Ah I see.
I wonder if prefixes could work here.
When you have lib-p-5, consumers could add second tailwind compile step, that handles these classes with the prefix to generate the library css?
@xent.dev.bsky.social
Web, Cloud & Architecture
Ah I see.
I wonder if prefixes could work here.
When you have lib-p-5, consumers could add second tailwind compile step, that handles these classes with the prefix to generate the library css?
What do you mean with βtranspiling awayβ the classes?
Non tailwind users would still include your Stylesheet with the tailwind classes that were used in your lib?
Consumers can just add the libraryβs dist folder to their tailwind config and it will generate the classes in its own css file. Of course, this could break styles when size classes are different.
13.01.2025 09:28 β π 0 π 0 π¬ 1 π 0Is upgrading to websockets an option?
TCP should already guarantee order of packets.
Iβm sure I read more than 10 posts from you explaining path traversal attacks but I canβt remember a single one that shows the solution to preventing them π
09.01.2025 16:01 β π 0 π 0 π¬ 1 π 0No, sadly you are correct.
The docs on static export explicitly mention that dynamic params without genereagteStaticParams is not supported.
nextjs.org/docs/app/bui...
Is the exploration part coming? :D
09.01.2025 06:12 β π 1 π 0 π¬ 2 π 0Graph of the State of JS 2024 results showing frontend frameworks on a chart with two dimensions: used /not used, and positive vs negative sentiment. Over the years, all of them seem to drift towards the center.
Frontend world seems to be converging.
General trend seems to be moving towards the center of the chart. And this aligns with the framework directions as well. Many are adopting similar reactive primitives and similar control flow approaches. So people seem to care less about the specific choices.
Hm, and of course I canβt find the twitter thread anymoreβ¦
16.12.2024 10:25 β π 1 π 0 π¬ 1 π 0Didn't we discuss that JSX isn't a valid return type for Server functions? @tkdodo.eu , @nilshartmann.net , @adamrackis.bsky.social
Looks like the docs changed: react.dev/reference/rs...
Itβs about arguments but also links to Server Component return types as valid for server actions.
Inngest is such a cool concept. Directly integrating into your app, where all dependencies are already set up is brilliant.
11.12.2024 06:23 β π 1 π 0 π¬ 0 π 0But curly braces are so hard to type on German MacOS keyboards π€ͺ
06.12.2024 07:38 β π 1 π 0 π¬ 0 π 0While working on this, the LLM use case became a quite obvious candidate for streaming so I wrote up this article:
www.esveo.com/en/blog/stre...
Bye bye forwardRef!
Hello cleanup functions for refs!
Pretty amazing how such small changes can improve your life so much!
Same :D
05.12.2024 17:18 β π 1 π 0 π¬ 0 π 0React is fast because of the virtual DOM?
02.12.2024 17:47 β π 0 π 0 π¬ 0 π 0Are you already at the place where you bore yourself because you realise that you are repeating the same jokes and stories?
Happened to me a couple of years ago when I gave a training for the 4th timeβ¦
$scope.apply()
29.11.2024 14:52 β π 0 π 0 π¬ 1 π 0Recently, we had so many discussions about react hooks, signals and more. How fitting that gitnation now published my lighting talk where I discuss why React should NOT adopt signals.
gitnation.com/contents/why...
Wow, this content looks really cool!
I almost skipped this because the only thing above the fold is the table of contents and the enroll button. A simple or continue free lessons with a down arrow could fix this risk.
What really surprised me was that there are some starting positions,that need EIGHT clicks to get to the end position...
Sadly, this kind of killed the game for me, because it feels like it's really unlikely that I will memorize these harder states, so I will always have to fall back to guessing...
github.com/andrewgreenh...
Whole source code can be found here.
So after a long day of work in front of my laptop, I decided to not spent time with my wife and open up my macbook again to build a solver...
And now, 2 hours later, it's done:
andrewgreenh.github.io/moriarty-con...
Each grid displays a number (the steps to finish) on the next cell to click.
It became apparent quickly, that there are "easy" starting positions and "hard" ones. For the harder ones, I always ended up randomly clicking around until I found a familiar place, where I knew how to finish.
And then I wondered what the optimal way was to solve each starting position
In this minigame, you have a 3by3 grid of cells. Each cell can be on, or off. By clicking on a cell, you toggle the cell and the direct neighbours (NOT DIAGONALLY).
Your task is turn the whole grid "on" as quickly as possible on as many random start positions as possible.
I got nerd sniped...
Recently someone (I believe it was @erikras.com ) wrote something about murdle.com. Besides the excellent logic puzzle, there is this mini game called moriarty console: murdle.com/console/
No no, thatβs taxidermy. You are looking for taximeter.
26.11.2024 07:04 β π 1 π 0 π¬ 0 π 0I absolutely love what you all are currently doing with node.js
In the last, I did not really care about node releases but now, every release contains something awesome!
public static void main() {}
Me too!
ππππ¦΄π¦΄
ππππͺ«πͺ«