The design of new inter-domain optical routing protocols may start from scratch, or on the contrary exploit all the research already developed in IP networks with the Border Gateway Protocol (BGP). Even though the network premises under which BGP was conceived have drastically changed, the pervasive deployment of BGP makes almost impossible its replacement, hence everything indicates that BGP-based routing will remain present in the coming years. In light of this, the approach often used for distributing reachability information and routing inter-domain connections below the IP layer has been to propose extensions to the BGP protocol, what unfortunately exports all well-known BGP weaknesses to these routing scenarios. In this paper we deeply analyze all these problems in order the reader to get a clear idea of the existing limitations inherent to the BGP, before exploring the routing problem in optical networks. Then, focusing on the optical layer we will demonstrate that current optical extensions of BGP do not meet the particular optical layer constraints. We then propose minor, though effective, changes to a path vector protocol overall offering a promising line of work and a simple solution designed to be deployed on a multi-domain and multi-layer scenario.