Finalize design and UI strategy

[jsonsivar]

Background terms for ppl like me:

• Headless/Radix: logic for UI components, no css just html and JS
• Stitches: styling framework, styled components
• Tailwind CSS: predefined styles in CSS format, they enable more granular css styles that make it composable instead of making broad classes directly (iiuc)

Questions we need to answer?

1. How does design <> frontend development process look like?
2. What style framework to use?
3. What theme to use?
4. Other important points
   • abstract components into interfaces so they can be changed later easily

**** MOST IMPORTANT PART **** Go forward model:

• design hand-off with dev will be more loose than in the past, it wont be exact or perfect but we use the design as a basis and make sure it looks good, then we polish later but that shouldnt block releases
• in the stack, we are using stitches and tailwind css, and then based on requirements pick from headless or radix
  • if it turns out radix makes sense for that particular use case/component, then maybe just use that piece from radix to save time (eg color palette)
  • we dont have to take an all or nothing approach
  • if we use the abstractions Petr mentions we can stay agile with how we implement it
  • also found this plugin that connects tailwindcss and radix so that gives us even more flexibility
• trial run
  • lets do a trial run with the mint epic for NFA with the following approach
  • it will give us data to adjust based on and more clearly predict how long the FE will take
  • we have a design base we will use from FIGMA
  • the look and feel are whats in drip so we can borrow the css and implement it in tailwind so we get best of both worlds
  • based on this experience we can improve and adjust whether thats to use more tailwind or less
• drip: we dont have to use it all or at all. we can just take the css and put it into tailwind so we follow the theme for the company, but we implement it the way its best for the project via stitches and tailwind css
• in the end NFA team will go with an approach that will make sure the team can do as fast as possible, while also hitting the quality we want for an MVP

And I posted this last week and it's even more relevant now if anything it's going to be even more looser: