diff --git a/modules/ovs/__init__.py b/modules/ovs/__init__.py
new file mode 100644
index 00000000..fcbe95dc
--- /dev/null
+++ b/modules/ovs/__init__.py
@@ -0,0 +1,24 @@
+# -*- coding: utf-8 -*-
+
+# Copyright(C) 2013      Vincent A
+#
+# This file is part of weboob.
+#
+# weboob is free software: you can redistribute it and/or modify
+# it under the terms of the GNU Affero General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# weboob is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+# GNU Affero General Public License for more details.
+#
+# You should have received a copy of the GNU Affero General Public License
+# along with weboob. If not, see <http://www.gnu.org/licenses/>.
+
+
+from .backend import OvsBackend
+
+
+__all__ = ['OvsBackend']
diff --git a/modules/ovs/backend.py b/modules/ovs/backend.py
new file mode 100644
index 00000000..a3c7a844
--- /dev/null
+++ b/modules/ovs/backend.py
@@ -0,0 +1,104 @@
+# -*- coding: utf-8 -*-
+
+# Copyright(C) 2013      Vincent A
+#
+# This file is part of weboob.
+#
+# weboob is free software: you can redistribute it and/or modify
+# it under the terms of the GNU Affero General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# weboob is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+# GNU Affero General Public License for more details.
+#
+# You should have received a copy of the GNU Affero General Public License
+# along with weboob. If not, see <http://www.gnu.org/licenses/>.
+
+from weboob.tools.backend import BaseBackend, BackendConfig
+from weboob.tools.browser import BrowserForbidden
+from weboob.tools.value import Value, ValueBackendPassword
+from weboob.capabilities.messages import ICapMessages, ICapMessagesPost, Message
+
+from .browser import OvsBrowser
+
+
+__all__ = ['OvsBackend']
+
+
+class OvsBackend(BaseBackend, ICapMessages, ICapMessagesPost):
+    NAME = 'ovs'
+    DESCRIPTION = u'OnVaSortir website. Handles private messages only'
+    MAINTAINER = u'Vincent A'
+    EMAIL = 'dev@indigo.re'
+    VERSION = '0.h'
+
+    CONFIG = BackendConfig(Value('username',                label='Username', default=''),
+                           ValueBackendPassword('password', label='Password', default=''),
+                           Value('city',                    label='City (subdomain)', default='paris'))
+    # TODO keep list of possible cities
+
+    BROWSER = OvsBrowser
+
+    STORAGE = {'seen': {}}
+
+    def create_default_browser(self):
+        return self.create_browser(self.config['city'].get(),
+                                   self.config['username'].get(),
+                                   self.config['password'].get(),
+                                   parser='raw')
+
+    def iter_threads(self):
+        with self.browser:
+            for thread in self.browser.iter_threads_list():
+                yield thread
+
+    def get_thread(self, id):
+        with self.browser:
+            thread = self.browser.get_thread(id)
+
+            messages = [thread.root] + thread.root.children
+            for message in messages:
+                if not self.storage.get('seen', message.full_id, default=False):
+                    message.flags |= Message.IS_UNREAD
+
+            return thread
+
+    def iter_unread_messages(self):
+        with self.browser:
+            for thread in self.iter_threads():
+                # TODO reuse thread object?
+                thread2 = self.get_thread(thread.id)
+                messages = [thread2.root] + thread2.root.children
+                for message in messages:
+                    if message.flags & Message.IS_UNREAD:
+                        yield message
+        # TODO implement more efficiently by having a "last weboob seen" for
+        # a thread and query a thread only if "last activity" returned by web
+        # is later than "last weboob seen"
+
+    def set_message_read(self, message):
+        self.storage.set('seen', message.full_id, True)
+        self.storage.save()
+
+    def post_message(self, message):
+        if not self.browser.username:
+            raise BrowserForbidden()
+
+        with self.browser:
+            thread = message.thread
+
+            if message.parent:
+                # ovs.<threadid>@*
+                self.browser.post_to_thread(thread.id, message.title, message.content)
+            else:
+                # ovs.<recipient>@*
+                self.browser.create_thread(thread.id, message.title, message.content)
+
+# FIXME known bug: parsing is done in "boosted mode" which is automatically disable after some time, the "boosted mode" should be re-toggled often
+
+# TODO support outing comments, forum messages
+# TODO make an ICapOuting?
