How to avoid Spark to start if connection is Incomplete?

Hi All,

We are testing Openfire 3.6.3 and spark 2.5.8 (and 2.6.0 Beta 2) to replace our actual Exchange 2000/ Windows Messenger installation. We want to use only as an internal chat (no broadcast, no file transfer, etc), and to use the (Spark) client administration centralized and logging of conversations in the Openfire Server.

All the tests until now was OK, but we just found one problem: when the users login to Spark, if the Response Timeout configured is reached, then the Spark client open the interface but ignore the forced (in Openfire Server) configuration.

As an example; as we disable the broadcast from the server, then when users login to spark, the broadcast icon do not appear (this is ok). But if the Response Timeout configured time is reached (may be because a slow connection or too much network traffic) and the configuration data of the server is not completely transferred to the client, then spark open with default options and show the broadcast icon (see images attached). Also the contacts presence is not correct and it shows as offline, users that are online.

The expected behavior is the following: if the Response Timeout is reached and the forced configuration data provided by the server is not completed, then the client** go back to the login screen** and ask for the user/pass again.

Then, is there a way to do this?

As an alternative, is there a way to change the startup default options in the start of the spark client to hide the undesirable icons at least until the forced server configuration is completed ok?? (for example in the Spark.vmoptions file)



Well, i dont think you can do anything with the vmoptions file. I can of course file this as a ticket for the improvement, but i’m confused on what side it should be fixed. Whether it is Spark which has to repeat requests for the rules from the server until it gets them, or maybe it is Openfire which has to repeat rules push to a client if it gets no response.