Shellsharks Blogroll - BlogFlock

Create A Static Site Using 11ty & Deploy to Neocities (2026 Refresh) - The Weblog of fLaMEd

2026-06-16T12:00:00.000Z

What’s going on, Internet? Way back in 2022 I wrote a guide on building a static site with 11ty and deploying it to Neocities. It’s been one of my most-read posts, but it’s also aged: Eleventy has moved to v3 with a brand new module system, the dev server changed, and my whole workflow has shifted away from GitHub toward Forgejo and Codeberg. So here’s the refresh.

I haven’t hosted my own site on Neocities for years now, but it’s still home to a huge community of personal sites and homepages, especially folks in the 32-Bit Cafe, so this guide is still very much for them.

This guide aims to help you create a homepage using the static site generator (SSG) 11ty, keep the code in version control, and deploy it to Neocities, first by hand, then automatically.

The homepage that we are creating will take advantage of the Nunjucks templating language, allowing us to create a shared header, navigation and footer across all the pages on our homepage.

We will be creating an about, links, and contact pages before diving in and creating the ability to add a blog and a list of all blog posts on the blog page!

We will structure and style the page with a standard HTML5 boilerplate and some basic CSS that should allow you to add in your unique flavour that we all know you love to do.

Create a new project

First off, from a terminal, confirm that you have Node and NPM installed:

node -v && npm -v
v22.11.0
10.9.0

Create a new directory and cd into it:

mkdir 11ty-neocities && cd 11ty-neocities

Initiate a new project:

npm init -y

Install 11ty:

npm install @11ty/eleventy

Once the 11ty installation is complete, open the project in your favourite code editor:

codium .

You should now be in VSCodium with the following project structure:

11ty-neocities/
├── node_modules/
├── package.json
└── package-lock.json

Open package.json and update the scripts section to the following:

  "scripts": {
    "start": "npx @11ty/eleventy --serve",
    "build": "npx @11ty/eleventy"
  },

We also need to tell Node that this is an ESM project. Add "type": "module" to package.json. The file should look like this:

{
  "name": "11ty-neocities",
  "version": "1.0.0",
  "description": "",
  "type": "module",
  "main": "index.js",
  "scripts": {
    "start": "npx @11ty/eleventy --serve",
    "build": "npx @11ty/eleventy"
  },
  "keywords": [],
  "author": "",
  "license": "ISC",
  "dependencies": {
    "@11ty/eleventy": "^3.1.6"
  }
}

Create an 11ty config file

From the terminal (or VSCodium), create a new file eleventy.config.js at the project root:

touch eleventy.config.js

Open the file in VSCodium and add the following and save:

export default function (eleventyConfig) {
  return {
    dir: {
      input: "src",
      output: "public",
      includes: "_includes",
    },
  };
}

.gitignore

As we’re going to be keeping our homepage code in version control, create a .gitignore file in the project root:

touch .gitignore

Open the file in VSCodium and add the following and save:

# dependencies installed by npm

node_modules

# build artefacts

public

Start building the homepage

Now comes the fun part, building our homepage. 11ty supports a number of templating languages, but the two you’ll reach for most are Markdown and plain HTML. Markdown is the popular choice for content like blog posts: you just write, without <html> tags getting in the way. HTML is handy when you need precise structure. The best part is you can drop HTML straight into a Markdown file and 11ty renders it correctly, so it’s never one or the other.

For the pages that make up the site’s structure (home, about, links, contact) we’ll use HTML, because it maps neatly onto the layouts and partials we’re about to build. When we get to the blog, we’ll write the posts in Markdown, where it shines. Use whichever fits the job.

Create a src directory at the project root and cd into it:

mkdir src && cd src

Create an index.html file in the terminal or VSCodium:

touch index.html

Open the file and add some content:

<html>
  <head>
    <title>My New 11ty Homepage on Neocities!</title>
  </head>
  <body>
    <h1>Hello World</h1>

    <p>
      Check out your cool new static site built with
      <a href="https://11ty.dev">11ty</a> on
      <a href="https://neocities.org/">Neocities</a>.
    </p>
  </body>
</html>

Now from the terminal start 11ty:

npm start

If everything has been configured right so far you should see the following:

> 11ty-neocities@1.0.0 start
> npx @11ty/eleventy --serve

[11ty] Writing public/index.html from ./src/index.html (liquid)
[11ty] Wrote 1 file in 0.03 seconds (v3.1.6)
[11ty] Watching…
[11ty] Server at http://localhost:8080/

Now you can open up http://localhost:8080 and check out your new 11ty homepage! It should look like this:

A Basic Hello World HTML Page

Amazing! But what we want to avoid is having to write out the <html> and <head> and <body> tags on each and every page, and be able to include a site header, navigation and footer so we don’t have to copy and paste the changes across every page each time we update.

Let’s checkout templating a layout!

Create a base layout

Create a new directory _includes/ in the src/ directory and cd into it:

mkdir _includes && cd _includes

Create a file base.njk in the terminal or VSCodium:

touch base.njk

Open the file and add the following:

<!DOCTYPE html>
<html lang="en">
  <head>
    <meta charset="UTF-8" />
    <meta name="viewport" content="width=device-width, initial-scale=1.0" />
    <title>{{ title }}</title>
  </head>
  <body>
    <header>
      <h1>{{ title }}</h1>
    </header>
    <main>{{ content | safe }}</main>
  </body>
</html>

Now, head back to the index.html file you created earlier, delete the contents and add some front matter and some content:

---
title: Hello World!
layout: base.njk
---

<p>
  Check out your cool new static site built with
  <a href="https://11ty.dev">11ty</a> on
  <a href="https://neocities.org/">Neocities</a>.
</p>

<p>This homepage template is perfect for:</p>

<ul>
  <li>Creating your own space on the web</li>
  <li>Expressing yourself</li>
  <li>Displaying all the gifs you've collected</li>
</ul>

<h2>Why do you want a homepage?</h2>
<p>The web was made for personal homepages, make this one yours</p>

If you’ve kept 11ty running and the browser running it should look like this:

A Basic Hello World HTML Page Using a Template

Amazing! Now lets create the additional pages for our homepage.

