--- a/src/pages/common/blog/page_meta.py	Wed Jan 24 09:58:00 2018 +0100
+++ b/src/pages/common/blog/page_meta.py	Wed Jan 24 09:58:00 2018 +0100
@@ -20,6 +20,8 @@
 
 RE_TEXT_URL = re.compile(ur'[^a-zA-Zéèêôà,_]+')
 TEXT_MAX_LEN = 60
+TEXT_WORD_MIN_LENGHT = 4
+URL_LIMIT_MARK = 90  # if canonical URL is longer than that, text will not be appended
 
 def microblog_uri(self, uri_data):
     args = [uri_data[u'path'], uri_data[u'node']]
@@ -192,9 +194,11 @@
 
     # if we know the profile, we use it instead of service + blog (nicer url)
     if target_profile is None:
-        blog_base_url = self.getPageByName(u'blog_view').getURL(service.full(), node or u'@')
+        blog_base_url_item = self.getPageByName(u'blog_view').getURL(service.full(), node or u'@', u'id')
+        blog_base_url_tag = self.getPageByName(u'blog_view').getURL(service.full(), node or u'@', u'tag')
     else:
-        blog_base_url = self.getURLByNames([(u'user', [target_profile]), (u'user_blog', [])])
+        blog_base_url_item = self.getURLByNames([(u'user', [target_profile]), (u'user_blog', [u'id'])])
+        blog_base_url_tag = self.getURLByNames([(u'user', [target_profile]), (u'user_blog', [u'tag'])])
         # we also set the background image if specified by user
         bg_img = yield self.host.bridgeCall(u'asyncGetParamA', u'Background', u'Blog page', u'value', -1, template_data[u'target_profile'])
         if bg_img:
@@ -210,22 +214,29 @@
     template_data[u'items_http_uri'] = items_http_uri = {}
     template_data[u'tags_http_uri'] = tags_http_uri = {}
     for item in items:
-        # we add text from title or body at the end of URL
-        # to make it more readable
-        text = item.title or item.content
-        text = RE_TEXT_URL.sub(u'-', text).lower().strip(u'-')
-        while len(text) > TEXT_MAX_LEN:
-            if u'-' in text:
-                text = text.rsplit(u'-', 1)[0]
+        blog_canonical_url = u'/'.join([blog_base_url_item, utils.quote(item.id)])
+        if len(blog_canonical_url) > URL_LIMIT_MARK:
+            blog_url = blog_canonical_url
+        else:
+            # we add text from title or body at the end of URL
+            # to make it more readable
+            text = item.title or item.content
+            text = RE_TEXT_URL.sub(u' ', text).lower()
+            text = u'-'.join([t for t in text.split() if t and len(t)>=TEXT_WORD_MIN_LENGHT])
+            while len(text) > TEXT_MAX_LEN:
+                if u'-' in text:
+                    text = text.rsplit(u'-', 1)[0]
+                else:
+                    text = text[:TEXT_MAX_LEN]
+            if text:
+                blog_url = blog_canonical_url + u'/' + text
             else:
-                text = text[:TEXT_MAX_LEN]
+                blog_url = blog_canonical_url
 
-        # now we can compute item URL
-        blog_url = u'/'.join([blog_base_url, u'id', utils.quote(item.id), text])
         items_http_uri[item.id] = self.host.getExtBaseURL(request, blog_url)
         for tag in item.tags:
             if tag not in tags_http_uri:
-                tag_url = u'/'.join([blog_base_url, u'tag', utils.quote(tag)])
+                tag_url = u'/'.join([blog_base_url_tag, utils.quote(tag)])
                 tags_http_uri[tag] = self.host.getExtBaseURL(request, tag_url)
     template_data[u'allow_commenting'] = data.get(u'allow_commenting', False)