Imported from previous forum
The 3.0 spec clearly states that the sequence number is NOT incremented for heartbeats or a logon. The same is not stated for Logout. Does the same rule apply here?
Also, if an out of synch on sequence numbers occurs, what would prevent the side which experinced the logout from resequencing intraday?
You are correct that 3.0 never clearly stated it. The minutes from the January 1996 Technical Committee meeting held at DTC in New York did state that sequence numbers should NOT increment for Logouts (FIX 2.7 and 3.0 were the only versions out at that time). You will likely find that most (ambiguity restricts one from saying all) 3.0 implementations do not expect sequence number to increment for the Logout. A rule of thumb is that if you increment sequence number for a message then you need to be able to handle a resend request for that seq num. You obviously wouldn’t want to resend the Logout message itself and I, personally, found very few 3.0 implementations which support Sequence Reset and that there are risks in using Sequence Reset without the Gap Fill feature. Hope that helps.
> The 3.0 spec clearly states that the sequence number is NOT incremented for heartbeats or a logon. The same is not stated for Logout. Does the same rule apply here?
> Also, if an out of synch on sequence numbers occurs, what would prevent the side which experinced the logout from resequencing intraday?
>