Daniel K.

TypeScript friend; ReactJS lover; NodeJS follower; wannabe gamedev; not so scary monster anymore

[Daniel K. @danielk_cz]

HBO: Remake Game of Thrones Season 8 with competent writers. - Sign the Petition! https://t.co/kNEONeUOH8 via @Change

[Sizzy - The Browser For Developers @sizzyapp]

Sizzy 0.32 is LIVE 🥳 We also have a new changelog 😍 🎉

Full Changelog: https://t.co/3eoz3eWwxL

[Daniel K. @danielk_cz]

@mpocock1 Looks great Matt, keep it pushing :) I also wonder if there is some good way for handling nested states vs `state.matches()` call with TypeScript being able to check that. It could, in theory, generate all possible paths.

[Daniel K. @danielk_cz]

@mjackson @cassiozen It's cool, you just need to remember what was your previous branch and if it's safe to run that command :)

[Daniel K. @danielk_cz]

@mpocock1 @DavidKPiano What about Babel macros, would that work? That would be certainly preferable because of CRA.

[Daniel K. @danielk_cz]

@mpocock1 @DavidKPiano I don't follow why would need Babel plugins? Isn't it just a codegen that runs separately from any compilation?

[Daniel K. @danielk_cz]

@mpocock1 @DavidKPiano Yes, I do React and I think it's the most popular option anyway. Others can come later.

[Daniel K. @danielk_cz]

@mattiasgeniar @slyduda @ctrlshifti Prettier would not like that for sure...

[Daniel K. @danielk_cz]

@mpocock1 @DavidKPiano I have been fighting with this a lot. Please, make it public soon🤯

[Daniel K. @danielk_cz]

Must have!

[Matt Pocock @mattpocockuk]

I built it, and it's awesome. Full type safety on even exceptionally complex machines 😍

@DavidKPiano

[Daniel K. @danielk_cz]

@kentcdodds Sounds good, but I don't have that many followers for anyone actually notice such post :)

[Daniel K. @danielk_cz]

@kentcdodds idea for blog post how the useMemo doesn't provide semantic guarantee. I often keep running to snippets where people assume otherwise. That little notice in React docs is fairly invisible ... https://t.co/AcZO87tHEz

[Daniel K. @danielk_cz]

@kentcdodds Well, it does matter if store some state related value in there. I see it a lot with MobX observables actually. And I know it's not a real thing, but it's going to be hell to debug such random situation if we don't prepare for it now.

[Michel @mweststrate]

Can't stress this enough. Not all reconciliation is good. As a rule of thumb, you probably don't want to reuse a react tree for anything that has a different identity, even when it is structurally similar! As you might spill state over from one identity to the next

[David K 🎹 @DavidKPiano]

Why XState is different (according to them):

- Completely framework-agnostic
- Based on a W3C standard: https://t.co/ijq5TGTNNF
- Serializable logic
- Visualization
- Time-travel in the *future* (know which next states are possible)
- Absolutely no new nor invented ideas

[Michel @mweststrate]

To summarize: Recoil & MobX are similar: many, individual subscribable atoms (Redux has always one), enabling sideways comp updates + model that favors deriving & memoizing data. MobX offers higher level abstractions like maps and arrays based on atoms + transparent tracking

[Daniel K. @danielk_cz]

@portexe @dan_abramov If only "read" was enough. That person would need to evaluate importance, translate it to Dan and he would need to provide actual answer. Someone should make a SaaS like that :D

[Daniel K. @danielk_cz]

@brian_d_vaughn @dan_abramov Each one of those people who writes that should try to be OSS maintainer once :) Like a basic education.

[Daniel K. @danielk_cz]

@thekitze I suppose you did not figure out audio issues and cancelled it?

[Daniel K. @danielk_cz]

@kore_sar If only it was a large project :) There is like 20 dependencies right now.

[Daniel K. @danielk_cz]

I want to remove peerDependencies from existence! This is from my small project, everything up to date and it's a total nightmare trying to track everything down.