Caleb Hailey

The chicken and egg of RSS and "podcasts"

2026-07-15T23:11:38Z

Anil Dash (@anildash.com) explains why "'Wherever you get your podcasts' is a radical statement" in simple terms:

Being able to say, "wherever you get your podcasts" is a radical statement. Because what it represents is the triumph of exactly the kind of technology that's supposed to be impossible: open, empowering tech that's not owned by any one company, that can't be controlled by any one company, and that allows people to have ownership over their work and their relationship with their audience.

By my count Anil uses the term "podcast" 22 times in this post. He never once mentions the technology that makes it possible: RSS enclosures.¹ Anil is a fearless advocate for the open web. The omission of "RSS" is not a "dis" on open standards. He simply doesn't mention "RSS" because he doesn't have to. Celebrating the success of "podcasting" is inherently a celebration of the success of RSS! If Anil were to write a similar post about the enduring success of RSS, he might not mention HTTP or XML. Just like most posts you will read about HTTP won't mention TCP. I could keep going but I won't belabor the point.

I love the open web. One of the things I love most about the open web is that it's built from layer upon layer of open standard technologies. I especially love the products that not only adopt these technologies but go further and embrace the spirit of openness in the underlying standards. By giving creators and consumers more control and more choice over the way they experience the world wide web, they help to advocate for the open web.

Conversely, one of my least favorite things about the open web is when established voices attempt to suppress new ideas. That's the opposite of advocacy and it only hurts the open web. How does building an app based on the ideals of interoperation hurt the open standards? Conversely, how does criticizing new ideas that are designed to onboard new users to open standards actually help promote said open standards? 🤔

I love RSS. In fact, you could say I have demonstrated irrational dedication to RSS by quitting my day job and spending the last year of my life working on an app that adheres very closely to the RSS and OPML standards (among others). My goal with HyperTexting is to extend the reach of "RSS feed readers" beyond the group of people who have already known and loved them for the past 20 years. I just happen to think that the number of people who haven't tried an RSS feed reader is significantly larger than the number who have. And the group that hasn't tried RSS is made up of people who are overwhelmed with or intimidated by technology. HyperTexting is my experiment to try reaching that audience by focusing on the "why" of RSS (reclaiming control of your news feed) rather than the "how" (Really Simple Syndication as an open standard).

My theory is informed by graphs like this one:

The blue line represents search activity for "podcast", and the red line represents search activity for "rss".

When I see this graph I see the growing success of RSS as a technology, even if the terminology itself is fading in popularity. I'm going to try my best to help and contribute to that success with some new ideas. I'm also going to ignore any gate-keeping criticism² that freezes RSS in time and can only hurt its success.

see the original post by Dave Winer on thetwowayweb.com, via @web.archive.org. ↩︎
hi there, @doc.searls.com 👋 I know you read the footnotes, so here's a footnote just for you. I'm sorry you think my work is a "dis" on RSS, but you're welcome to your opinion. Everybody has one. 💩 I would however like to clarify that I have not once attempted to "re-brand RSS with [my] own app". I have only standardized on the term "feeds" instead of "RSS, Atom, and JSON Feed" (all three of which are supported by HyperTexting) in my app and marketing content. I have already documented the origin of the name HyperTexting and it has nothing to do with rebranding RSS. I have also not blamed anyone for the success of RSS, or suggested that there is "a stink on the standard." I don't know how you reached that conclusion when all I've attempted to do so far is "rally around conventions like Dave [Winer] is promoting with OPML blogrolls." I'm not the first person to suggest that there are still new ideas to be tried around RSS, and I'm already not the last. It's amazing that there are still opportunities (!) for RSS to continue growing after all these years. For my part I'm going to try to help them grow rather than criticize where that growth comes from. 🍻 ↩︎

The social-media-like interface

2026-06-25T20:37:55Z

There's a feeling you get when you are building something that you're very passionate about that can haunt you if you let it: "What if someone else makes something [similar / better / faster]?" It's a particular brand of imposter syndrome that will be familiar to anyone who has ever written¹ an "Introducing X" blog post or tweet.

