_vote.py
@@ -10,9 +10,16 @@ from werkzeug.test import Client
 from werkzeug.wrappers import Response
 
 from isso import Isso, notify, utils, core
+from isso.utils import http
 
-utils.heading = lambda *args: "Untitled."
-utils.urlexists = lambda *args: True
+class Dummy:
+
+    status = 200
+
+    def read(self):
+        return ''
+
+http.curl = lambda method, host, path: Dummy()
 
 loads = lambda data: json.loads(data.decode('utf-8'))
 
diff --git a/tox.ini b/tox.ini
index 4d836e2..119a277 100755
--- a/tox.ini
+++ b/tox.ini
@@ -15,6 +15,7 @@ deps =
 deps =
     nose
     ipaddress
+    doctest-ignore-unicode
 commands=
-    nosetests --with-doctest isso/
+    nosetests --with-doctest --with-doctest-ignore-unicode isso/
     nosetests specs/