Hi Adam,

from my point of view it'd be great to have more people involved. I've been doing some development, but more or less out of necessity (I enjoy working on the project when I can, but don't really have the time to commit to it much more than I have been unfortunately).

To give you some background around why there's the renshawbay repo as well as the main repo; Initially I created the renshawbay repo as there were various changes I needed to make to the project while it was hosted in sourceforge, and for me it was easier to manage that in github. I added Python 3 support, and later some stuff around subclassing .net classes in Python. It was then decided to move the project to github, so we created the pythonnet repo directly from the sourceforge repo rather than fork the renshawbay repo. The thinking back then was that we should do a 2.0 release based off that fork, and then once that was stable look at merging in the Python 3 work from the renshawbay fork.

So, to answer your first question - pythonnet/pythonnet is the official repo, but most new development (new features etc) has taken place on the renshawbay/pythonnet fork in the python3 branch (which maintains support for Python 2).

There's no official roadmap that I'm aware of. There are some issues and milestones in the github repo, but AFAIK no one is actively working on those right now. There are only a couple of issues remaining for the 2.0 release however.

Pull requests to the official repo are reviewed and merged by the project owners (of which I'm one). So far those have been bug fixes or changes necessary for the 2.0 release. I keep the renshawbay fork up to date with any commits to the main repo.

For what it's worth, here's what I'm aware of that needs attention:
    - Finalizing and releasing to PyPI the 2.0 release for Python 2.x only
    - Testing and getting the renshawbay python3 branch working on non-windows platforms (the linux build currently has problems, and I've not tested any mac builds)
    - Merging the renshawbay python3 branch into the main pythonnet fork
    - Updating the docs

If you're able to help at all that would be much appreciated.

Best regards,
Tony


On Tue, Jun 16, 2015 at 12:08 AM Tribble, Brett <btribble@ea.com> wrote:

Adam, I’m ecstatic that there’s a player out there who is making good use of Python.net, and who would like to help contribute. Organizational and logistical issues aside, I’m all for anything you can throw at the project! Someone correct me if I’m wrong, but I believe Brian Lloyd has largely yielded this project to Tony and the community. Based on Brian and  Tony’s past posts, I’m fairly sure they’ll welcome any and all contributions to the project.

 

It may sound sacrilegious to some, but I would love to see the PTVS (https://pytools.codeplex.com/) folks get involved with the project. They’re turning out a solid product, and this fits solidly in with what Microsoft is trying to do with PTVS, .NET Core, Azure etc.

 

 

From: PythonDotNet [mailto:pythondotnet-bounces+btribble=ea.com@python.org] On Behalf Of Brad Friedman
Sent: Monday, June 15, 2015 2:36 PM
To: A list for users and developers of Python for .NET
Subject: Re: [Python.NET] surveying the landscape ...

 

I'll chime in and say the lack of these kinds of legitimate "stake-holder" systems and responsibilities has forced me to turn away from depending my work on this project. I still keep up on it in hopes that it will turn around.  If a legitimate player were to step up and contribute to a responsible, active and stable future for the project, I'd likely reconsider my stance and begin active support again. It's hard to justify putting much into it as one guy with limited resources.  It needs full multi-platform release and development support both as a python module and a .net embedding toolkit, both for Python 2.x and 3.x.  That's a lot of work to commit to getting set right and maintaining.


On Jun 15, 2015, at 3:57 PM, Adam Klein <aklein@bluemountaincapital.com> wrote:

Hello all,

 

We are using Python.NET at BlueMountain to interface between our large .NET code base and the cpython ecosystem for interactive, exploratory computing. By way of background, I was a major contributor to the pandas library for a time; my firm is behind the Deedle library (https://github.com/BlueMountainCapital/Deedle).

To state the obvious, the project has proven hugely valuable. BlueMountain has an interest in making sure the library doesn’t languish. To that end, we’re interested in contributing to the project in terms of manpower and possibly funding development. I’d like to get a better sense of a few things:

 

- is there a BDFL … is Brian Lloyd still active; or is Tony Roberts steering the ship (being the top code contributor recently on github?) It looks like python 3.x integration and more recent work is happening on on renshawbay/pythonnet? Is pythonnet/pythonnet still the official repo?

- who manages releases to PyPI?

- is this PythonDotNet mailing list the appropriate clearinghouse for all discussions related to the project?

- are there other institutions that are public users of this project?

- is there an official vision or roadmap for future releases?

 

I see that python 3.x support looks like it’s happening on renshawbay/pythonnet (awesome). For other wish-list items or proposed contributions, is it best to start opening issues on the pythonnet/pythonnet github site? How are pull requests managed?

 

I’m also wondering if there’s any collective / documented knowledge about the inherent limitations and pitfalls of the library and/or where development resources are needed?

 

Best,

  Adam

 

 

 

_________________________________________________
Python.NET mailing list - PythonDotNet@python.org
https://mail.python.org/mailman/listinfo/pythondotnet

_________________________________________________
Python.NET mailing list - PythonDotNet@python.org
https://mail.python.org/mailman/listinfo/pythondotnet