I just published one such blog post this morning, announcing the launch of @hypertexting.com for iPhone on the App Store. 🎉 It's exciting and scary at the same time. But it almost didn't happen.

I have never felt the aforementioned fear more acutely than this past January when Terry Godier (@terrygodier.com) published an essay called "Phantom Obligation":

Why do RSS readers look like email clients?

If you've used almost any RSS reader in the past two decades, you know this layout intimately. There's a sidebar with your feeds organized into folders. There's a list of items, sorted by date, with little dots indicating what you haven't read yet. There's a reading pane where the content appears when you click.

The shape is so ubiquitous that it feels inevitable. But of course nothing in design is inevitable. Someone made a choice, and then other people followed that choice, and eventually the choice calcified into convention.

As I read the rest of Terry's thoughtful blog post, I experienced simultaneous surges of joy and dread. 🎢 Never before had a single blog post captured so much of my feelings about a technology that I have spent so much time thinking about (RSS). I was practically hugging my laptop as I read it.

And then the dread:

I built an RSS reader with these principles. If you're interested, you can check it out now.

Did Terry beat me to it? Did the project I had been working on for over a year just become an afterthought? Thankfully, it turns out that Terry and I made very similar observations, but those observations inspired very different products.

To be clear, I agree with every single word in Phantom Obligation. But I also think the opportunity for RSS readers is much bigger than solving the unread count problem. Allow me to explain.

The Outlook-like interface

Pull up a chair kids, because "back in my day..." I remember when NetNewsWire came out (still the 🐐 IMO) and the hottest buzzword in tech wasn't "AI", it was "the Outlook-like interface". 🤣 During the early-mid 2000's I worked in tech support at a small software company (my first tech job) and the founder was obsessed with making our app(s) look like Microsoft Outlook. It drove me absolutely crazy. I wouldn't understand why that mattered until later in my career as a Product Manager. It was about meeting people where they were. The "Outlook-like interface" was just an earlier iteration of "Uber for X".

Outlook wasn't "intuitive", but it was familiar to a LOT of people. Today there's nothing more familiar than the social media news feed. On September 5th, 2026 the Facebook News Feed will turn 20 years old! Even though it didn't invent the feed, that launch marked a seminal moment in internet culture history as it shaped much of how we experience the internet down to this day. It's all feeds, all the way down.

For the past few years I've spent an inordinate amount of time thinking about whether a "social-media like interface" is possible for "RSS" (and Atom, and JSON Feed, and OPML). The conclusion I've reached is that this requires more than just a departure from the three-column/inbox-style feed reader and obligation-inducing unread counts. It means taking inspiration from social media profiles, responses, mentions, threads, video, discovery, and even personalization. It also means embracing user privacy and the unique attributes of the open web that social media platforms ~~can't~~ refuse to give us. Finally, it should also lower the barrier to entry for publishing in addition to subscribing.

Executing on that thought process over the past year is what led me to HyperTexting. I hope you will try it, and I hope it helps you escape the algorithm.

Oh, and there's one more thing.

Talking about RSS without saying "RSS"

RSS is a tremendous success story. Without it, the expression "Wherever you get your podcasts" would not be possible. And yet it is! But why has podcasting reached escape velocity in a way that "RSS readers" have not? Why don't more people know that RSS readers are a compelling alternative to algorithmic social media newsfeeds? Is it a branding problem, a product problem, or something else?

I think the answer is much bigger and more complicated than Current or HyperTexting. I also think we can get there by starting small. In order for RSS to ~~succeed~~ reach escape velocity, I think we need to stop saying "RSS".² That's why, if you look closely, you'll see that I've gone out of my way to avoid using terms like "RSS", "Atom", "JSON Feed" or "OPML" in the HyperTexting app and website.