+
diff --git a/modules/ovs/browser.py b/modules/ovs/browser.py
new file mode 100644
index 00000000..eda55f24
--- /dev/null
+++ b/modules/ovs/browser.py
@@ -0,0 +1,102 @@
+# -*- coding: utf-8 -*-
+
+# Copyright(C) 2013      Vincent A
+#
+# This file is part of weboob.
+#
+# weboob is free software: you can redistribute it and/or modify
+# it under the terms of the GNU Affero General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# weboob is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+# GNU Affero General Public License for more details.
+#
+# You should have received a copy of the GNU Affero General Public License
+# along with weboob. If not, see <http://www.gnu.org/licenses/>.
+
+
+from weboob.tools.browser import BaseBrowser, BrowserIncorrectPassword
+from weboob.tools.parsers.iparser import IParser
+import BeautifulSoup
+
+from .pages import PagePrivateThreadsList, PagePrivateThread, PageLogin, PageIndex, DummyPage, PageUserProfile
+
+
+__all__ = ['OvsBrowser']
+
+
+class SoupParser(IParser):
+    def parse(self, data, encoding=None):
+        return BeautifulSoup.BeautifulSoup(data.read().decode(encoding or 'utf-8'), convertEntities=BeautifulSoup.BeautifulStoneSoup.ALL_ENTITIES)
+
+
+class OvsBrowser(BaseBrowser):
+    PROTOCOL = 'http'
+    DOMAIN = 'paris.onvasortir.com'
+    ENCODING = 'cp1252'
+
+    def __init__(self, city, username, password, *a, **kw):
+        self.DOMAIN = '%s.onvasortir.com' % city
+        self.PAGES = {
+            '%s://%s/' % (self.PROTOCOL, self.DOMAIN): PageIndex,
+
+            r'%s://%s/message_read.php\?Id=.+' % (self.PROTOCOL, self.DOMAIN): PagePrivateThread,
+
+            '%s://%s/vue_messages_recus.php' % (self.PROTOCOL, self.DOMAIN): PagePrivateThreadsList,
+            '%s://%s/vue_messages_envoyes.php' % (self.PROTOCOL, self.DOMAIN): PagePrivateThreadsList,
+
+            '%s://%s/page_action_connect.php' % (self.PROTOCOL, self.DOMAIN): PageLogin,
+
+            r'%s://%s/\?Langue=EN' % (self.PROTOCOL, self.DOMAIN): DummyPage,
+            '%s://%s/page_action_boost.php' % (self.PROTOCOL, self.DOMAIN): DummyPage,
+            '%s://%s/vue_profil_all.php.php' % (self.PROTOCOL, self.DOMAIN): DummyPage,
+            r'%s://%s/message_msg_envoi_ok.php\?.*' % (self.PROTOCOL, self.DOMAIN): DummyPage,
+            '%s://%s/message_action_envoi.php' % (self.PROTOCOL, self.DOMAIN): DummyPage,
+
+            r'%s://%s/profil_read.php\?.+' % (self.PROTOCOL, self.DOMAIN): PageUserProfile,
+        }
+
+        kw['parser'] = SoupParser()
+        BaseBrowser.__init__(self, username, password, *a, **kw)
+
+    def iter_threads_list(self):
+        self.location('/vue_messages_recus.php')
+        assert self.is_on_page(PagePrivateThreadsList)
+        for thread in self.page.iter_threads_list():
+            yield thread
+
+        self.location('/vue_messages_envoyes.php')
+        assert self.is_on_page(PagePrivateThreadsList)
+        for thread in self.page.iter_threads_list():
+            yield thread
+
+    def get_thread(self, _id):
+        self.location('/message_read.php?Id=%s&AffMsg=all' % _id)
+        assert self.is_on_page(PagePrivateThread)
+        return self.page.get_thread(_id)
+
+    def login(self):
+        assert not self.is_logged()
+
+        self.page.login(self.username, self.password)
+        if not self.is_logged():
+            raise BrowserIncorrectPassword()
+        self.location('/?Langue=EN')
+        self.location('/page_action_boost.php')
+        self.location('/')
+
+    def is_logged(self):
+        return (self.is_on_page(DummyPage) or self.page.is_logged())
+
+    def post_to_thread(self, thread_id, subject, body):
+        self.location('/message_read.php?Id=%s' % thread_id.encode(self.ENCODING)) # FIXME
+        assert self.is_on_page(PagePrivateThread)
+        self.page.post_to_thread(thread_id, subject, body)
+
+    def create_thread(self, recipient, subject, body):
+        self.location('/profil_read.php?%s' % recipient.encode(self.ENCODING)) # FIXME
+        assert self.is_on_page(PageUserProfile)
+        self.page.create_thread(recipient, subject, body)
diff --git a/modules/ovs/favicon.png b/modules/ovs/favicon.png
