Am Freitag, den 16.09.2011, 23:31 +0200 schrieb Thomas Arp: > netmd_secure_leave_session: No error > netmd_secure_set_track_protection: No error > netmd_secure_enter_session: No error > netmd_secure_send_key_data: No error > netmd_secure_session_key_exchange: No error > netmd_secure_setup_download: No error > netmd_prepare_packets: No error > netmd_secure_send_track: Command rejected > netmd_set_title ("test" to track 52): Not implemented > netmd_exch_message: netmd_poll failed > netmd_secure_commit_track: Command failed, unknown error > netmd_secure_session_key_forget: No error > netmd_secure_leave_session: No error > > So we should break the code on NETMD_COMMAND_FAILED_REJECTED errors. In general, yes. But we should add some exceptions on that: netmd_secure_leave_session is expected to fail if there is no "secured" session available. Recent recorders always indicate success on this command, but you can not count on it. We send that command in the beginning to cancel a previous unterminated session. To make this even more robust, a netmd_secure_session_key_forget command should be added before the leave_session command (errors also ignored). The second thing to consider is that older MD recorders do not implement the "set_track_protection" feature, so a "NOT IMPLEMENTED" error at that stage should also not be treated fatal. When you encounter a "REJECTED" error and abort the transfer, please still terminate the session (session_key_forget and leave_session). Regards, Michael Karcher
Attachment:
signature.asc
Description: This is a digitally signed message part