I hope this part of the HyperTexting experiment works, and even more so, I hope it inspires my fellow RSS advocates to keep building! Don't let the imposter syndrome win!

I built an RSS reader with these principles. If you're interested, you can check it out now. 😊

I emphasize the writing here because I suspect there are thousands if not millions of "Introducing X" blog post drafts and abandoned Tweets resulting from letting the imposter syndrome win. Hitting "Publish" requires a certain kind of courage! ↩︎
Hi Dave. 👋 I see you and I'm a huge fan! I suspect you may have thoughts™ on this take, but I think it's possible we might agree more than we disagree. ↩︎

The files (apps) are in the computer (iPhone)

2026-06-24T16:21:03Z

There's a scene in the movie Zoolander where Owen Wilson's character has an epiphany about where some critical information is located. And then he delivers this line:

"[The files] are in the computer. It's so simple."

Where Google and Microsoft's AI strategies revolve around the idea that your data lives inside of their ecosystems, Apple unveiled a mostly device-centric approach that I think more closely aligns with how people think about their "files".

Even though a younger generation of computer users – and I'm using the term "computer" loosely to include smart phones, ~~tablets~~ iPads, and laptop/desktop computers – don't know where their files are saved, it still matters that the files exist.

To the extent that our personal context exists in regular files on disk, Siri AI will have access to that information. Some of those files just so happen to be databases containing our most personal information, including Messages (dot app) we've sent to family and friends, or Notes (dot app) we've jotted down because we wanted to remember them, or Reminders (dot app) we've set so we don't forget to do something.

So far this is all table stakes. All of the major players are racing to build personal AI agents that operate within their walled garden ecosystems. Your mileage will vary depending on how deep you are invested in the ecosystem. What I suspect will differentiate one from another is how capable they are when the context they need is in a third party application, which brings us back to Owen Wilson.

The files are apps, and the apps are in the ~~computer~~ iPhone. It's so simple!

Our phones are the most important computers in our lives. Where is [insert some important piece of information] you need? It's "on your phone". Even if it's really in some third-party app that stores your information using encrypted cloud storage, the app is on your phone, and so the information is accessible from your phone.

The operating system is the super app

Every time I see a headline about the frontier AI labs internal "super app" initiatives, I always think to myself "they're mispronouncing 'operating system'". What Apple showed at WWDC26 with Siri AI, and the underlying OS-level indexing improvements, reveals to me that they are finally entering the "super app" race. But Apple's super app isn't an app at all - it's an operating system.

Will Siri AI work better if you use Messages.app instead of WhatsApp, and Mail.app instead of Gmail, and Notes.app instead of OneNote? Yes, of course. But will it also work better than the alternatives if you use a third-party notes app? Maybe!

Apple's plan to connect Siri AI to the information in third-party apps is the App Intents framework. Whether third-party app developers adopt App Intents or not will probably be determined by how good Siri AI is. If Apple can deliver a compelling personal assistant experience via their super app (iOS 27 with Siri AI), then offering support for Siri AI may become a differentiator that third-party app developers race to adopt.

Developer Betas are still early doors, but the reports are trickling in and it sounds like Siri AI might actually be good (enough).

It's shaping up to be an exciting summer and fall!

Legibility of Effort and The Social Contract of Writing

2026-06-24T14:24:40Z

Nolan Royalty (@eieio.games), writing without the assistance of generative AI, in a post entitled "Legibility of Effort":

What software (and writing, to an extent) is missing now is legibility of effort - the ability to tell at a glance whether something took a human meaningful work.

First of all I cannot emphasize enough how much I appreciate that this was a blog post and not a tweet. The dense hypertextual writing in itself suggests some greater degree of effort. But as Nolan points out, even that isn't a trustworthy signal:

I wrote this blog post in my text editor. It took me a while.

Until recently, “someone cared enough to write this” was an ok heuristic. Plenty of writing on the internet was bad, but you could convince me that you cared about something just by writing it down.

