I like your vision for Hypen and I sincerely hope that you can avoid the faith of every project that tried it before you! (all the toolkits that started as declarative, with clear logic/ui separation, eventually and inevitably add scripting and reinvent js/html).
That made me chuckle though:
Accessibility and i18n tooling:
This is in the plan - accessibility should be the easier one, with i18n tooling coming as more "platform calls" support comes online.
In my experience accessibility is far more difficult to get right, but maybe I've just been unlucky. :)
Thanks!
Honestly, in one iteration of Hypen I actually added a `Script` component, but after trying it, it already felt both dirty and confusing, so it's a definitive no for the near future :)
Regarding accessibility - I've also burned myself quite a few times on that bridge and that is why there won't be 1.0 without it being fully solved - I don't want devs to have to retroactively think about it or for it to have a second class implementation.
I believe there is a solution out of the box that should cover at least 80% of the most common needs, and provide primitives for more custom cases. Hopefully this time I can nail it... fingers crossed!
Accessibility is also difficult to add retroactively. You should consider it from the start, not simply add it to a plan!