I appreciate this is probably a little more of an edge case, but I'm rocking docker-readsb-protobuf on an IPv6-first Kubernetes cluster. As services are advertised primarily by their IPv6 address, and this container forcibly sets --net-bind-address to 0.0.0.0 (the IPv4 catchall), anything that tries to dial this container over IPv6 just gets turned away at the door.
Simple proposed solution: Make READSB_NET_BIND_ADDRESS a configurable value overruling this line here:
and potentially consider setting the default to ::, as almost all systems should tolerate it (?), and it should also bind to the wildcard v4 address at the same time.
(if readsb doesn't support binding to :: then by all means close this issue - that's the main reason I haven't done a PR!)
I appreciate this is probably a little more of an edge case, but I'm rocking docker-readsb-protobuf on an IPv6-first Kubernetes cluster. As services are advertised primarily by their IPv6 address, and this container forcibly sets
--net-bind-address
to0.0.0.0
(the IPv4 catchall), anything that tries to dial this container over IPv6 just gets turned away at the door.Simple proposed solution: Make
READSB_NET_BIND_ADDRESS
a configurable value overruling this line here:https://github.com/sdr-enthusiasts/docker-readsb-protobuf/blob/ff0d464a5ef0bdd1f3e6d7b6f44d1601da9d38d5/rootfs/etc/s6-overlay/scripts/readsb#L14
and potentially consider setting the default to
::
, as almost all systems should tolerate it (?), and it should also bind to the wildcard v4 address at the same time.(if readsb doesn't support binding to
::
then by all means close this issue - that's the main reason I haven't done a PR!)