updated reference https://issues.apache.org/jira/browse/TINKERPOP-1392
On Fri, Jul 29, 2016 at 10:40 AM, Dylan Millikin <[email protected]> wrote: > +1 sounds good > > On Fri, Jul 29, 2016 at 5:25 AM, Stephen Mallette <[email protected]> > wrote: > > > I just created this issue: > > > > https://issues.apache.org/jira/browse/TINKERPOP-1391 > > > > As part of TINKERPOP-1278, I'd went ahead with deprecating the old method > > of sending a Traversal to the server. The approach taken there was to use > > java serialization to ship the Traversal to the server. The more I work > on > > this though, the more it just looks like dead code that we really don't > > need to keep for any reason at all. > > > > Deprecation isn't really helping anyone out in this case as there are no > > drivers (save TinkerPop's gremlin-driver) that support this capability > > which is now wholly replaced by Bytecode serialization. Users won't know > > the difference as it is at a lower level of their concern and third-party > > drivers will be unaffected because they couldn't really support this > > "undocumented feature" anyway outside of the JVM. > > > > So, with all that reasoning, I'd just like to remove it all together and > > make some code look cleaner in the driver and server. The java > > serialization approach was a good experiment that helped lead us to the > > Bytecode solution and I think that's all it was ever really good for. > > > > Unless there are objections, I'll assume lazy consensus in 72 hours > > (Monday, August 1, 2016, 8:30am EST) and move forward with dropping it > out. > > > > Thanks, > > > > Stephen > > >
