Jempbox is slow on xmp with large event histories

In looking at the timeouts in a recent run against 8 million PDFs, I found one file where the processing time was caused by extremely slow parsing of the media management schema.

If I do enough subclassing and put a hard limit inside getEventSequenceList(), the processing time is fairly quick.

I realize that Jempbox is not going to be supported going forward and understand if this is a "do not fix".