Of course, generating plausible-looking text - or a plausible-looking website - is trivial now.

This reminded me of a post I bookmarked last week by Johanna Larsson (@jola.dev), entitled "The Social Contract of Writing", and the @oxide.computer Request For Discussion document RFD 576 that Johanna linked to:

LLM-generated writing undermines the authenticity of not just one’s writing but of the thinking behind it as well. If the prose is automatically generated, might the ideas be too? The reader can’t be sure — and increasingly, the hallmarks of LLM generation cause readers to turn off (or worse).

Finally, LLM-generated prose undermines a social contract of sorts: absent LLMs, it is presumed that of the reader and the writer, it is the writer that has undertaken the greater intellectual exertion. (That is, it is more work to write than to read!)

I love this "social contract" framing. I've been think about this in terms of the difference between writing for a social media platform versus writing for a website (e.g. a blog post). When content is written for social media, I automatically assume that less effort was put into the writing because less care was put into the longevity and accessibility of the writing. When is the last time you read a month-old tweet? If you did I bet it was embedded in or linked from a blog post, where writing is generally meant to last. And if someone doesn't care about their writing enough to put it somewhere durable, then why should I take the time to read it? Was it posted on social media primarily for the engagement? The likes and retweets and hot takes from people who only ~~read~~ skim the introduction, or who are replying in the comments to try and steal some of the fleeting attention the original post is getting? Conversely, the only way to reply to most of the serious effort writing I read is via email. Writing that can only be replied to by email says something. A writer who wants to avoid effort doesn't knowingly invite the general public to fill their inbox.

Back to Nolan's post, where he concludes:

For me, this means it’s harder to find good stuff worth engaging with online. And on a larger scale, systems built on easily gauging effort at a glance are falling apart.

I think there's more good stuff worth engaging with online than there ever has been. Good stuff that people worked hard to create¹. I don't think that content is harder to find than it was before, I just think you're not going to find it in on social media platforms where the algorithms don't care about effort.

Social media platforms are enormous systems full of mixed effort content. Gauging that effort at a glance has always been necessary to find the good stuff. Social media will probably never fall apart, but I do think the value proposition of social media has a crack in the foundation. Fortunately, I think the path to (re)discovering the good stuff is just a hyperlink away, via the destination internet.

even if the good stuff is made with the help of AI, as Nolan concedes, "Making a good thing still requires making lots of decisions, and making lots of decisions is still very hard, despite LLMs, so making good things is still very hard." ↩︎

Change Happens Quick

2026-06-04T23:36:48Z

In Monday's episode of @dithering.fm entitled "Electric Cars and Meta Subs" John Gruber made some comments about an "alternate universe where cars went electric a long time ago" that I quite enjoyed, quoted here via lightly edited transcription:

There's some other world where cars weren't gas driven for decades, and were driven by [battery technology having made advances decades earlier]. And in that alternate universe where the world was already all electric, or that was the default kind of car, cars would look very different. All vehicles would look very different than the cars we have today.

The cars we have today, and what people think cars are supposed to look like, whatever the type – sports car, family sedan, SUV – is all kind of defined by the fact that they are internal combustion engine cars, and they have drivetrains that take up a thing in the middle.

This reminded me of a satirical 2022 blog post entitled Test Drive of a Petrol Car that was written from a similar perspective. What if we lived in an alternate universe where cars were always electric and internal combustion engines were the newfangled technology?

Automakers do not sell the cars themselves, only through independent car repair shops as middlemen. It may sound like a bad omen to buy the car from a car repair shop that you want to visit as seldom as possible. But you apparently can’t buy the car directly from the manufacturer but must go through such intermediaries. The seller was very “pushy” and tried to convince us to buy the car very forcibly, but the experience is perhaps better elsewhere.

