Mastodon

happy try-not-to-lose-your-shit-as-you-endure-terrible-instruction-manuals-and-extremely-poor-toy-design day, for those that celebrate

it’s going wonderful over here 😅😅😅

(merry christmas, y’all 🎄)

If folks could refrain from comparing no-build-step Web Components (oranges) to *build-step* Component Frameworks (apples), I would be very happy.

Feel free to compare no-build-step Web Components (oranges) to no-build-step Component Frameworks (oranges) though.

After over a year of using Mastodon almost exclusively for social media—it’s *still* a shock to go back to visit an algorithmically-driven social network.

Scrolling and reading on other sites feels like eating a candy bar while drinking a soda—the candy bar has very little flavor—your taste buds are already overloaded by sugar.

And yet sugar is the algorithmically incentivized form of content, so sugary posts are the only ones that get traction.

Importantly: Eleventy is *still* punching far above its weight despite the Netlify-induced setbacks we faced this year.

Many of the frameworks listed are VC-funded with multiple full-time developers. We don’t have that—intentionally. Our project goals are different.

As I’ve said before: Eleventy and Astro are allies in the web development framework melee. When Astro wins, Eleventy wins too. It isn’t a zero-sum game.

Alright, this graph from Netlify’s “State of the Web” survey is… intellectually dishonest. And I have to weigh in on it now because a few folks have been making some *specious* assumptions and tagging me 😬

I get it: Netlify wants to highlight graphs that make Next.js look bad but c’mon—leave me out of it.

(I also understand that their resident statistics person departed the company so… it’s unlikely that this was malicious)

🧵🔽

Rendering framework or site generator usage and satisfaction changes (2023 vs 2022) In the bottom left corner (negative changes in satisfaction scores and usage) is 11ty, Gridsome, Preact, Gatsby (obscured), and Next

If a “survey” displays results from questions about developer perception of frameworks without filtering out folks that have never used the framework, it ain’t much use—is it?

Are we interested in Amazon reviews from people that have never bought the product? Or movie reviews from people that haven’t watched the movie?

It’s frustrating out here, y’all.

I see this experimental limbo for Node.js’ ESM `vm.Module` causing large scale problems elsewhere too, right?

Jest seems to be trying to stick to Node.js APIs (which is a good goal!) and people are migrating to Vitest for “stable” ESM support—but Vitest is using `esbuild` to sidestep the experimental status.

Here’s how I think about it and I think this is the only way to survive as an indie open source project in a VC-funded world:

The #1 competitor to Eleventy will always be the previous version of Eleventy.

Fixed what I think may have been the gnarliest @eleventy bug yet: a mixture of multiple subtle things when using pagination, computed data, and (internally) data mutability in proxies on the data cascade.

I learned a few things and left everything cleaner than it used to be but I’m going to rest now 🧠😴

Vercel’s Next.js v14 has shipped and (as one does) I had a look at the JavaScript footprint of the production build of Hello World:

v14 (2023): 299 kB
v13 (2022): 272 kB
v12 (2021): 228 kB

Unfortunately the same trend continues, even with a new bundler: the bundles are growing!

(and with Tailwind included by default, too)

A screenshot of Create Next App with the Chrome Devtools showing 299 kB of JavaScript loaded on the page from 6 requests (91.8 kB transferred)