suggestion: reboot with rescript core?
Opened this issue · 3 comments
Seems that the Rescript team is going to remove Belt from the Rescript compiler and language(see), I suggest moving forward to rescript-core bindings, which is the new standard library in Rescript ecosystem.
I guess some of the bindings and APIs currently we've got won't match to rescript core's, but it's a matter of time the belt is going to be removed from the rescript, we have to move ASAP.
And, I think it will bring good momentum to reboot this library.
@JUSTIVE, thanks for letting me know! It's indeed a great moment to start over, I have already researched a few things, for instance: moving to Reason + Melange, and it's something I do really consider (but I haven't thoroughly tested the performance yet)
PS. thanks for your email! ❤️
Reason + Melange?! That's a lot more bolder move than I thought. I thought this library's root is rescript(more precisely, Bucklescript) and all the core concepts of this library come from the rescript's GenType. Wouldn't it be too large to continue on this project(ts-belt)? Maybe it should be a whole separate project like 'ts-caml', or 'ts-reason'(just found this prior work, ts2reason).
My idea was to extend Belt's API with rescript core's, not to change the whole build system or architectures.
btw, besides of rebooting the whole library, we need to improve current state. there are plenty of bugs and glitches to resolve. could you review, merge PRs and release new rc version?