Create the following pages in the src/ directory with the terminal or VSCodium:

touch about.html && touch links.html && touch contact.html

Open each of them up and add in some front matter and content:

about.html:

---
title: About Me
layout: base.njk
---

<p>Heya 👋 this is my homepage.</p>

links.html:

---
title: Links
layout: base.njk
---

<p>These are some of my favourite websites 🔗</p>
<ul>
  <li><a href="https://flamedfury.com">fLaMEdFury.com</a></li>
  <li><a href="https://11ty.dev">11ty</a></li>
  <li><a href="https://neocities.org">Neocities</a></li>
  <li><a href="https://32bit.cafe/">The 32-Bit Cafe</a></li>
</ul>

contact.html:

---
title: Contact Me
layout: base.njk
---

<p>Heya 👋 this is my contact page</p>

You should now be able to browse each of these pages if you kept 11ty running on the following urls:

http://localhost:8080/about/
http://localhost:8080/links/
http://localhost:8080/contact/

Great stuff, but that’s no use without a navigation! Let’s take a look at partials and create a shared header, navigation, and footer to bring our homepage together.

Creating our partials

In the terminal cd into _includes/ and create three partial files:

cd _includes && touch header.njk && touch navigation.njk && touch footer.njk

Open each of them up and add some content:

header.njk:

<h1>Welcome to my Homepage</h1>

navigation.njk

<a href="/">Home</a>
<a href="/about/">About</a>
<a href="/links/">Links</a>
<a href="/blog/">Blog</a>
<a href="/contact/">Contact</a>

footer.njk:

<p>This is my footer | © 2026 Me.</p>

Once our partials are created, open base.njk again and update it to include our new elements and partials:

base.njk:

<!DOCTYPE html>
<html lang="en">

<head>
  <meta charset="UTF-8" />
  <meta name="viewport" content="width=device-width, initial-scale=1.0" />

  <title>{{ title }}</title>
</head>

<body>
  <header>{% include 'header.njk' %}</header>

  <nav>{% include 'navigation.njk' %}</nav>

  <main>
    <h1>{{ title }}</h1>
    {{ content | safe }}
  </main>

  <footer>{% include 'footer.njk' %}</footer>
</body>

</html>

If you’ve kept 11ty running and the browser running it should look like this:

A Basic Hello World HTML Page Using a Template and Partials

Amazing! Now lets add the blog.

Creating the blog

Blog posts are mostly prose, so this is where Markdown earns its keep. We’ll write the posts as .md files and let 11ty turn them into pages.

Create a new directory blog in the src directory and cd into it:

mkdir blog && cd blog

Create the following files in the src/blog directory with the terminal or VSCodium:

touch my-first-post.md && touch my-second-post.md && touch my-third-post.md && touch blog.json

Awesome, Open each of them up in VSCodium and add the following:

my-first-post.md:

---
title: My First Blog Post
---

This is my first blog post

my-second-post.md:

---
title: My Second Blog Post
---

This is my second blog post

my-third-post.md

---
title: My Third Blog Post
---

This is my third and final blog post

blog.json

{
  "layout": "blog"
}

We better create a blog layout so it renders!

Head back to the _includes directory to create a new layout file:

cd ../_includes && touch blog.njk

Open blog.njk up in VSCodium and add the following:

blog.njk:

---
layout: base.njk
---

<article>{{ content | safe }}</article>

Check that your blog posts are loading:

Amazing right? But to make it a blog, we need a blog page that lists all of our blog posts. We can do this with a tags collection:

Open blog.json again and add a key called tags with a value of blog:

blog.json:

{
  "layout": "blog",
  "tags": "blog"
}

Now 11ty has created a collection called blog and all we have to do is list it.

Head back to the src/ directory and create a blog.html file:

cd .. && touch blog.html

Open it and add the following:

blog.html:

---
title: This Is My Blog
layout: base.njk
---

These are all of my amazing blog posts, enjoy!
<ul>
  {% for post in collections.blog | reverse %}
  <li>
    <a href="{{ post.url }}">{{ post.data.title }}</a>
  </li>
  {% endfor %}
</ul>

If you’ve kept 11ty running and the browser running it should look like this:

A Basic Blog List Page

Amazing huh?

Add some styles

Great, so far we have a fully functional home page, but it doesn’t look quite right. We need a style sheet. You can use the one below as an example, it’s basic styling with some modern techniques, or just throw in your own!

Create a new css directory in src, cd into it and create styles.css:

mkdir css && cd css && touch styles.css

Open styles.css in VSCodium and add the following:

styles.css:

:root {
  /* Let the browser handle light and dark automatically */
  color-scheme: light dark;

  /* Two system-font stacks: sans for body, serif for headings */
  --font-body: system-ui, sans-serif;
  --font-head: ui-serif, Georgia, "Iowan Old Style", serif;

  /* Purple/pink palette. light-dark() picks the light value first, dark second */
  --bg: light-dark(#fdf4fa, #1a141f);
  --text: light-dark(#2a1f2d, #ece0ef);
  --muted: light-dark(#6f5d77, #a892b0);
  --accent: light-dark(#a21caf, #ff7ab6);
  --border: light-dark(#ecd9ec, #3a2f40);
}

*,
*::before,
*::after {
  box-sizing: border-box;
}

body {
  font-family: var(--font-body);
  color: var(--text);
  background: var(--bg);
  font-size: 1.15rem;
  line-height: 1.6;
  max-width: 40rem;
  margin-inline: auto;
  padding: 0 1rem;
}

/* Site header */
header {
  text-align: center;
  padding-block: 2rem;
}

/* Navigation */
nav {
  display: flex;
  flex-wrap: wrap;
  justify-content: center;
  gap: 1.5rem;
  padding-bottom: 1.5rem;
  border-bottom: 1px solid var(--border);
}

/* Main content */
main {
  padding-block: 2rem;
}

/* Site footer */
footer {
  margin-top: 4rem;
  padding-block: 2rem;
  text-align: center;
  color: var(--muted);
  font-size: 0.9rem;
  border-top: 1px solid var(--border);
}

/* Links */
a {
  color: var(--accent);
  text-underline-offset: 0.18em;
}

a:hover {
  text-decoration: none;
}

/* Headings */
h1,
h2 {
  font-family: var(--font-head);
  line-height: 1.1;
}

h1 {
  font-size: 2.5rem;
}

h2 {
  font-size: 1.8rem;
  margin-top: 2.5rem;
}

Now we need to include the style sheet in our base.njk layout file. Open it up and add <link rel="stylesheet" href="/css/styles.css" /> to the <head>:

