Imported from previous forum
[ original email was from Davide Khalil - Reuters ]
Regarding that the clarity of the FIX 4.0 specification on when to reset sequence numbers. The idea is good to define a start and stop time on a per session basis. Is this time GMT, the timezone of the originator or the timezone of the receiver? It would remove ambiguity to simply modifiy the
last sentence of a portion of the specification:
"It is recommended that a new FIX session be established once within each 24 hour period." … "by establishing stop and start time, the timezones of these times between the communicating parties." Ambiguity only extends testing time. There is no reason not to modify the spec.
The spec is also equally unclear regarding the properties of certain Time-in-Force values. Is a DAY order valid in the time zone of the connector vs that of the reciever? How long is GTC. Can an order be outstanding for a year? The more FIX is considered internationally, the more these subtle omissions of specific definition come out.