So we sat in the car and pressed the START button. The car’s gasoline engine coughed to life and started to operate. One could hear the engine’s sound and the car’s whole body vibrated as if something was broken, but the seller assured us that everything was as it should. The car actually has an electric motor and a microscopically small battery, but they are only used to start the petrol engine – the electric motor does not drive the wheels. The petrol engine then uses a tank full of gasoline, a fossil liquid, to propel the car by exploding small drops of it. It is apparently the small explosions that you hear and feel when the engine is running.

It's a humorous post that well illustrates how much human bias is built up around the familiar, and how hard it is think objectively. Understanding this innate human bias can work in your advantage when developing new products. The common refrain amongst product managers is that they are meeting people where they are. The best products give people enough of what they're used to while also introducing something genuinely new. This is a lot harder than it sounds though.

Gruber continued:

And now that the world is inexorably moving towards an electric future, [sooner rather than later], all cars are going to be electric. They're lurching in awkward steps towards the shapes and form factors that electric cars should be, while trying to be appealing to people's sensibilities that were forged by a different universe.

They're just so different at a fundamental technology level that, that there's no real way to launch something that is truly native electric.

Then he compared the ongoing shift from internal combustion engines to electric vehicles with the shift from wired to wireless ear buds.

Very similar to AirPods when they came out and everybody said they look stupid. Everybody wearing AirPods looks stupid. But you look exactly like you did with the [wired ones] everybody had been wearing since 2001, except you cut the wires off. And it's like, I don't know, it still looks stupid.

And now, of course, you look a little bit more distinctive and noticeable if you have wired earbuds. Change happens quick.

I couldn't agree more!

customElements.apply()

2025-02-12T11:00:00-08:00

`customElements.apply()`

A (naive?) approach to workaround customElements.define() "same tag name" errors.

Web components have completely changed the way I write vanilla JavaScript. It turns out classes are quite useful for encapsulation (who knew?)! Complementary approaches like HTML web components (and CSS web components) have reduced the amount of JavaScript I'm writing by about 95%. These ideas from @adactio.com and @hawkticehurst.com together with a steady stream of web component content from @gomakethings.com have revived my interest in modern web development.

But recently I started running into a little speed bump. I've been exploring HTML templating workflows that enable me to compose sections and entire pages from a collection of layout "partials" (ala server side includes). This sometimes causes multiple <script> tags to be added for a given web component, resulting in redundant calls to customElements.define() and a "cannot define multiple custom elements with the same tag name" error.

NotSupportedError: Cannot define multiple custom elements with the same tag name.

My initial reaction to this was to move all <script src> tags into the <head>, but this inevitably resulted in failing to load a component's JavaScript on certain pages (dependency management fail). Then I read @jakelazaroff.com's excellent Define a custom element blog post and it opened my eyes to the many different approaches to managing custom elements. But I also felt like my challenge might have been slightly different than the problem Jake was solving. So I tried something different that has been working really well for me.

It started with reviewing the CustomElementsRegistry reference documentation and discovering that get() and getName() methods already existed. All that was needed was an idempotent method for registering custom elements. It turns out this was trivial to implement.

1CustomElementRegistry.prototype.apply = function(tag, component) {
2    if (!(!!component.constructor && component.toString().substring(0,5) == "class")) { return }; // guard
3    if (this.get(tag)) { return }; // lookup by tag name
4    if (this.getName(component)) { return }; // lookup by component class
5    try { this.define(tag, component) } catch(err) {
6        console.debug(`component ${component.name} already registered as <${tag}>`);
7    };
8};

This is already working so well for me that it's just about the only JavaScript I keep in main.js for new projects. I'm replacing calls to customElements.define with customElements.apply in all of my web component JavaScript.

1customElements.apply("my-element", class MyElement extends HTMLElement {
2    // this is where the magic happens 🪄
3});

In doing so, I can add <script src='/js/my-component.js'> tags in any layout partial without worrying about "same tag name" errors.

Do you find this helpful? Or is this a horrible idea? 😅 Send me an email and let me know what you think: hello [at] calebhailey.com

The Best Chef in the World

2022-10-08T14:30:00-07:00

Originally posted on @sheesh.blog.

