I just published an article on React Server Components. While they promised much, real-world use revealed some gaps.
The Limits of RSC: A Practitioner’s Journey
www.nirtamir.com/articles/the...
Just shipped a big upgrade to eslint-plugin-default-import-name! 🚀
✨ Auto-renames usages of default imports (incl. JSX + vars)
⚔️ Handles naming conflicts
🧠 Customizable with importPathRegexToTemplate
Try it — your code will thank you 💅
github.com/nirtamir2/es...
In this manner - revalidating is like creating another API request. We don't have interactivity because it's like an API request and response. This also makes me think that in some cases it's better to compute things on the client synchronously instead of sending an API request
I'm thinking of RSC as a server API that gets props returns JSX, and tells React to render it in a specific node. Good use case for it is rendering Markdown - when the added JS to compute and render the markdown costs more than doing it in the server and just returning the result
I'm happy to use Astro for my website. I wish Astro would have Quik's like ability to take <script> tag in Astro components, bundle it separately, add its id to the markup (like QRL), and load it lazily - only when the user hovers over the component (like URL but for bundled JS)
Started a new project with @tanstack.com start!
I'm super excited about it - I plan to migrate from Next.js for one of my projects if it will work well. RSC did not work well for me - even the most minor infinite loading breaks its model and it's very limited in client state updates
The new area of MPC makes me think that IDEs like Cursor can do more than writing code. I may be able to control Blender and other tools, do research, and write more effective notes. It may be the gateway to interact with AI models.
Sometimes, I explore creative coding using Claude / ChatGPT to write some code just to see the result. I think doing it in Cursor will do a better job because it is aware of the project context, and it saves me time copying and pasting, and cursor rules can give me better results
I'm using Webstorm as IDE mainly for the refactoring abilities and shortcuts. AI autocompletions do not work for me effectively yet. I want to be accurate and know what to do most of the time. AI may help me do more specific refactoring that is not available now.
I think the best way to learn something new is by building a side project that integrates it 🚀
Interesting - you are talking about v8 hidden classes 👀
IDK about the gain in performance it can have, but I'm interested to see the results in benchmarks
The rationale was to fix the happy paths (like explicit types) and give options.
I also considered an option to allow it only on variables under specific comment / variable names type regexes.
I need to run it on some of my projects to get more insights.
I created a new ESLint rule: `sort-object-properties-by-type` 🎉. It reorders object properties based on their TypeScript type, even inside arrays. Useful for keeping configs and large objects organized. Includes auto-fix 🔧
github.com/nirtamir2/es...
@typescript-eslint.io
@eslint.org
I find TDD (Test Driven Development) works well for developing ESLint plugins. It helps me stay focused, catch bugs as I build, and automate checks, making the process more efficient and smoother
I think it's time to use effect.website for one of my projects, which includes a lot of try-catch blocks. Curious to see how it fits
Good read on productivity systems and how someone executes effectively. I appreciate the idea of accountability partners for staying on track.
ricos.site/posts/my-pro...
Integrated pkg.pr.new into my ESLint package. It lets me test npm packages in a live environment before publishing—like Vercel for npm. No need to publish multiple versions just to test, and it makes releases much smoother and more reliable
I contributed to
github.com/urish/vscode... in the past,
But now with the power of AI we can probably convert native language to AST selectors. But not sure about how to integrate modules inside.
I’m in Miami right now! Any devs up for a coffee? Also, what tech meetups or events should I check out while I’m here?
Just create something useful that you care about
Start by taking the first step. Begin with a modest goal, such as earning your first $10. There's no one-size-fits-all approach; simply take action and get started.
Remember what brought you joy and success in your business. It's perfectly fine to revisit and refine it. Don't feel pressured by an imaginary audience; focus on making your vision a reality, as it's your own creation.
Delegate or die. If you leave for a year - you want your business to improve without you.
Do it for love, not just for financial gain. Set up your business as if you didn't need the money – a mindset that prioritizes being over having. This approach will bring me happiness because it's about becoming who you want to be, rather than accumulating what you want to have.
Prioritize the needs of your customers above those of your business. Even if you have competitors, this mindset is essential. Avoid creating situations that might cause your customers difficulties, simply to ensure your company's survival.
Some ideas to get excited about: "Hell yeah!" or not? Instead, do something useful that truly matters to you. Do it because you're passionate about it.
During my flight to Barcelona, I listened to the podcast “Anything You Want - 40 Lessons for a New Kind of Entrepreneur.” I highly recommend it; it provided me with a lot of inspiration. I’d love to follow in his footsteps.
An excellent video explaining and integrating new CSS features into a web app
www.youtube.com/watch?v=oXSF...
After navigating through lengthy bureaucratic procedures on Google Play, my memory game app has been successfully re-published in the Play Store!
play.google.com/store/apps/d...
