On Feb 10, 5:27 pm, Henrik Schröder <[email protected]> wrote: > I could probably add the above to the project, but then you would have to > convince me that there's a legitimate usecase for noreply. :-)
It's not just use cases, it's bugs. There are times when a response might come back even though the client intended to send a noreply. We had a case where it was, from the server point of view, a malformed request. The client thought it asked for noreply. The server was getting replies, and now the whole thing is out of sequence and you just have to hang up. I'd just think that if there's an error performing an operation, it's OK to say what that error is. We fixed that in the binary protocol. :)
