This page lists part of the collection. The page links below work without JavaScript. With JavaScript, a filter searches everything and pages the matches.
Podcast content can carry ads, but you can't drop an ad in the middle of a three-minute song. In music, ads only fund the platform, so for directly benefiting the musician value-for-value is the only model that works.
you can't drop an ad in the middle of a three minute song. There's no there's no possibility to even do an ad based music model that did go and I'm talking about in a way that directly contribute directly benefits the musician.
The feedback loop and the feeling of connection between listener, host and guest are what make podcasting special. A well-produced Fresh Air episode felt like reading a Wikipedia article: informative, but not compelling enough to subscribe.
The the feedback, the feedback loop within a podcast is it changes everything. And it's, it makes all the difference.
You can hear the difference between podcasters doing it from passion and those who reverse-engineered a monthly income target. Only the first is compelling.
you can tell when people are doing it, because they have a passion. You can tell when people are doing it, because they've, they've put together a math problem where they can end exit the end of the month with certain X number of dollars, right? in those in those things are just not compelling to me at all.
His dad, a church deacon, taught that you choose deacons from people already doing a deacon's work. Dave applies it to passion projects: the best contributors were doing it before any reward existed.
my dad said a long time ago he was a deacon in the church and he was saying that when you're looking around for people to be deacons. What you do is you you look for the people who are already doing the work of the deacon. And those are the people you nominate to be a deacon.
Writing code is only half the job; the other half is meetings and evangelism, which is hard and exhausting, so Dave tends to do too little of it. Going fast means things get lost, so sometimes stop building, take a breather, and explain how things fit together.
You have to be a coder and an evangelist. And the evangelism part is hard and exhausting. And so that I tend to not do it as much as I should.
Apple published the iTunes tags to control how feeds behave inside Apple's ecosystem and never promised anything beyond that. The industry adopted it as a de facto standard, causing conflicts like the iTunes block tag being honored by some and ignored by others. The only fix is a truly open namespace, and you shouldn't answer private APIs with more private APIs.
you should think of the items namespace not as a namespace not as an open spec, but as an A API. The iTunes namespace is the API that you use to control
Dave changed index behavior out of anger and was criticized for it. Setting the rhetoric aside, he called the criticism fair: apps depend on the index, so decisions that affect others belong to the group, not to him alone.
But But what if that's, that's fine. But rhetoric rhetoric aside, it was a fair criticism. I mean, like I, I, I get angry and I just in a change something and I should have I should have talked it out first
Citing Dave Winer: once people depend on a feature you shouldn't remove it. Shims and stopgaps are dangerous for that reason. Have a theory of when to shim, start manually if volume is a trickle, and if you do shim, say up front that it's temporary.
Dave, Dave Winer has as good advice here, you know, he always said, he always said you never take a feature away. Right? You're you always add, you don't want it, but you got to be careful what you add
Dave hates Macs, but deliberately uses one at home because new hires who grew up on Macs connect to work from them, and he needs the knowledge to support them.
So I'm like okay, we we need to be able to troubleshoot Mac so I'd use the I made the decision I'm like I'm gonna force myself to use a Mac at home so I can have the knowledge to help these people
Feed verification got bogged down because people called it 'authorization' and reached for OAuth. It's really just proof of feed ownership. Reframed that way, the simple txt-tag approach that includes self-hosted and WordPress feeds wins, and you don't build a second mechanism on top.
we're referring to it as authorization, when it really is something more simple, which is just pure feed ownership verification. And if you think about it, in those terms, the the problem gets a little bit less complex.
The W3C works because a few giants control the one or two rendering engines everything filters into. Podcast standards must filter through dozens of hosts, apps and platforms, and developers bristle at being told what to do, so a top-down standards body is nearly impossible.
The podcast industry is so much more scattered than that. It's like herding cats. It's basically impossible to get any and some people I think, also, this is clear that this happened when the podcast industry reached out to certain individuals
When Pocket Casts raised its premium tier to about $40 a year, Dave welcomed it. People pay $100+ a year for streaming services, and podcast listening hours are at least comparable. Raising the perceived value of podcast apps lifts all boats.
I just appreciate that they're raising the bar on what on what on the what the value of a podcast app in general is I think that'll lift all the boats a little bit.
In a free market the consumer decides what they can bear. Letting listeners pay what they want feels better to them than lock-in, and people hate paywalls.
what matters in a free market is the consumer and what they can bear. Do the best thing you can do is give your give your audience the ability to pay you what they feel. Because they feel better about that arrangement.