_includes/base.njk:

<!DOCTYPE html>
<html lang="en">

<head>
  <meta charset="UTF-8" />
  <meta name="viewport" content="width=device-width, initial-scale=1.0" />
  <link rel="stylesheet" href="/css/styles.css" />
  <title>{{ title }}</title>
</head>

<body>
  <header>{% include 'header.njk' %}</header>

  <nav>{% include 'navigation.njk' %}</nav>

  <main>
    <h1>{{ title }}</h1>
    {{ content | safe }}
  </main>

  <footer>{% include 'footer.njk' %}</footer>
</body>

</html>

You would have noticed that the stylesheet hasn’t been applied, we have to do one more thing in eleventy.config.js, something called file passthrough copy.

Open eleventy.config.js in VSCodium and add the following:

export default function (eleventyConfig) {
  eleventyConfig.addPassthroughCopy("./src/css");

  return {
    dir: {
      input: "src",
      output: "public",
      includes: "_includes",
    },
  };
}

Because this will come up we may as well create the directories and add in the configuration for our images, fonts and JavaScript files.

Create the following directories in src:

mkdir img && mkdir fonts && mkdir js

Update eleventy.config.js again:

export default function (eleventyConfig) {
  eleventyConfig.addPassthroughCopy("./src/css");
  eleventyConfig.addPassthroughCopy("./src/img");
  eleventyConfig.addPassthroughCopy("./src/fonts");
  eleventyConfig.addPassthroughCopy("./src/js");

  return {
    dir: {
      input: "src",
      output: "public",
      includes: "_includes",
    },
  };
}

Just make sure you put all your static files in the appropriate directory and you’ll be good.

So finally, if you’ve kept 11ty running and the browser running it should look like this:

A Nicely Styled Homepage

Yours will look a little different depending on the colours and fonts you chose above. Now we have a homepage we’re happy with, let’s get it online.

Deploy to Neocities

There are two ways to get your site onto Neocities. We’ll start with the simplest, pushing it from your terminal by hand, then automate it so a deploy happens every time you commit.

Build the site

Whichever method you choose, first build a fresh copy of your site:

npm run build

This writes the finished HTML, CSS and assets to the public directory. That’s the folder we deploy.

The simple way: the Neocities CLI

Neocities provides a command-line tool that lets you push your site straight from your terminal. It’s a Ruby gem, so you’ll need Ruby installed.

gem install neocities

The first time you run a command it’ll ask for your Neocities username and password, then store an API key locally so you don’t have to log in again.

Push the contents of your public directory:

neocities push public

That’s it, your homepage is live. For a lot of people this is all you need. Build, push, done.

The automated way: Forgejo Actions

Pushing by hand is fine, but it’s even nicer to have your site rebuild and deploy itself every time you commit a change. We can do that with Forgejo Actions, the built-in CI for Forgejo. If you self-host Forgejo this runs on your own runner; if you don’t self-host, Codeberg offers the same thing (more on that below).

First, push your project to a repository on your Forgejo instance. Then grab your Neocities API key from your account settings (Manage Site Settings → API Key) and add it to your repository as a secret named NEOCITIES_API_KEY (Repository → Settings → Actions → Secrets).

Now create a workflow file at .forgejo/workflows/deploy.yml:

name: Deploy to Neocities

on:
  push:
    branches:
      - main

# only run one deploy at a time
concurrency:
  group: deploy-neocities
  cancel-in-progress: true

jobs:
  deploy:
    runs-on: docker
    steps:
      - name: Checkout
        uses: https://code.forgejo.org/actions/checkout@v4

      - name: Set up Node
        uses: https://code.forgejo.org/actions/setup-node@v4
        with:
          node-version: 22

      - name: Install and build
        run: |
          npm ci
          npm run build

      - name: Deploy to Neocities
        uses: https://github.com/bcomnes/deploy-to-neocities@v3
        with:
          api_token: ${{ secrets.NEOCITIES_API_KEY }}
          dist_dir: public
          cleanup: true

Commit and push the workflow file. From now on, every push to main rebuilds your site and deploys it to Neocities automatically.

Not self-hosting? Use Codeberg

If you don’t run your own Forgejo instance, Codeberg is a free, community-run home for your code and runs the very same Forgejo Actions. The workflow file above works as-is. Push your project to a Codeberg repo, add the NEOCITIES_API_KEY secret in the repository settings, and you’re away. You may need to enable Actions for your repository first; see the Codeberg CI documentation for details.

Bringing your existing site across

Already have a homepage you’ve been hand-coding on Neocities? You don’t have to start from scratch. Eleventy is happy to take what you’ve got and slot it into this structure.

Copy each existing page into src/ (your old index.html becomes src/index.html, and so on). Then move the parts every page repeats, the <head>, header, nav and footer, into base.njk and the partials you built earlier. Delete that boilerplate from each page and add a little front matter at the top:

---
title: About Me
layout: base.njk
---

Whatever’s left in the file is just that page’s own content, and the layout wraps it.

Your CSS goes in src/css/, images in src/img/, and fonts in src/fonts/. The passthrough copy we set up earlier ships them straight to public/.

If a page is mostly writing, paste the body into a .md file instead of .html. Any fiddly HTML, like an embed or some custom markup, can stay exactly as it is and 11ty will render the Markdown around it.

Run npm run build, check public/ looks the way you expect, then push it live with the Neocities CLI or your Forgejo Actions workflow. Same site you already had, now with layouts, partials and a build step doing the repetitive work for you.

Reference: I created the original version of this guide based heavily on these existing guides, and they’re still well worth a read:

Without these, I wouldn’t even know how to write down what I needed to.

Hey, thanks for reading this post in your feed reader! Want to chat? Reply by email or add me on XMPP, or send a webmention. Check out the posts archive on the website.

Two Way TV - product photos of 1997's hottest gadget - Terence Eden’s Blog

2026-06-16T11:34:18.000Z

Back in the late 1990s, I did a brief stint of work experience at the BBC. One of the most memorable moments was sitting in on a meeting about early forms of Interactive TV.

