Patch corrections

Patch is a correction package that fixes business-critical issues.

We list here the issues fixed from RTT (release to testing) to RTC (release to customer) and onwards.

To see which version your tenant has, open Communication Panel and go to User Menu > Settings > About. The patch number is the last digit in the Version field.

The latest patch includes fixes from all previous patches of the release. For example, corrections made in the first patch release are also in the second one, and so on.

24Q4.3 (build number 10.0.2404.3)

Sending a chat ended prompt failed if there were over 500 chat queues in the system.

24Q4.4 (build number 10.0.2404.4)

In certain cases, two emails were offered simultaneously to an agent who wasn't the preferred agent for those emails.

24Q4.5 (build number 10.0.2404.5)

In SAP Service Cloud (C4C), customer identification did not work for direct calls to agent's extension number.

24Q4.6 (build number 10.0.2404.6)

We have introduced several improvements and fixes to enhance Communication Panel (CP) phone functionality, ensuring greater reliability, especially when recovering from connection interruptions.

  • Session Initiation Protocol (SIP) recovery has been made faster by immediately checking user media when the WebRTC timer (with a 5- or 30-second interval) is started or restarted.

    In practice, SIP can now reconnect immediately without waiting for the next polling round, provided the network connection and backend are functioning properly.

  • If SIP doesn't recover automatically after the backend reconnects (for example, due to three consecutive SIP registration failures), CP fully reinitializes SIP to ensure a fresh start.

  • SIP logs its debug data directly to the browser console when the debug or trace log level is enabled. Tracing actions related to an asynchronous SIP connection in the CP log is now easier.

  • Reconnecting to the backend has been made faster by trying to reconnect already before starting the reconnection loop with the defined interval (by default 5 seconds).

  • Due to a backend issue, CP failed to update the phone/sip_connected data in the model preventing the phone from recovering properly.

  • A delay has been added to the SIP re-registration to ensure that the phone function is set to true before trying to register again.

    This affects the known Mac issue (3 failed re-registration tries in one second).

  • A rare case where SIP needed to be reinitialized has been fixed. In this case, the phone function is set to false for SIP to reconnect properly.

  • An internal model update retry mechanism caused issues by occasionally updating the following data items up to 30 seconds after the backend reconnected:

    /users/me/properties/phone/sip_connected = false

    /users/me/properties/capabilities/phone = false

    The update retry mechanism now prevents this.

  • A 3000 ms delay has been introduced to allow CP to attempt SIP reconnection. This prevents conflicts with previous disconnection-related events, such as the disconnected event with the error value false.

  • An operating system related action has been added when the backend has reconnected but SIP is not connected automatically in 5 seconds:

    • Mac: we will re-initialize SIP.
    • Windows: we restart WebRTC timer to initialize connection again.

Known issues

The following items are known issues, which we plan to address in the 25Q2 release:

  • If an agent loses phone connection while there is an incoming call, the call is rejected (returned to a queue) and is seen in Dashboard with the disconnection reason noPhone.

  • If an agent loses phone connection while talking to a customer, the call is lost (no voice stream). The agent must either disconnect it (and call back ) or try transferring it to a queue.