Why? For the glory of Horta of course!
Download the logs using wget -m and put them in a folder from where
you're going to execute the converter. Put a prepared h2 database
file in the same folder. The database should contain a log table
defined like
this. And
finally do sbt run. The converter recursively finds all files that
match \d{2}\.html and converts them into the database.
You may also pass a path to logs as a first parameter of sbt run:
$ sbt 'run path/to/logs'
NOTE: due to naive implementation of /me messages parsing it is recommended to convert one conference at a time.