[sr-dev] Kill t_relay() return codes. Long life to failure_route.

[Juha Heinanen]

Iñaki Baz Castillo writes:

 > So I wonder, shouldn't be better if we forgot the t_relay() codes and
 > instead handle the error of the winning branch (even if it failed or
 > it was no branch) in failure_route?

inaki,

this has been discussed before and then the response has been that it is
impossible to enter failure route if transaction has not been created.
so there has not been any other choice that to live with this terrible
mess.  now calling drop() in branch route seems to add one more
complication and produce errors to syslog.

 > Then any code we actually run into "if ! t_relay() {  ... }" could be
 > executed in failure_route and we get an easier and solid way to handle
 > errors (instead of mixing t_relay negative values, local generated
 > negative responses in failure_route and so).

this has been proposed before, and the answer has been "can't be done".

-- juha