Improvements to performance reliability and maintainability historically almost always get released, provided there are not downsides, but you already covered maintainability, so perhaps there aren’t.
An amusing exchange on maintainability, readability, and correctness of rewrite is in a pull request.
It would ideally be maintainable by volunteers who aren’t always top database experts unless we’re fortunate enough to always have them volunteering. Another question would be how standalone the database work is. If it gets into big revisions in the C# design or the destination files, risks increase.
I’ll describe the historical model, although this has hit snags recently due to a limit from a volunteer.
There has been a code signing certificate issue, but really what we’d like is a new release manager,
exact duties TBD, but it probably includes doing release notes, releases, and perhaps moving code.
Generally a Canary release is the first after automated testing done at pull request submission, then willing volunteers on that channel run the release. After enough confidence, it would move to a Beta.
It’s certainly a bad thing if breakages get this far, exposing the entire user base, so care is needed…
I’m trying to get you and @tarianjed working together, and I now notice that I forgot about @vmsh0
Database rebuild did an analysis, followed by my comments and my concise DB writeup at end here.
Not exactly (that I know of). There are some less formal pieces around, such as the above and the Wiki:
Documentation for the local database format
Anything that can be done to smooth the entry path for volunteers would help. It’s difficult to get started. Maintainability of documentation matters too. I think the above Wiki doc is slightly out of date already… Anything that needs special software to create might be more difficult for even willing helpers to update.
What’s not to like about that (if it can be done safely)? If by support, you mean help with, resources are unfortunately limited – and some volunteers would deeply benefit things, but I mentioned possible help.
The original Duplicati author was not a database professional, participates less these days, and would probably be thrilled (I know I would be) if a team of DB volunteers could make headway on such things.