new file mode 100644
index 00000000..0783b9ce
Binary files /dev/null and b/modules/ovs/favicon.png differ
diff --git a/modules/ovs/ovsparse.py b/modules/ovs/ovsparse.py
new file mode 100644
index 00000000..a16f44cd
--- /dev/null
+++ b/modules/ovs/ovsparse.py
@@ -0,0 +1,121 @@
+#!/usr/bin/env python
+
+# Copyright(C) 2013      Vincent A
+#
+# This file is part of weboob.
+#
+# weboob is free software: you can redistribute it and/or modify
+# it under the terms of the GNU Affero General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# weboob is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+# GNU Affero General Public License for more details.
+#
+# You should have received a copy of the GNU Affero General Public License
+# along with weboob. If not, see <http://www.gnu.org/licenses/>.
+
+
+import BeautifulSoup
+import datetime
+
+
+def nearest_parent(node, expected):
+    return node.findParent(expected)
+
+    while node and node.name != expected:
+        node = node.parent
+    return node
+
+def all_text_recursive(node):
+    return ''.join(node.findAll(text=True))
+
+def all_next_siblings(node):
+    ret = []
+    while node:
+        ret.append(node)
+        node = node.nextSibling
+    return ret
+
+def parse_date_from_site(sitedate):
+    parts = sitedate.split() # [d, m, Y, '-', 'H:M:S']
+    if len(parts[0]) == 1:
+        parts[0] = '0%s' % parts[0]
+    months = {'january': '01', 'february': '02', 'march': '03', 'april': '04', 'may': '05', 'june': '06', 'july': '07', 'august': '08', 'september': '09', 'october': '10', 'november': '11', 'december': '12'}
+    parts[1] = months[parts[1].lower()]
+    del parts[3]
+    return datetime.datetime.strptime(' '.join(parts), '%d %m %Y %H:%M:%S')
+
+def image_to_text(src):
+    smileys = {'chat/e/grin.gif': ':D',
+               'chat/e/unsure.gif': ':s',
+               'chat/e/smile.gif': ':)',
+               'chat/e/shocked.gif': ':|',
+               'chat/e/sad.gif': ':(',
+               'chat/e/huh.gif': ':h',
+               'chat/e/suprised.gif': ':o',
+               'chat/e/cool.gif': 'B)',
+               'chat/e/redface.gif': ':red',
+               'chat/e/confused.gif': ':=',
+               'chat/e/razz.gif': ':p',
+               'chat/e/wink.gif': ';)',
+               'chat/e/mad.gif': ':x',
+               'chat/e/rolleyes.gif': ':b',
+               'chat/e/lol.gif': ':lol',
+               'chat/e/wub.gif': ':$',
+               'chat/e/bouche.gif': ':K',
+               'chat/e/sick.gif': '+o('}
+
+    return smileys.get(src)
+
+def html_message_to_text(nodes):
+    parts = []
+
+    for node in nodes:
+        if isinstance(node, BeautifulSoup.NavigableString):
+            parts.append(unicode(node).replace('\r', ''))
+        elif node.name == 'img':
+            parts.append(image_to_text(node['src']))
+        elif node.name == 'a':
+            parts.append(node['href'])
+        elif node.name == 'br':
+            parts.append('\n')
+        else:
+            assert not ('%s not supported' % node.name)
+
+    return ''.join(parts)
+
+def create_unique_id(proposed_id, used_ids):
+    if proposed_id not in used_ids:
+        return proposed_id
+
+    def make_id(base, index):
+        return '%s-%s' % (base, index)
+
+    index = 1
+    while make_id(proposed_id, index) in used_ids:
+        index += 1
+
+    return make_id(proposed_id, index)
+
+# public
+
+def private_message_form_fields(document):
+    ret = {}
+    form = document.find('form', attrs={'name': 'envoimail'})
+
+    def set_if_present(name):
+        item = form.find('input', attrs={'name': name})
+        if item:
+            ret[name] = item['value']
+
+    set_if_present('Pere')
+    set_if_present('Sortie')
+    set_if_present('Dest')
+    set_if_present('Titre')
+    return ret
+
+def is_logged(document):
+    return (not document.find('form', attrs={'name': 'connection'}))
diff --git a/modules/ovs/pages.py b/modules/ovs/pages.py
new file mode 100644
index 00000000..bfa75645
--- /dev/null
+++ b/modules/ovs/pages.py
@@ -0,0 +1,177 @@
+# -*- coding: utf-8 -*-
+
+# Copyright(C) 2013      Vincent A
+#
+# This file is part of weboob.
