[websocket-dev] Connection closure behaviour

[Mark Thomas <markt@xxxxxxxxxx>]

Hi,

In researching a PR[1] I noticed that the TCK is attempting to send
messages in message annotated with @OnClose[2] although the associated
client never tests for that message being received (or not being
received) by the client.

I think the tests can be made more robust here but before starting down
that path, I wanted to check that there was consensus on the expected
behaviour in this case:

Is it legal to send/start a new message in a method annotated with @OnClose?

Give the following from RFC 6455:

- 6.1   The session must be OPEN to send a message
- 7.1.3 Upon either sending or receiving a Close control frame, it is
        said that The WebSocket Closing Handshake is Started and that
        the WebSocket connection is in the CLOSING state.

I do not believe it is legal for an endpoint that has received (and
processed) a CLOSE from to attempt to send a DATA frame.

For locally closed sessions the Jakarta WebSocket 2.0 specification
states (2.1.5):
If the session is closed locally, the implementation must attempt to
send the WebSocket close frame prior to calling the onClose() method of
the WebSocket endpoint.

This is consistent with RFC 6455 and implies messages may not be sent at
that point.

My only concern is with the text in section 4.5 which states:

The container notifies the [@OnClose] method before the connection is
brought down [WSC-4.5-1].

4.5 could be read as allowing messages to be sent from the @OnClose
method when a close control frame is received. Such a reading does not
appear to be consistent with RFC 6455. Expanding that text to read:

The container notifies the [@OnClose] method after the Close control
frame has been received but before the network connection is brought
down [WSC-4.5-1].

would, I think, be clearer and consistent with RFC 6455.

Before we start thinking about how closely the TCK currently tests /
follows the above, do the people here think I have interpreted the
specifications correctly? If not, what did I miss and what is the
correct interpretation?

Thanks,

Mark


[1] https://github.com/eclipse-ee4j/jakartaee-tck/pull/357
[2]
https://github.com/eclipse-ee4j/jakartaee-tck/blob/master/src/com/sun/ts/tests/websocket/ee/jakarta/websocket/session/WSTestServer.java#L140
[3]
https://github.com/eclipse-ee4j/jakartaee-tck/blob/master/src/com/sun/ts/tests/websocket/ee/jakarta/websocket/session/WSClient.java