-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 On 16/04/14 19:34, Michael Rogers wrote: > This variant doesn't require retransmissions or nacks. A message > may be acked by some but not all of the members in one or more > rounds before it's finally acked by everyone in the same round and > committed.
After sleeping on the ideas I posted yesterday I've realised they don't work. There are at least two serious problems: * A member can send inconsistent acks to other members, in order to make them believe that different messages have been acked by everyone * There's a race condition between everyone receiving everyone else's acks for a message, and everyone receiving a better message for the same slot Sorry for posting before sleeping. :-) However, I still think this cut-down version of the problem is worth pursuing. Cheers, Michael -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.12 (GNU/Linux) iQEcBAEBCAAGBQJTT5mhAAoJEBEET9GfxSfMlzkH/R9GlHeJnF2srNXuTR5D8q4c 6mLkU5Zwdd0zczAJR4oKfKrTGc6fAuv/gQj8AHsyD63leIZxhRytsbh/pMUH3oXF wgz/f8ZQOxwSJS9W8EJoV3FLckFfoQXEqpo5R9nBgJtNFgEDM+8KwSBzG193ejIK tTzAv+VR+/qOnjhN/8dhNgVr+2MsVypRcmo7kjSxxjf+BjPPqsqTCOlOi2L/hS+y mSTEaNO/OwW5hN9BIM+l6EseQjVlEaLfsH8cbrOGHF25dGmUiIaY9Strf6Cnsw7Z rzeCXx0ZkFsw2wH0cd/u5UMbHGF5VmVrlMpSwQCbk61FRBA6MUkFep52GoRuojk= =2Weu -----END PGP SIGNATURE----- _______________________________________________ Messaging mailing list [email protected] https://moderncrypto.org/mailman/listinfo/messaging
