Thanks for the support they86!
I think you and others are right, that we should focus more on the strengths and maybe cut the features that take a lot of work and maintenance (integrating with constantly changing external services).
Legend has a lot of legacy code in it since it was started 10 years ago. We had built a lot of things from scratch because there were no open source libraries or common best practices at the time: the editor, the virtual list, email parsing, sync integrations with google services, custom designed the whole UI, etc… So for example in comparison with newer competitors who use a big and well maintained open-source editor (TipTap) and get tons of features from that for free, I have to add editor features manually.
I’ve been working on open source projects that were inspired by the core tech in Legend (Legend State and Legend List), which can power making a very fast UI with a robust sync system. And lots of other people are using them so there’s a lot more test coverage than just Legend. I already pulled some of that work into Legend last year for the sync improvements, but there’s still a lot of old code that could be upgraded.
So I’ve been noodling on the idea of starting from scratch, rebuilding the core most important features (outliner, panes, boards), to be able to add some of the other ideas I’ve been thinking about much faster. And by using AI to assist in developing it, I think I could build a lot of that really fast. And there’s some core changes I’d like to make that would be much easier when designing around them from the start.
I’ve also been wondering if it could make sense to start a new project as open source so more people could contribute to it. There could be a paid sync layer or premium non-open source features on top to help fund it?
But anyway, I just want to reassure you that I am still thinking about this and want to come back to it soon 😀
