Description
I feel like we are close to having a decent MVP version with a (close to) zero-config parse-anything mode with predictable behaviours, and a strict mode that detects many common user errors so the author does not have to.
We are at risk of stalling though. @darcyclarke who did the initial proposal in this repo is currently doing his own thing in https://github.com/darcyclarke/minargs (taking into account some discussions here), and I think @bcoe is taking a break from trying to coax some consensus (he did mention threads with "4 people with 8 opinions in them" as annoying, but good!). I suspect also the wider tooling group who are behind this initiative have been busy elsewhere.
I have been leaving it up to common agreement, or Benjamin to move things forward. We have covered a lot of good ground in discussions, but there are multiple viewpoints and goals and desires. I am going to try being more active to push towards a first milestone with key functionality implemented while we have people engaged here. Potentially with some approach rather than a universally agreed approach. It doesn't mean it is ready for Node.js consideration, or everything it could be, but I would like to try for something reasonably solid and implemented and usable.
I'll put 🔨 in any comments where I am pushing for a way forward. Speak up if I am out of order on a decision or process, and I'll cease and desist. 🙊
(Disclaimer: I am a maintainer of Commander. When issues come up I often research what other libraries are doing and what the POSIX and GNU guidelines say. I believe I have solid knowledge and experience of how command line parsing works in practice. I don't have experience with building a parser on top of a parser as such, which although not one of the stated goals of this repo is definitely of interest to people!)