I saw a demo of "Two Way TV". A flimsy grey box which (somehow) integrated with your OnDigital TV Box and connected to a server via a modem. If you were watching "Who Wants To Be A Millionaire" you could play along at home, send in your answers in realtime, and win REAL CASH PRIZES!!!

An anonymous benefactor read my blog post about the tech, had a pootle through their loft, and found one of the trial boxes they'd been sent in June 1997. With their kind permission, here are some photos of the future we never got.

The unit came in a chunky box with enough logos to convince you it was safe to plug in to the phone network.

Flipping it over, we see a little more of the tech-specs and a defunct barcode (5033936000023 for anyone searching) and product number (SD2044N).

Mmmm! NICAM! So, what did the Set Top Box (STB) actually look like:

Pretty much the same as every other Digital TV STB of the era. A featureless grey slab. Here's a closer view.

How was it all connected? Here's the rear of the box:

Aerial in and out - I assume it had a digital decoder in it, but could pass the analogue channels through to the TV.

Stereo out, for plugging in to your sound system.

SCART in and out. I assume that let you connect your VCR or games console in pass-through mode.

Keyboard looks like a PS/2 port - which would have been the standard at the time. Comms and Remove Receiver both appear to be 8 pin serial connectors.

Finally, there's a standard telephone port for connecting to the dial-up service which makes it all work.

So, you've plugged in all the wires, how do you actually play the games? The unit comes with two controllers - one red and one blue.

The four buttons (triangle, square, circle, lozenge) were for answering on-screen questions. There's an up and down scroller in the middle and a help button above it. I don't know what the large grey circle does.

But what's the peculiar button on the coloured background? That's an early cursor control! Commercially available touchscreens were still in their infancy. This physical controller allowed you to position a digital cursor on screen. Nifty!

Also in the box was a TV guide:

A getting started leaflet:

There's also a service manual:

I don't know what the difference was between SD2044 and SD2044N. Do you?

How much would this amazing interactive experience cost you? I've no idea about the upfront payment for the kit, but there's a Direct Debit form for the monthly subscription.

That's about £14 in today's money.

So there you have it! A snapshot of 1997's vision of the future. From my understanding, the box was never a hit with the public. Two Way TV pivoted to other forms of interactive content like premium-rate phone-ins before going bust in 2003.

If you have any more memories of the service, or interesting photos, please leave a comment or get in touch with me.

The Filing Association of New York - Johnny.Decimal

2026-06-16T08:53:20.000Z

Via Don, news of the annual meeting of the Filing Association of New York. The after-dinner at Gonfaroni's – closed for 96 years, sadly – sounds like a hoot.

Of course this is charming and quaint, but it makes obvious something that I think about all the time: the fact that, in the past, people used to actively work on the problem of filing. It was a profession; there was a local Association that met regularly.

These filing associations turned into organisations like ARMA, the Association of Records Managers and Administrators. So it's not like we totally forgot how to file stuff. But – at least in my professional experience – it's not something the normal office worker is ever exposed to.

And when filing is something you're forced to do, it tends to be in service of checking a compliance box. Those of us who've had the pleasure of using an enterprise 'document management system' – looking at you, HP TRIM – know that they rarely make a thing easier to find. They just make your mistakes harder to correct.¹

Yet it remains the situation that the average office worker is the one creating volumes of stuff; this worker being given little to no training in the still-essential field of filing. No wonder our systems are such a mess.

No wonder nobody can find anything any more.

Every single such system hosting a folder named z_deleted. ↩

📝 2026-06-16 08:11: Sun's out, so there's only one way to travel to the office... ☀️ - Kev Quirk

2026-06-16T07:11:00.000Z

Sun's out, so there's only one way to travel to the office...

☀️

Thanks for reading this post via RSS. RSS is ace, and so are you. ❤️

You can reply to this post by email, or leave a comment.

📝 2026-06-16 07:55: Like we don't have enough animals already. This little black blob will be joining us... - Kev Quirk

2026-06-16T06:55:00.000Z

Like we don't have enough animals already. This little black blob will be joining us on August.

Thanks for reading this post via RSS. RSS is ace, and so are you. ❤️

You can reply to this post by email, or leave a comment.

Read "Anthropic's Safety Superpower" - Molly White's activity feed

2026-06-16T01:19:25.000Z

Read:

“Anthropic's Safety Superpower”.

Ben Thompson in Stratechery. Published June 15, 2026.

Here’s the thing about these safety justifications: I think they work because, to Anthropic, they aren’t justifications. The company really believes that they are the only ones who believe in super intelligence, and thus are the only ones who are sufficiently concerned about the dangers. That excuses decision after decision, policy after policy, and confrontation after confrontation that, to people on the outside, look like a bizarre combination of cynicism and naiveté. The contrast to OpenAI is massive: I think that one way to understand how and why OpenAI lost its lead is that, in the years following the release of ChatGPT, the company has been at war with itself internally as what used to be a research lab was suddenly seized with the burden of being the accidental consumer tech company; to the extent OpenAI solved that conflict, it was by bleeding huge amounts of talent to Anthropic in particular. Anthropic, on the other hand, has perfect alignment between talent and mission and business. The company gets to sell to researchers the creation of a machine god, with the mantle of being the sort of person who cares about the dangers and is smart enough to navigate them on behalf of humanity; that every policy change that falls out of that happens to be great for business is the most beautiful coincidence in the world.

Read "Choosing to Stay Human" - Molly White's activity feed

2026-06-16T00:55:57.000Z

Read:

“Choosing to Stay Human”.

Ethan Mollick in One Useful Thing. Published May 26, 2026.

To be clear, I am cool with a lot of cognitive surrender. I don’t remember phone numbers anymore because my phone does that for me. I am happy my kids didn’t need to learn cursive. I am fine with calculators doing my daily math and my computer figuring out how to schedule my classes. These were once useful skills, but we were probably right to get rid of them. AI is different because the technology is general enough that virtually any cognitive task can be offloaded into it to some degree. I don’t want to be too precious about writing: there is no principle that says a polished email draft has to come out of a human mind any more than a column of arithmetic has to. But we don’t want to give up everything, and that we mostly don’t know yet, for any specific task, what is important and what is not. Deciding that is going to be a real challenge.