+#
+# weboob is free software: you can redistribute it and/or modify
+# it under the terms of the GNU Affero General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# weboob is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+# GNU Affero General Public License for more details.
+#
+# You should have received a copy of the GNU Affero General Public License
+# along with weboob. If not, see <http://www.gnu.org/licenses/>.
+
+
+import datetime
+import re
+import urllib
+from weboob.tools.browser import BasePage
+from weboob.capabilities.messages import Message, Thread
+
+import ovsparse
+
+__all__ = ['PagePrivateThreadsList', 'PagePrivateThread', 'PageLogin', 'PageIndex', 'DummyPage', 'PagePostMessage', 'PageUserProfile']
+
+
+class OvsPage(BasePage):
+    def is_logged(self):
+        return ovsparse.is_logged(self.document)
+
+    def login(self, username, password):
+        self.browser.select_form(name='connection')
+        self.browser['Pseudo'] = username.encode(self.browser.ENCODING)
+        self.browser['Password'] = password.encode(self.browser.ENCODING)
+        self.browser['Retenir'] = ['ok']
+        self.browser.submit(nologin=True)
+
+
+class PagePrivateThreadsList(OvsPage):
+    def iter_threads_list(self):
+        # site is sorted from latest to oldest
+        for message_a in reversed(self.document.findAll('a', href=re.compile(r'message_read.php\?'))):
+            ovs_id = re.search(r'Id=(\d+)', message_a["href"]).group(1)
+            id_ = ovs_id
+
+            thread = Thread(id_)
+            thread.title = ovsparse.all_text_recursive(message_a)
+            thread.flags = Thread.IS_DISCUSSION
+
+            #~ parent_tr = message_a.findParent('tr')
+            #~ username = all_text_recursive(parent_tr.find('a', href=re.compile(r'profil_read.php\?.*')))
+            #~ notread_self = (parent_tr.get('class') == 'newmails')
+            #~ notread_other = (parent_tr.find('span', **{'class': 'new_sortiepartenaire'}) is not None)
+
+            yield thread
+
+
+class PagePrivateThread(OvsPage):
+    def get_thread(self, _id):
+        thread = Thread(_id)
+
+        thread.title = self.document.find('div', 'PADtitreBlanc_txt').find('center').string
+        thread.flags = Thread.IS_DISCUSSION
+        root = True
+
+        for message in self._get_messages(thread):
+            if root:
+                message.children = []
+                thread.root = message
+                thread.date = message.date
+                message.title = thread.title
+                root = False
+            else:
+                message.title = 'Re: %s' % thread.title
+                message.children = []
+                message.parent = thread.root
+                thread.root.children.append(message)
+
+        return thread
+
+    def _get_messages(self, thread):
+        thread_div = self.document.find(True, 'PADpost_txt')
+        used_ids = set()
+
+        rcpt = self.document.find('input', attrs={'type': 'hidden', 'name': 'Dest'})['value']
+        sender_to_receiver = {rcpt: self.browser.username, self.browser.username: rcpt}
+        # site is sorted from latest to oldest message
+        for message_table in reversed(thread_div.findAll('table')):
+            for td in message_table.findAll('td'):
+                profile_a = td.find('a', href=re.compile(r'profil_read.php\?.*'))
+                if not profile_a:
+                    continue
+
+                first_br = td.find('br')
+                assert first_br.nextSibling.name == 'br'
+                text_nodes = ovsparse.all_next_siblings(first_br.nextSibling.nextSibling) # TODO