The same advertiser runs on five shows with five codes. When you finally sign up, you use whichever code ranks when you search, so the code reflects SEO, not which show actually made you aware.
And it has nothing to do with really the show. It's just got to do with your awareness of the show. And then what how much Google juice has
Asked what he really wants in the index, Dave boiled it down to 'if somebody searches for a thing, they can find it,' regardless of content. Then he reverse-engineers that goal into concrete criteria, because with too much junk you can't find the junk or the treasure.
There has to be something where you can sort of take the philosophy of that idea and convert it into an answer of something that you can take action on.
When a friend wanted Spotify to 'fix' RFK Jr.'s voice (spasmodic dysphonia), Dave disagreed: because he's difficult to listen to, you have to pay very close attention to understand him.
No, see, this is this is part of his superpower. Correct? He's, he's difficult to listen to. So you have to you have to pay very close attention to understand what he's saying.
On Spotify and Dawn Ostroff's Hollywood strategy: Dave has seen companies merge that look like a perfect match on paper, but one culture irritates the other. They never fully integrate, and the thing falls apart.
I've seen this in my industry. With a few really good examples of two companies that merge and own paper, it looks like a perfect match. But you get in there and the the culture of one does is like really irritating to the culture of the other one.
Podcast hosting companies compete ruthlessly yet work together on open standards, and Dave can't think of another industry like it. They understand that a rising tide lifts all boats.
They they all understand that. That a rising tide lifts all boats. Yeah, it's a great thing to see.
Ad money is moving to YouTube, which is why the industry insists on calling YouTube videos podcasts. Its 'podcast' ad products depend on location, weather and driving data that only closed platforms provide, and on IP-address data brokers that privacy features like Private Relay could wipe out.
the podcast industrial complex, is not podcasting. It's not what you think it is. Chris said that all the ad market is moving to YouTube. And it also dawned on me that that's why they are so insistent about calling YouTube podcasts?
From an expert witness's testimony in the Bitcoin Fog case: proving a chain of custody from UTXO to UTXO, and linking any of them to an actual person, is very hard. Stolen keys make it worse, and the 'totally transparent' claim is mostly fiction.
But it's actually really difficult to prove a chain of custody from UTX o to UTX. Oh, and that, especially when you think about key, you know, keys, getting stolen all these kinds of things. You really cannot there's no linkage between an actual person and these UTX owes.
Not in the audio. Dave relays the reasoning of a guest on the Monero Talk podcast (named in the transcript as 'Francisco Cabana'); spelling and details of the case are uncertain, and Dave himself mixed it up with Tornado Cash on air.
Apple charges to use its platform under its own terms, like Reddit's API. It's even worse because the rules are ideological rather than monetary: you must obey the guidelines (e.g. moderation requirements) or be kicked off. You don't own your iPhone.
Apple is Reddit, they charge us to use their API according to their terms of service. Just because it just happens. It just so happens to be on our own piece of hardware. Right, that we still don't control it. They do.
The free-money era of big tech is ending. Reddit and Twitter raising API prices and cutting staff are what happens when companies must turn a profit, and it will spread across tech. Twitter went from 8,000 to 1,500 employees and little changed except getting closer to profitability.
When you actually have to make money, everything changes. And so Reddit, they don't have a choice, they have got to make money, and they are going in they're going to do it. And you know what, it's going to be just fine.
If Apple killed value-for-value buttons in iOS apps, it would be a setback, not the end. Focus on the problem you're solving and there are a hundred workarounds; for example, a thin progressive web app on top of the content puts you back in business.
when you focus on, what is the problem that you're trying to solve? There's 100 ways around it.
Onboarding friction stops people signing up at all. The second kind builds slowly: rising ad load, monetization plays, a platform changing under you, less compelling content. People just float away and don't come back, which is why he let Netflix lapse after months of not opening it.
You have like the initial friction of on. And I mean, with an app or a service or anything, you have initial onboarding friction, that can turn users away from ever signing up in the first place. That's one kind. There's another kind of friction, which is a which is a slow build up
Dave's overall media consumption is down. His family dropped Netflix after going four months without opening the app.
we dropped our Netflix subscription. And we didn't drop it. The reason we dropped it is because we went like four months and didn't even didn't even open the app, because because it's almost like they just simply ran out of compelling content.