Note published on June 15, 2026 at 5:30 PM UTC - Molly White's activity feed

2026-06-15T17:30:00.000Z

Bloggers, can we make better titles for our posts? - Kev Quirk

2026-06-15T14:29:00.000Z

Bloggers, can we make better titles for our posts?

by Michael Harley

Michael makes the case for us bloggers to use better titles when writing our posts as it helps discovery.

Read post ➡

I agree with Michael on this, but I realised that since adding other post types to my RSS feed I too am guilty of this, as my notes posts only show the date and time of the post in the RSS feed.

No more!

I've just pushed an update to my RSS feed that shows the first 15 words of the note after the date and time, which hopefully makes things more descriptive.

That aside, good post by Michael, you should go check out his blog. 🙃

P.S. apologies for any RSS reader spam.

Thanks for reading this post via RSS. RSS is ace, and so are you. ❤️

You can reply to this post by email, or leave a comment.

[RSS Club] What happens to old posts? - Terence Eden’s Blog

2026-06-15T11:34:34.000Z

Welcome to RSS Club! These posts are only available to RSS and Atom subscribers. You can read more about the idea at Dave Rupert's site.

I recently received an email from a distraught reader:

I was going through my recent bookmarks and I found https://shkspr.mobi/blog/2026/04/rss-club-for-wordpress which I had clearly saved to reference again later.

However, as with the nature of RSS Club™, I can't just revisit that post in my browser. I'd long since deleted the RSS item in my feed and re-adding your entire feed only fetched the last ~20 items, of which this wasn't one.

Oh no! It is true that my feed only goes back 20 entries. As a consequence, older posts are lost in the Time Vortex - much like several episodes of Doctor Who.

My first thought was "is this a bug or a feature?" Perhaps those posts should be ephemeral. It is possible that they've been archived - but it is equally possible that they've drifted away on the breeze like a child's balloon. Do I want people rummaging in the archives to get old club posts?

I think my answer is… I'm happy for them to be inaccessible for the casual reader.

But there are a few ways the determined scholar can find older posts.

Manually Save The Contents

Your feed reader probably lets you store a permanent copy of a post. You should do that if you want a local version available.

Archive.org

The Internet Archive regularly grabs a copy of my RSS feed. For example - https://web.archive.org/web/20260000000000*/https://shkspr.mobi/blog/feed/atom/

You can peruse older versions to your heart's content.

JSON API

WordPress's JSON API contains the full contents of the post (albeit in a slightly verbose JSON format). You need to know the post's ID.

For example, the JSON for this post itself can be found at https://shkspr.mobi/blog/wp-json/wp/v2/posts/72429

Feed Pagination

By default, you only get 20 items per feed. If you'd like to go to the 2nd page of the feed, you can use:

https://shkspr.mobi/blog/feed/atom/?paged=2

Then ?paged=3 for the next, etc.

Category Feeds

WordPress has build-in support for category-specific feeds. So, if you want to subscribe to only the RSS Club posts, you may use https://shkspr.mobi/blog/category/rss-club/feed/

This also works with tags, for example https://shkspr.mobi/blog/tag/rss-club/feed/

They should both be the same, but I primarily use categories to differentiate between club and non-club posts.

Email

Feel free to email me if you can't find something. My contact details are on https://edent.tel/

This week at JDHQ – 2026-06-15 - Johnny.Decimal

2026-06-15T04:55:55.000Z

Originally sent to the mailing list.

As promised (to ourselves), we've had a bit of time off here in Taiwan. So this is just a quick note to remind you that the first of the weekly Small Business sessions is this Thursday 18th June at 16:00 +8.

That's morning if you're in Europe, daytime for EMEA and into Asia, and evening in Oceania. The session next Tuesday 23rd June at 06:00 +8 is nicer for those of you in the Americas.

Zoom links for both of those are on the SBS events page at JDHQ – subscribe to the calendar there for the most up-to-date information.

We'll be kicking off the annual cycle with something easy, just to test out the process: your travel administration and records. As a reminder, the calendar has been designed to touch every aspect of your business over the course of the year. It's a light check-in to make sure that everything's in order and that you haven't forgotten, say, to renew your company's annual travel insurance policy.

I'll be sending those of you who asked for the Baserow template a link in the next couple of days. We'll go through it in the session so there's no pre-work required.

Of course if you're not already in the Small Business System, there's still time to upgrade your account. See you Thursday.

AI Blog Questions Challenge - Kev Quirk

2026-06-14T15:21:00.000Z

Ava did it, then Manu did it, then Thomas did it. So now I'm joining in. I haven't written a lot about AI on this site, so maybe this will give you all some insights as to what my opinions are on it all.

1. How was your first experience with AI models?

I don't remember my first time actually using an LLM, but I do remember what my feelings were in the early days of it all. Like many others, I tried Chatty Geeps and just talked shit to it. I found it to be far too sycophantic, but interesting nonetheless.

2. Do you use AI or are you completely against using it?

I do. I currently have a subscription to Google's Gemini and I use it for 2 main purposes:

To help when I get stuck coding.
As a search engine when I'm looking for something niche or complex.

3. Do you have any preference among different models, for example Claude vs ChatGPT? If yes, how do you choose?

Not really. I've used Chatty Geeps in the past, then Claude, and like I said above, I'm currently using Gemini. I've found them to all be similarly capable for my needs.

The main reason I'm currently using Gemini is because it comes with my Google One subscription.

4. What aspect of AI models do you like and what do you not like?

99.99% of the time I use an LLM in one of the two ways I mentioned above. And I think they work really well for this. As a result of using an LLM, I've been able to learn so much about coding because I can get it to explain problems to me, which I find really helpful.

Similarly, when I'm searching for something niche or complex, an LLM is very helpful. It saves me from wading through page after page of SEO garbage, Reddit posts, or stale forums. It does all that for me and generally gives me either the correct answer, or points me in the right direction so I can go from there.

There's two things that really irks me about LLMs though. The first is that they're so damned sycophantic. Not everything is a great idea, despite what the models say.

Second is the way they always end a reply with a question. Why is this necessary? Just give me the answer I need, then pipe down.

