I guess the javadoc isn't as clear as it should be. It would be better as:
> An internal function that signals a read or a write is about to occur.... On Thu, Mar 10, 2016 at 11:52 PM, Jonathan Ellithorpe <[email protected]> wrote: > Hi all, > > Quick question, is Transaction.readWrite() supposed to be used by > implementers to signal that a read/write is *about* to happen, or that a > read/write *has* happened. The comments indicate the latter: > > /** > * An internal function that signals a read or a write has occurred - not > meant to be called directly by end users. > */ > public void readWrite(); > > But the Neo4j Gremlin implementation is using it in the way of the former > (that is, using it to signal that a read/write is *about* to happen). > > Thanks for any clarification on this, want to make sure I'm implementing > transactions in the way that's intended. > > Jonathan >
