minor TODO update

TODO

@@ -2,6 +2,8 @@ Bugs :
     When no peer is avalaible, the setup crash without the --no-boot option
     G : see below
     U : this is still a bug to be solved
+    G : since the server is now a node it shouldn't ( the server will always
+        have a peer availabel which is himself )
 
 To be done :
     Use an algorithm to choose which connections to keep and/or establish
@@ -23,6 +25,7 @@ To be done :
     Use a timeout for the peersDB
     G : removing the dead peers should be enough
     U : I was speaking of the server peers DB
+    G : the timeout is the duration of the certificate delivered ( 1y for now )
 
     Specify a lease duration in ForwardViaUPnP
 
@@ -76,7 +79,7 @@ To be discussed:
     G : Well when no peer is available the SQL request ( + next() method ) raise
         a StopIteration exception
     U : Well, I don't think this is a good thing. When not in debug, a node
-        cannot now if their is anyone else already connected
+        cannot now if there is anyone else already connected
 
     G : don't reconnect to server each time we repopulate in peers_db ?
     U : From what I've read on the internet, when you create a server object,