DBozhinovski/supertokens-wc-exploration

Dragons be here. Very much WIP.
For me: // framework comparison

// Preact
// Solid

• Uniform components (JSX ideally)
• Async calls during init (fetch during mount)
• Async event handlers

• SSR support
• Bundle size / embedding

Summary:

1. We are sort-of biased towards Solid
2. SSR is "hackable"
   1. Long-term, it's definitely "standards" solvable, so we're not locking ourselves out of anything
3. Bundle size is in favor of Solid
4. Async stuff feels nicer in Solid (think react-query)
5. Moving fully to Solid is a non-trivial amount of work, but doable

Full vs. partial

• Partial:
  • Users suffer only an increased bundle size (until the transition is complete)
  • It's harder to maintain (basically, two codebases)
  • A potential path to this would be compile to web components from Solid, replace React components in the SDK.
• Full:
  • A lot of upfront work
  • A single transition (to a single codebase)
  • Feature halt or duplicated work until the transition is complete