It occurred to me (after noting that doc generation failed as a result of
my changes) that removing Python 2 support means that we can't have Jython
support as it doesn't support Python 3. Even if it did, I'd wonder if we
needed to keep maintaining Jython support in 3.5.0. It seems like a good
time to leave that behind as our focus for gremlin-python is about running
native Python and Jython for server support of Python based scripts hasn't
really evolved (nor do we want to expand script processing). The fact is
that the Jython ScriptEngine isn't nearly as advanced as the Groovy one and
therefore not especially performant or feature rich. I've not encountered
anyone using our Jython components since we first released them and most
Python developers who use our tools wonder why we even have Jython support
at all.

So, with all that said, I expect to expand this body of work to include a
full restructuring of gremlin-python to include removal (and deprecation in
earlier branches) of any Java/Jython related functionality for 3.5.0. Note
that we will lose automatic script processing for Python based docs by
doing this, but we don't have a lot of python generated docs so I tend to
think that's ok. Please let me know if there are any concerns with this
direction.



On Mon, Nov 25, 2019 at 9:02 AM Stephen Mallette <[email protected]>
wrote:

> I've alluded to it many times now without any objection but to make things
> more official I've created this issue in JIRA for the removal of Python 2
> support in 3.5.0:
>
> https://issues.apache.org/jira/browse/TINKERPOP-2317
>
> This will be a nice change as it means less test runs, less code to
> maintain, support for the latest python libraries, and we can be more
> consistent in how we write our code without having to lean back on Python 2
> odds/ends.
>

Reply via email to