> I tried to merge this through to master and the test in this PR fails.
It only failed when you had it merged into `master` locally? Any idea why it doesn't already fail on `3.4-dev` / in this PR? > I guess that .NET doesn't end up using the same channel for sessions the way > that the Java driver does. The Java driver only allows a session to be used by a single channel and therefore a single connection? That increases the complexity for this feature for Gremlin.NET as it now needs to assign one connection for each session. If this gets too complex, then we will need to make the decision of whether we actually want to support this feature in .NET, but maybe it's also actually easier to implement and maintain then I'm thinking right now. [ Full content available at: https://github.com/apache/tinkerpop/pull/1263 ] This message was relayed via gitbox.apache.org for [email protected]