As soon as you have 20 minutes to spare, I encourage you to go watch this short film about Sally Schmidt, produced by Oscar® award winning documentary filmmaker Ben Proudfoot, for the @nytimes.com¹.

Speaking of restaurants, Sally started a little establishment you might have heard of called The French Laundry. 🤌

I particularly like this comment in the accompanying article:

Talking to Ms. Schmitt that morning, I learned she held a different kind of wisdom: that success may have other definitions.

Now that's what I call life-work balance!

Op-Docs is a New York Times production of "Oscar-winning series of short documentaries by independent filmmakers". This might be one of my new favorite places on the internet. ↩︎

Long time listener, first time caller

2022-04-24T20:26:18-07:00

Originally posted on @sheesh.blog.

Long time listener, first time caller

I can't remember the last time I read a book from front to back¹, but I am a voracious reader. The thing is, I prefer to read "the internet" instead of books.

At some point early on in my internet journey I discovered #RSS and I immediately fell in love². My RSS feed became like my own personal newspaper, where I was the curator. I could add and remove ~~columnists~~ blog feeds as my interest graph would change over time. I was one of the many³ who decried the death of Google Reader. RSS lives on, but it's not as ubiquitious as it once was.

For some years my feed shifted from RSS to social media — almost exclusively Twitter, and mostly via Twitter Lists — but a few trends in recent years have brought me back to RSS:

RSS services are great again⁴.
RSS apps are making a comeback!
"Doomscrolling" became a thing™️, prompting a renewed need to curate my feed.

With my newfound replacement for Google Reader's service (Feedbin.com), and my favorite old RSS reader app making a glorious (and open source!!) comeback, I decided to start the curation process from scratch. I created folders for "News", "People", "Portland" (my local happenings), "Products", "Startups", and "Tech".

And then I started subscribing. And I discovered that the very first feeds I added were the very same feeds I've been reading all along. A combination of overlapping interest graphs (I'm assuming) and writing styles have kept me reading certain blogs for over a decade now. These are the bloggers who's writing has both intimidated me – because I hold them in high esteem – and inspired me to start this blog.

I have feared that if I ever started a blog I would appear as a copy cat, a cheap imitation of these now-veterans who probably started like I am starting right now. But I decided that instead of shying away from writing, I should lean into it. Give credit where credit is due. Start my blog, in my own voice (however heavily "inspired by" it may be), and acknowledge those who went before me.

Starting a blog is easy. Buy a domain ✅ write an /about page ✅ write a "hello, world" ✅ smash the publish button. 💥 But for me, I could not have started this blog without acknowledging who this blog was inspired by.

✌️

I mean, I do remember the last book I read, but it's embarrasing. 😅 ↩︎
FUN FACT: fast forward many years later and I somehow had the privilege of working with Matt Shobe – one of the co-founders of Feedburner (acquired by Google in 2007). And by "working with" I mean he became an angel investor in my first company, and a critical advisor in our early stages. ↩︎
Or few, I guess. Otherwise we would still have Google Reader? ↩︎
In the years following the death of Google Reader (RIP), I tried a handful of apps and services that built replacements for Google Reader — from Feedly to Flipboard (powered by a carefully curated Twitter list) – but nothing stuck. Then a few years ago I tried Feedbin, and subsequently discovered its support for Newsletters (at a time when it seemed like everyone was starting a SubStack), and I almost forgot that Google Reader ever existed. ↩︎

Hello, world

2022-04-23T18:15:55-07:00

Originally posted on @sheesh.blog.

Every new blog should start with an obligatory "hello, world" — so here I am doing just that.

"I’ve started seven different blogs.
Every one had one post."

My experience with blogging has generally rhymed with the experience a former co-worker of mine had who said "I've started seven different blogs. Every one had one post." – Derek Newsom.

Something feels different this time, but only time will tell.

✌️