5. How do you feel about AI generated images? Does it annoy you if someone uses them in a blog post?

I've actually written about this before. Using them for feature images on a blog…meh. More generally, I think generated images have their use at times.

For example, we're about to have some work done on the stables and my wife wanted to get an idea of what it would look like. So she took a photo of the stables, then asked Gemini to add the details of the work we're getting done. It did a great job and gave us a really good idea of what to expect.

I don't generate images for fun though. I think it's wasteful and I don't really see the point of it.

I also don't think that AI generated images are art.

6. Internet is flooded with AI slop now, full of generated text, images, audio, and videos. How do you filter it from authentic human creation? Do you have a strategy?

I have no strategy. When I come across it, I generally just ignore it and move on.

For the record, I really dislike the word "slop" too. I think a lot of what LLMs generate is really useful, so to bundle it all in the "slop" bucket is unfair in my opinion.

7. Are you hopeful for a better future with A.I. or a dystopian one?

I think it will go the same way as all the fads that have come before it. The bubble will burst, many of these AI companies will die off, and we will be left with a handful of large providers that continue to iterate and improve upon it.

It's very similar to blockchain - 5 years ago companies were popping up and trying to shoehorn it into everything. That bubble burst and we rarely hear of it now. It still has it's uses, but it generally goes under the radar these days.

Long-term I think LLMs and AI have a future. They're very useful, but I don't think they're a magic pill for all our problems. Far from it.

Your turn!

Feel free to write your own response to this post - I've enjoyed reading the responses I've seen so far, which is why I decided to get involved in the first place.

If you do decide to write a reply, here's the questions in a handy-dandy, pastable (is that even a word?) format:

1. How was your first experience with AI models?
2. Do you use AI or are you completely against using it?
3. Do you have any preference among different models, for example Claude vs ChatGPT? If yes, how do you choose?
4. What aspect of AI models do you like and what do you not like?
5. How do you feel about AI generated images? Does it annoy you if someone uses them in a blog post?
6. Internet is flooded with AI slop now, full of generated text, images, audio, and videos. How do you filter it from authentic human creation? Do you have a strategy?
7. Are you hopeful for a better future with A.I. or a dystopian one?

Thanks for reading this post via RSS. RSS is ace, and so are you. ❤️

You can reply to this post by email, or leave a comment.

📝 2026-06-14 12:49 - Kev Quirk

2026-06-14T11:49:00.000Z

Can someone who's more green fingered than me tell me is this is giant hogweed please?

It's all over one of our fields, so if it is I'll need to get someone in to get rid of it.

Thanks for reading this post via RSS. RSS is ace, and so are you. ❤️

You can reply to this post by email, or leave a comment.

Did Frank Sinatra really think "Something" was a Lennon/McCartney song? - Terence Eden’s Blog

2026-06-14T11:34:16.000Z

Read enough articles about The Beatles and you'll repeatedly hit the claim that Frank Sinatra frequently introduced his cover of George Harrison's "Something" as his "favourite Lennon & McCartney number."

Much like the misquote about Ringo not being the best drummer in The Beatles, I think this might be one of those semi-apocryphal lines which has taken on a life of its own.

Here's what Paul McCartney has to say in The Beatles Anthology, Episode 4.

https://shkspr.mobi/blog/wp-content/uploads/2026/05/Sinatra.webm

That was broadcast in 1995 - so we need to look for sources from before that.

There's not much Internet before the mid-1990s. Google's mismanagement of the USENET archives is a cultural obscenity. Nevertheless, we can find a few references which predate McCartney's broadcast.

1994-12-26

Frankie used to introduce "Something" as his "tribute to Mr. Lennon and Mr. McCartney" ;^)

1990-03-05

In fact, a friend of mine (a supposed Beatle fan; turns out she's really just a L/M fan), were having a discussion about this very subject, she, just like Frank Sinatra, didn't know that George wrote "Something." Duh.

So it was certainly a proto-meme back then.

Of the thousands of Beatles books, I can't find any from before the mid-1990s which mention Sinatra's misattribution.

For example, 1994's The Complete Guide to the Music of the Beatles simply says:

Similarly, there are plenty of books and articles about Sinatra - lots of them talk about Something, but never this supposed misrepresentation. In 1980's New York Magazine, Sinatra is interviewed and says:

There are many videos of Sinatra singing Something on YouTube - none of them have him introducing the song as a Lennon/McCartney number.

Indeed, here's one where he introduces it as being by George Harrison.

I think that's 1982's The Concert for the Americas - in the Dominican Republic.

Here's a 1985 concert where he introduces it as being by George Harrison of The Beatles.

Way back in 1978 at Sinatra's Caesar’s Palace Concert, he introduces it with "George Harrison wrote it" and finishes with "by George Harrison".

Even back in 1975, during a concert in Jerusalem he was crediting Harrison, saying:

Every one of The Beatles was a very talented young man individually. And here's an example of George Harrison with a great love song."

I've now listened to dozens of recordings of Sinatra singing Something live and in none of them does he so much as mention John Lennon or Paul McCartney.

So is the quote apocryphal? Possibly not!

Less than a year after John Lennon was murdered, Sinatra treated Carnegie Hall⁰ to a series of 11 concerts.

On 10th September 1981, John Rockwell published Pop: Sinatra at Carnegie - a review of the opening night of Sinatra's concert series at New York City's Carnegie Hall:

Also on the 10th, a clutch of US papers reproduced a story by the inimitable Mary Campbell of the Associated Press.

Most of the syndicated versions leave out the parenthetical remarks.

On the 11th, Patricia O'Haire published a somewhat snide review of the September 9th concert in The New York Daily News

On 14th September 1981, a British newspaper re-reported the comment:

That's the Daily Express by Rob Benson, their Los Angeles correspondent¹.

By the 29th of September 1981, the story had made it to Australian Financial Times' The Bulletin.

It's unclear how many of those journalists were actually at the concert. I assume John Rockwell, Mary Campbell, and Patricia O'Haire were as they published fairly detailed reviews.

Tracking down a set-list for that long-gone concert is tricky. Carnegie Hall themselves get the dates wrong in their archive and say the first performance was on the 8th, and their set-list is sourced from Setlist.fm rather than their own records. The Sinatraphiles mailing list has a set-list for the 9th which does include "Something".

