You said you need JDBC support in ha_connect/Connect Engine? Then it has to be enabled on that also because currently it's not supported and there is no JNI/JDBC support in Connect Enginen it's worth of nothing to have jar file as there is no endpoint for it.

Tuukka... I'm not sure what you mean. We installed the Connect package on Debian and only use it for Connect JDBC... the only thing we needed to do to get it to work was grab the JavaWrappers.jar from an older install.

Connect engine does have JNI/JDBC support if you were not aware... we have been using it for years now.
You docs:
https://mariadb.com/kb/en/connect-jdbc-table-type-accessing-tables-from-another-dbms/

rdyas yes that was what I was after as Connect Engine needs support for JDBC/Java connect when linking. Hmm.. strange that is works without but it's kind of good thing. As Debian does not want just 'binary' blobs if source is available so it makes it little more complicated but in PR we'll sort it out.

otto should we get this on target or should I take this bug ownership?

illuusio I will dive into this over the weekend and try to finalize https://github.com/MariaDB/server/pull/1890 over the weekend.