Send messages during connect immediately rather than only after we’ve
performed our reverse DNS lookup.
Stop logging client reads/writes.
1.11.0 (2019-01-01)
No longer automatically rehash once a week. I changed my mind about this!
1.10.0 (2018-08-18)
Rehashing now reloads the TLS certificate and key.
Rehashing now automatically happens once a week. This is so we load new
certificates.
Require valid TLS certificates on outbound TLS connections. This means
servers we connect to must have valid certificates that match the name we
connect to them as.
1.9.0 (2018-07-28)
Started tracking changes in a changelog.
If a message is invalid, send a notice to opers about it rather than just
log. This is to catch bugs arising from this behaviour.
Send a notice to opers if there is an unprocessed buffer after a read
error. This is again to catch bugs from this behaviour.
Failing to set read deadlines now logs rather than triggers client quit.
This is to allow for the read to happen which can have data in the
buffer.
Delay before flagging the client as dead if there is a write error. This
is to give us a window to read anything further from the client that
might be available.