There's a purported recording of the September 10th concert with a set-list on the reverse:

There's no "Lennon" song - the only Beatles number is "Something". Let's take a listen to the introduction from that bootleg recording.

🔊 Something
🎤 Frank Sinatra

💾 Download this audio file.

"A beautiful song by George Harrison. Maybe one of the best love songs ever written."²

So, that's a handful of contemporary sources who mention that Frank Sinatra once introduced "Something" as being composed by someone other than Harrison.

The only recording is of the concert the next day - and it doesn't includes that "blooper".

There's no other mentions I can find which directly cite a specific concert or performance.

Did Sinatra ever say it was his "favourite Lennon and McCartney song"? He sang in thousands of shows³, not all of which were recorded⁴, so it is entirely possible he mentioned it. But you'd expect more than a few reporters would write about it, wouldn't you?

The origin of the "quote", as far as I can tell, is from an interview Paul McCartney gave to David Hinckley in the New York Daily News on 21st October 1984.

That's the first time that I can see "Something" mentioned as Sinatra's "favorite Lennon-McCartney song".

I went rummaging through some reviews of Frank's concert performance which included "Something" in the set list.

His concert at the Palladium:

And Frank sings 'Something'. It's OK. The Vanilla Fudge were more adept at Beatle rewrites however.

Chris Salewicz. "Frank Sinatra: Palladium, London". New Musical Express (1975).

His concert at the Royal Albert Hall:

Superb renditions of Jim Webb's 'Didn't We?' and Harrison's 'Something' were recreated with a totally unique empathy. "Real Songs, beautiful songs", he said fervently, no trace of show-biz cant.

Max Bell. "Frank Sinatra: Royal Albert Hall, London". New Musical Express (1975).

And another report of the same gig:

Jimmy Webb's 'Didn't We' and the classic 'Nice And Easy', were exceptionally good, standing out easily among lacklustre renditions of 'Something', 'Strangers In The Night' and a David Gates song. In between, Sinatra delivered various controversial raps designed to instigate audience loyalties but proved that Sinatra should open his mouth only when singing.

Barbara Charone. "Frank Sinatra: Royal Albert Hall, London". Sounds (1975).

I've read dozens of gig reviews of old Sinatra concerts and they all contain various levels of snark about his performance, song choice, and politics - so you'd expect British reporters would have picked up on the misattribution, wouldn't you?

Instead, there's two slightly contradictory reports of one single concert and no suggestion that Sinatra himself said it was his "favorite Lennon-McCartney song". Given that he repeatedly credited George Harrison in the decade leading up to that concert, I think it is fair to say the "quote" has taken on a significance far beyond its actual importance.

If you have a recording of Sinatra introducing "Something" as a Lennon/McCartney number - or any other contemporary reports of that - please drop a comment in the box.

After all, Sinatra had a lot of practice! ↩︎
It is odd that the reporter describes Sinatra as "now" singing Something when it had been in his repertoire for over a decade. About the right level of journalistic rigour expected of the Express. ↩︎
Later on, in the introduction to "Luck Be A Lady Tonight", he sarcastically describes Marlon Brando as "America's great baritone!". There are quite a few jokey moments in the performance - so it is entirely possible his Lennon & McCartney remark was a quip. ↩︎
Incidentally, as far as I can tell, Sinatra first sang "Something" in December 1970 on The Dean Martin Show - about a year after its release on Abbey Road. Sinatra's performance doesn't contain him saying anything about the song. ↩︎
I spoke to one collector who said:

I also checked all of the other collectors lists I have, and they do not have it either, I do however have reference to its existence via a notecard that represents a massive collection. What this means is that the concert could exist, but more than likely has never been digitized. Many Sinatra concerts are still stuck on reel to reels from the 70s and 80s and have never been transferred to the digital realm and shared on the internet.

↩︎

Hobbling my app launcher - Johnny.Decimal

2026-06-14T01:47:30.000Z

I've been using an app launcher/switcher since at least 2011 – the date my Alfred v1 licence was created in 1Password – and for unknown years before that, when Quicksilver was the new thing.

App launchers do just that: they help you launch apps. Why grab the mouse and shuffle on over to the Dock when your fingers are already on the keyboard? We've had this capability for decades: Windows 3.0's File Manager had a File → Run… menu, with Windows 95 adding the keyboard shortcut Win-R. From there, type a thing and it'll launch.¹ Alfred, Raycast and their ilk just do this faster and better.

Speed ≠ efficiency

Lately, I've noticed these tools getting in the way. I have Raycast bound to Cmd-Space and testing shows that I can switch from whatever I'm doing to anything else in 150ms: essentially instantly.²

This sounds like a benefit. But when I can switch to anything, instantly, that's what I find myself doing, and way too often. It's like all friction has been removed, and what remains isn't good. What remains is a slippery mess.

The other behaviour I've noticed is an almost manic switching between two apps. Safari, code, Safari, code. Flick-flick-flick between the two. My fingers are so used to these keyboard sequences I feel like I do them subconsciously! I end up in an app and wonder, what am I doing here? (I was previously using the builtin Cmd-Tab switcher for this, but I broke that habit by turning off the keyboard shortcut.)³

Reintroducing friction

Fortunately, Raycast makes it easy to disable certain apps. Because I don't want to quit Raycast entirely: it's far too useful. I use it as an emoji 👋🏼 and Unicode symbol ▷ ■ ◁ picker. At the former it's way better than macOS' builtin feature, and at the latter it fills a gap that the OS simply doesn't provide.⁴ It's also my clipboard manager, my snippet expander, and the list goes on.⁵

But you can turn off its ability to launch an app by just unchecking the Enabled box in Preferences → Extensions for that app. So that's what I've done for Safari and Discord, to start. As I notice other apps that are distraction magnets, I'll disable those.

Instead, it's back to the trackpad. Swipe-swipe-swipe to the Dock I'll go, and I'll click the icon and it'll take about a second and that already feels like an eternity. And my brain is enjoying that.

