Various experience and opinion based changes

Changes

• The filewriter will now only run in the job pool mode. This means that the user is required to supply a job pool topic in the filewriter .ini configuration flag using the --job-pool-uri option.
• Improved log messages and thread names. Done to aid debugging.
• Changed fetch.message.max.bytes Kafka variable back to its default value as it was causing timeouts.
• Fix of bug in setting up the console logger interface.
• Warning (log) messages will now be produced if the first message from a source has a data type different than that configured for the current writer module instance. This has been implemented for the f142, senv and ADAr modules.
• The Grafana metrics prefix now has the form "kafka-to-nexus.hostname.service_name" if the service name is set. If not, it has the form "kafka-to-nexus.hostname.service_id".
• Fix ordering of elements in static data.
• Added functionality for automatically instantiating extra writer modules.
• The writer modules f142, senv and tdct will now automatically also instantiate ep00 writer modules at the same location. Disable this by setting the enable_epics_con_status config option to false.
• The --abort-on-uninitialised-stream command line option has been removed. An error in the JSON code for initialising a stream will now always cause an error that will stop further initialisation.
• Added documentation.
• Potentially fixed a bug where the file-writer gets into a bad state.
• Removed automatic "start_time" and "end_time" metadata fields.
• Simplified system test code.