I see that since moving to the new Jira, this issue has been removed and the source code now indicates it's not supported at all anymore. I would like to restate the case for allowing Bobo IndexReaders to be efficiently reopened, including deletes ánd additions.
We use Hibernate Search and we have found it challenging to integrate that with Bobo. So, we're left with creating new Bobo IndexReaders periodically and caching those for across thread results. This leads to a lot of memory overhead and garbage collecting as the IndexReaders tend to builld up a lot of cache.
We would be grateful if the reopen method was properly implemented.