In the early versions it wasn't quite this simple. Typing excel won't launch Excel, for example. You need to know the full name of the executable, excel.exe. In modern versions of Windows you should be using the newer search features to do this, although Win-R still brings up trusty Run. ↩
I recorded my screen as I invoked Raycast, typed s, and hit return. This switches me to Safari. Raycast is only on screen for 7 frames of the 50 frames/sec recording. ↩
Using the wonderful Supercharge by Sindre Sorhus. The setting Tweaks → Disable Command+Tab app switcher exists because I asked him to add it. :-) ↩
For emoji, assign a keyboard shortcut to the Raycast builtin feature Search Emoji & Symbols. I use Ctrl-Opt-I.

For Unicode, install the extension Unicode Symbols Search. I bind that to Ctrl-Opt-U. ↩
Bind the builtin Clipboard History to Ctrl-Opt-C. ↩

Read "Why "Book-Shaming" Won't Solve the Children's Literacy Crisis" - Molly White's activity feed

2026-06-13T15:22:31.000Z

Read:

“Why "Book-Shaming" Won't Solve the Children's Literacy Crisis”.

Jessica Winter in The New Yorker. Published June 12, 2026.

Barnett’s concerns about literary merit and professional esteem may be timeless, but they are not terribly timely; they seem to float high above the current on-the-ground realities of what many educators and researchers agree to be a literacy crisis. In urging his audience to see children’s books as “real books,” Barnett skips over larger, more pressing questions about why so many children aren’t reading books at all, real or otherwise.

SwissMicros calculators - Johnny.Decimal

2026-06-13T01:29:10.000Z

I have no use for a desk calculator, but as soon as I have a desk again I'm buying a SwissMicros DM41L.

Figure 0217A. SwissMicros DM41L.

If you need scientific, the DM42n has you covered.

Figure 0217B. SwissMicros DM42n.

Clouds; colour - James' Coffee Blog

2026-06-13T00:00:00.000Z

Approaching a junction, I looked up to the sky and saw a hole in the clouds unlike any I have seen before. I usually look up to the stone that must be at least a hundred years old: to the grey buildings – homes – in which I see so many stories. But today the blue sky and the white cloud stood out; the life of the city is made as much in Nature as it is in architecture.

I was on my way to a coffee shop, one I frequent because every time I go it is quiet. Classical music plays in the background. I have been there enough times to know their playlist is limited, but I don’t mind. On the playlist is a rendition of one of my favourite songs from childhood, a song that introduced me to the idea of lightning bugs. Nature illuminates? Awe lives on through the decades.

At a table close to the entrance, I saw the Great Wave; it was someone’s laptop cover. Now I see art wherever I go.

Later in the day, while walking, I heard in the background a child say “why are we here?” to their parent, presumably referring to the physical location. My mind couldn’t help but jump to the existential. Why are we here? Oh! how the same question can take on different meanings through life; we grow into new views and perspectives and answers.

“How is art history?” said someone I have had a crush on for several months, a person to whom, most recently, I have found myself being more open with than many. Earlier this year, I didn’t think I’d see them again, but chance had it that I have had the opportunity to do so a few more times (and potentially once more after today). To her question I replied “I see the world through new eyes everywhere I look.” “Everything is so colourful.”

I had been carrying a veil of tiredness today, accumulated from staying up late to have many wonderful late-night conversations this week. But in the moment when I was asked about my studies, the artistic lens I have been building over the last few years came to the fore. I saw colour in the moment and the world and life. The moment brought a smile to my face. I had been carrying that perspective for so long, bringing it into my writing. But there was something special sharing it with someone in spoken word. Maybe some day I’ll have the words to describe how I feel.

📝 2026-06-12 16:59 - Kev Quirk

2026-06-12T15:59:00.000Z

I can think of worse places to walk the dogs...

Thanks for reading this post via RSS. RSS is ace, and so are you. ❤️

You can reply to this post by email, or leave a comment.

Published on Citation Needed: "Second Circuit rejects Sam Bankman-Fried’s appeal" - Molly White's activity feed

2026-06-12T15:18:13.000Z

Published an issue of Citation Needed:

Second Circuit rejects Sam Bankman-Fried’s appeal

The Second Circuit upholds Bankman-Fried’s conviction and 25-year sentence, leaving few remaining options for the disgraced crypto executive

Shellsharks Blogroll - BlogFlock

Create A Static Site Using 11ty & Deploy to Neocities (2026 Refresh) - The Weblog of fLaMEd

Two Way TV - product photos of 1997's hottest gadget - Terence Eden’s Blog

The Filing Association of New York - Johnny.Decimal

Footnotes

📝 2026-06-16 08:11: Sun's out, so there's only one way to travel to the office... ☀️ - Kev Quirk

📝 2026-06-16 07:55: Like we don't have enough animals already. This little black blob will be joining us... - Kev Quirk

Read "Anthropic's Safety Superpower" - Molly White's activity feed

Read "Choosing to Stay Human" - Molly White's activity feed

Note published on June 15, 2026 at 5:30 PM UTC - Molly White's activity feed

Bloggers, can we make better titles for our posts? - Kev Quirk

Bloggers, can we make better titles for our posts?

[RSS Club] What happens to old posts? - Terence Eden’s Blog

This week at JDHQ – 2026-06-15 - Johnny.Decimal

AI Blog Questions Challenge - Kev Quirk

1. How was your first experience with AI models?

2. Do you use AI or are you completely against using it?

3. Do you have any preference among different models, for example Claude vs ChatGPT? If yes, how do you choose?

4. What aspect of AI models do you like and what do you not like?

5. How do you feel about AI generated images? Does it annoy you if someone uses them in a blog post?

6. Internet is flooded with AI slop now, full of generated text, images, audio, and videos. How do you filter it from authentic human creation? Do you have a strategy?

7. Are you hopeful for a better future with A.I. or a dystopian one?

Your turn!

📝 2026-06-14 12:49 - Kev Quirk

Did Frank Sinatra really think "Something" was a Lennon/McCartney song? - Terence Eden’s Blog

Hobbling my app launcher - Johnny.Decimal

Speed ≠ efficiency

Reintroducing friction

Footnotes

Read "Why "Book-Shaming" Won't Solve the Children's Literacy Crisis" - Molly White's activity feed

SwissMicros calculators - Johnny.Decimal

Clouds; colour - James' Coffee Blog

📝 2026-06-12 16:59 - Kev Quirk

Published on Citation Needed: "Second Circuit rejects Sam Bankman-Fried’s appeal" - Molly White's activity feed