I don't think we currently error if the user specifices a content-length header and then writes more bytes to the response. I think this is something we could detect and error. Some http clients have a hard time handling this case (especially with pipelining) and would be nice if we could help with reducing the probability of this occuring.