Dear authors,
Your work is impressive, and here are some comments I have. 1. In 2nd paragraph of the abstract, “it may be needed to discover an ALTO server outside…”, who is the “it”? Who is the initiator of the discovery? It can be by the server or the client, corresponding to approach 2.1, 2.2 (by servers) and 2.4 (by clients) in section 1.1.1. What is the difference in protocol between these two initiators of the discovery? 2. What is the detailed algorithm for the reverse DNS lookup? Only the rough procedure is given in the draft. However, considering that the searching space will be large, an optimized algorithm for speeding up is needed. 3. With regard to implementation, how to deploy the database and where to store the DNS information? 4. Suggestion. Some of the security issues can be solved by allocating different DNS servers with different policies to the queries. In a result, for instance, operator “A” can discover and get the information from the server of “B”, whereas “B” can not discover “A”. 5. Does the protocol support successive discovery? For example, operator “A” need the information from the server of “C”, but it can only discover “B”, then “B” will help “A” to discover “C”, and relays the information to “A”. 6. In 2nd paragraph of the section 1.1.2, besides the legal liability, the synchronization between different servers is also an inevitable challenge. Non-ideal synchronization will lead to conflicts and other issues. 7. In the 1st paragraph of the section 1.3, “a new registration administration authority would have to be established”, but why is it a problem? The reason is not well interpreted here. 8. With regard to the endpoint service (section 3.2), as for the destination IP addresses, (D1, D2, D3, ...), what if the destinations are in different domains? Then will the server start another discovery for such new domains? 9. Language typos. The 1st sentence in the 1st paragraph of the section 2.3. The 2st sentence in the 1st paragraph of the section 3.3, “ressources” ==> “resources”. Hope these are helpful. Best, Geng Computer Science, Yale
_______________________________________________ alto mailing list [email protected] https://www.ietf.org/mailman/listinfo/alto
