On 2020-02-27 14:51, Bill Cole wrote:
On 27 Feb 2020, at 14:24, Jim Popovitch via Mailman-Users wrote:
Personally, I'd like to see the GNU Mailman project have a formal Mailman 2.3 release that supports Python3, I feel that there would be a lot of support for that.
I'm sure there would be widespread applause and congratulations if such a thing were actually released. That sort of "support" is unhelpful towards actually making such a release.
The needed support is the actual skilled effort of writing the required Python3 code. I don't have the time to hunt down the specific statements, but I have vague recollections that both Barry and Mark have said repeatedly that doing so would be substantially more effort than they are willing to put into anything built on the MM2 architecture.
Rewriting without breaking is hard.
There is a Python framework called Twisted. It has a lot of useful features. Also a lot of vices, but a lot of useful features. As best I can determine, the task of updating it to be Python 3 compatible has now been under way for ten years (with most of that time, only one person working on it).
What has this yielded?
"Most of the most commonly used parts" of Twisted are now Python 3 compatible.