Investigate the possibility of running compression, etc. in a BoardReader
When we talked with the folks from the Muon g-2 experiment some time ago, we learned that they are planning to compress and/or zero-suppress their data before event building. (They will use MIDAS, not artdaq.) They can do this because they don't need full events to compress or zero-suppress.
At the moment, the natural place to run compression in artdaq is in art modules in the EventBuilder processes. However, there may be a need in future experiments to allow for functionality like that in the BoardReader processes so that we are not sending as much data over the network. We should investigate what would be needed to do this.
The initial time estimate that I'm putting in this Issue is simply for some initial discussions.