There are potentially two results that come back from a communications operation: (1) is the Result field in the message header itself. The other is the result of the operation. The distinction is not clear in the code.
I propose that this Result parameter should be the result of the operation. Thus, it should be removed from calls where there is no net result. The Result field in the message header should indicate if the communications was successful; either this needs to be reported separately or perhaps it should just be removed (since the operations always succeed in the current model).
Maybe there's a second issue here about how to fail the message processing for some reason, but for now I don't see much benefit to having multiple "layers" of error being possible. Some of this is a vestige of earlier communications models (where the message could fail independent of the operation).
There are potentially two results that come back from a communications operation: (1) is the Result field in the message header itself. The other is the result of the operation. The distinction is not clear in the code.
I propose that this Result parameter should be the result of the operation. Thus, it should be removed from calls where there is no net result. The Result field in the message header should indicate if the communications was successful; either this needs to be reported separately or perhaps it should just be removed (since the operations always succeed in the current model).
Maybe there's a second issue here about how to fail the message processing for some reason, but for now I don't see much benefit to having multiple "layers" of error being possible. Some of this is a vestige of earlier communications models (where the message could fail independent of the operation).