Imported from previous forum
[ original email was from John Cameron - jc@cameronsystems.com ]
When BusinessRejectReason indicates "application not available", is it still required to set BusinessRejectRefID in response to messages containing an ID field?
Currently the spec requires that the ref ID field is populated in certain cases. However, in the case of "application not available", I suspect that the intentions of the authors of this part of the spec would be that it should not be required (since there is nothing wrong with the message at all).
I disagree. Populating BusinessRejectRefID provides the recipient a reference to a specific message which they’ve sent for which the Business Message Reject has been sent. A good example would be if a broker sends “application not available” if they can’t forward your order on to the system supporting a particular trading desk. That broker may have one system which supports European orders, another supporting ADRs, etc. Thus if a customer sends them a FIX Order it may be possible to have some orders which can be successfully routed while others fail because “application not available”. In that situation, the customer would definitely want BusinessRejectRefID to identify the specific order(s) which were rejected. Similar analogy for Allocations, etc.
> When BusinessRejectReason indicates "application not available", is it still required to set BusinessRejectRefID in response to messages containing an ID field?
>
> Currently the spec requires that the ref ID field is populated in certain cases. However, in the case of "application not available", I suspect that the intentions of the authors of this part of the spec would be that it should not be required (since there is nothing wrong with the message at all).
>
[ original email was from John Cameron - jc@cameronsystems.com ]
Yes, your example is a good one. It should be required even on application unavailable rejects. Thanks for your usual rapid and well considered response. How do you ever find time to do anything else?!
> I disagree. Populating BusinessRejectRefID provides the recipient a reference to a specific message which they’ve sent for which the Business Message Reject has been sent. A good example would be if a broker sends “application not available” if they can’t forward your order on to the system supporting a particular trading desk. That broker may have one system which supports European orders, another supporting ADRs, etc. Thus if a customer sends them a FIX Order it may be possible to have some orders which can be successfully routed while others fail because “application not available”. In that situation, the customer would definitely want BusinessRejectRefID to identify the specific order(s) which were rejected. Similar analogy for Allocations, etc.
>
> > When BusinessRejectReason indicates “application not available”, is it still required to set BusinessRejectRefID in response to messages containing an ID field?
> >
> > Currently the spec requires that the ref ID field is populated in certain cases. However, in the case of “application not available”, I suspect that the intentions of the authors of this part of the spec would be that it should not be required (since there is nothing wrong with the message at all).
> >
>