Dave says he 'loves' the Apple Vision Pro because he thinks it will be remembered as the poster child for when things began to fall apart.
I love this product. Because in my opinion, this is the this will be in the future. This will be looked back upon as the poster child for when everything began to fall apart.
From years of help desk work, Dave knows a spec only makes sense to people who already know what it does, so every spec needs an explainer.
I'm reading this and I'm like, no, no, nobody, no human is going to understand this unless you already know what it does. Right. And so I think I even wrote in the spec. This thing needs, there's going to be an explainer for it
Dave subscribes to the 'rewrite it three times' approach: the first version you'll hate, the second teaches you what you should have done, and the third time is the charm.
use subscribe to the rewrite it three times methodology. You know, the third time is, like, right at once you're gonna hate it right? It's right at the second time, realize, you know, man, if I'd only done this, and the third time's the charm.
Dave would rather build Adam and volunteers a simple dashboard than keep SSHing in to run cleanup scripts. It doesn't have to be beautiful, just functional, and people who want to help need tools to help with.
And I don't think this is hard to write because it doesn't have to be beautiful. It just has to be functional.
Having lived through hype cycles (blockchain, ML, AI keynotes at tax conferences), Dave says the reality is mundane: new technologies spread by being built into existing tools, like Buzzsprout auto-generating chapters for people who wouldn't otherwise bother.
the reality of the way that things like this new technologies come into widespread use within industry, is they come in through by being baked into the tools you already use.
Each Podcast Index aggregator starts with a big batch, backs off by 100 feeds whenever the parser crashes, and after 48 stable hours tries to move up again. Each one finds its own 'happy place' and runs at maximum utilization.
it just self adjust as it goes until it finds each until it finds the happy place for each aggregator.
From 27+ years of end-user support: (1) all software sucks and is full of bugs (even Amazon's desktop site has a mobile-gesture bug), (2) users don't tell you the truth, (3) nobody reads your emails. A bonus fourth: if they're on Wi-Fi, that's at least 50% of the problem.
The three pillars of helpdesk are all software sucks, and is full of bugs. That's number one. Number two, users don't tell you the truth. Yep, total number three, nobody reads your emails. Those are the three pillars.
Dave credits Adam's 25-year habit of being kind to sysadmins and learning how their systems work: when you know how their stuff works, you understand their frustrations and can have empathy.
oh yeah yeah. So you you need to know how their stuff works so that you can understand their their frustrate you can empathize so you can have empathy.
On a contentious question (stopping auto-ingestion of Anchor feeds), Dave would rather let people form and voice opinions for a while than make changes now. He uses the same approach for proposed tags, which he gives time to 'bake'.
I don't want to do anything right now, I want to let this bake. And I want to let people have an opinion, you know, form opinions on it and continue to talk about it a little bit.
Dave's take on throwaway Anchor feeds: much uploaded audio will never get a single listener, so an index has to be choosy about what it keeps.
Nobody is more than just nobody cares. It's nobody in the history from now until the end of time. No human being will ever listen to that piece of audio content.
An XML attribute is one-and-done: one key, one value, and anything more needs delimiters. If a feature will need to expand, make it its own tag so it stays extensible.
The problem with attributes is they're just a one and done. You get it's a key value, you get it you get one value, right? And if you try to stuff something else in there, you have to use delimiters.
On YouTube ingesting RSS feeds (clean episodes, no pass-through, no stats): people were angry when a host harvested emails from feeds, yet would let YouTube sit between them and their customers. Given Google's graveyard of abandoned projects, be very cautious, and don't interleave podcasts with clickbait feeds.
this would be as if a cast became instead of just soliciting customers they became a proxy be between you and your customer. And but don't worry about it we're not we're not going to charge anything.
Because the comments are built on open protocols (ActivityPub), the same threaded comments under an episode show up in every app that supports them. Compare that with comments split across YouTube, Spotify, Apple reviews and Twitter, which podcasters must manage or hire someone to manage.
do we do we seriously want to live in a world where there's 10, different siloed batches of content, where all of our stuff lives and in a podcaster is going to have to either manage that themselves and jump back and forth between all these different platforms.
Nostr and similar tools can enhance podcasting in many ways without replacing RSS. Plug into RSS as it is, because RSS is honestly impossible to replace.
Like if you if you just integrate with RSS as it is, you can still do like, I think you can still do most of what you're wanting to do. It's just the the RSS is going to be just honestly impossible to replace.