I maintain a (formerly) popular Python third-party library. At one time it was in the top few hundred most-downloaded packages on PyPI. Unfortunately (thanks to work and life) I haven't actively worked it for several years now, even though it's still a dependency for a few very well-known projects.<p>Every time I try to get back to it and get started again, I remember all the roadblocks. These are probably not <i>that</i> hard to overcome, but psychologically they keep making it easier to do nothing rather than dig in:<p>- All the latest work is in an "next generation" branch that's been converted to Python3. I'm not sure whether I should release this as a separate "NG" package, or do a bunch of work to make it 2.x compatible again, or if I can use something like 3to2.<p>- There's a lot of changes on that "next generation" branch that I need to document.<p>- I haven't read any of my project email or mailing list traffic in years.<p>- The project is hosted in mercurial on Bitbucket, I want to migrate it to git on Github. I have an old attempt to transfer it on Github already, I need to figure out how to update that to the latest mercurial revision.<p>- I should release a new version of the "old" (released) codebase with bugfixes, but I'm not as familiar with it anymore, compared to the "next gen" code.<p>Has anyone else gone through this? Any advice on where to start? Should I try to find a partner to help motivate me? Even just words of encouragement might help at this point ;)
I feel you man. I would definitely say, this whole disastrous decades-long schism between python 2 and python 3 certainly does not help the situation. I would start by answering the following questions to myself:<p>1. What are the major issues users are most frequently facing with my packages and what minimal-effort things can I do to address these issues<p>2. As you said, should I make a "next-gen" version, and if so, should I deprecate the old version (I think this might be a good approach)<p>3. are there developers who, if given a nudge/opportunity, could help in this clean up?<p>Just some thoughts.
To get the juices flowing don't focus on "all the roadblocks" just focus on putting in Time - 20mins to 1 hour a day - same timeslot everyday (early mornings work for me). Doesn't matter on which issue. Its about building momentum up slowly. The longer you sustain it, the more likely you will get sucked back in. I usually need atleast a month of that, to get back into older projects.<p>And try to work on that documentation. Good docs brings in those collaborators. Even if they have no experience, any help is better than none and those interactions will be an extra source of motivation.