+                #~ print text_nodes
+
+            # date will be used as id
+            sitedate = profile_a.findParent('div').find(text=re.compile(',.*')).replace(', ', '')
+            sysdate = ovsparse.parse_date_from_site(sitedate)
+            compactdate = datetime.datetime.strftime(sysdate, '%Y%m%dT%H%M%S')
+
+            # but make it unique
+            msg_id = ovsparse.create_unique_id(compactdate, used_ids)
+            used_ids.add(msg_id)
+
+            message = Message(thread, msg_id)
+
+            message.sender = re.search(r'\?(.+)', profile_a['href']).group(1)
+            message.receivers = [sender_to_receiver[message.sender]]
+
+            message.date = sysdate
+
+            message.content = ovsparse.html_message_to_text(text_nodes)
+
+            notread_self = bool(td.find('span', 'ColorSurligne'))
+            notread_other = bool(td.find('span', 'new_sortiepartenaire'))
+            if notread_other or notread_self:
+                message.flags |= Message.IS_NOT_RECEIVED
+            else:
+                message.flags |= Message.IS_RECEIVED
+
+            yield message
+
+    def post_to_thread(self, thread_id, subject, body):
+        form = ovsparse.private_message_form_fields(self.document)
+        recode_dict(form, self.browser.ENCODING)
+        form['Message'] = body.encode(self.browser.ENCODING)
+        self.browser.location('/message_action_envoi.php', urllib.urlencode(form))
+
+        # html code is so broken that mechanize won't parse the forms
+        #~ self.browser.select_form('envoimail')
+        #~ self.browser['Message'] = body.encode(self.browser.ENCODING)
+        #~ self.browser['Pere'] = thread_id.encode(self.browser.ENCODING)
+        #~ self.browser['Titre'] = subject.encode(self.browser.ENCODING)
+        #~ self.browser.submit()
+
+
+class PageLogin(BasePage):
+    pass
+
+
+class PageIndex(OvsPage):
+    pass
+
+
+class DummyPage(BasePage):
+    pass
+
+
+class PagePostMessage(OvsPage):
+    pass
+
+
+class PageUserProfile(OvsPage):
+    def create_thread(self, recipient, subject, body):
+        form = ovsparse.private_message_form_fields(self.document)
+        recode_dict(form, self.browser.ENCODING)
+        form['Message'] = body.encode(self.browser.ENCODING)
+        form['Titre'] = subject.encode(self.browser.ENCODING)
+        self.browser.location('/message_action_envoi.php', urllib.urlencode(form))
+
+        #~ self.browser.select_form('envoimail')
+        #~ self.browser['Titre'] = subject.encode(self.browser.ENCODING)
+        #~ self.browser['Message'] = body.encode(self.browser.ENCODING)
+        #~ self.browser.submit()
+
+
+def recode_dict(dict_, encoding):
+    for k in dict_:
+        dict_[k] = dict_[k].encode(encoding)
diff --git a/modules/ovs/test.py b/modules/ovs/test.py
new file mode 100644
index 00000000..b479641a
--- /dev/null
+++ b/modules/ovs/test.py
@@ -0,0 +1,28 @@
+# -*- coding: utf-8 -*-
+
+# Copyright(C) 2013      Vincent A
+#
+# This file is part of weboob.
+#
+# weboob is free software: you can redistribute it and/or modify
+# it under the terms of the GNU Affero General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# weboob is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+# GNU Affero General Public License for more details.
+#
+# You should have received a copy of the GNU Affero General Public License
+# along with weboob. If not, see <http://www.gnu.org/licenses/>.
+
+
+from weboob.tools.test import BackendTest
+
+
+class OvsTest(BackendTest):
+    BACKEND = 'ovs'
+
+    def test_ovs(self):
+        raise NotImplementedError()