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 ad spending is dominated by a few VC-funded advertisers like BetterHelp and Squarespace. Dave says VC money isn't dependable; it can evaporate with slight economic disruption. If one advertiser making up 15 percent of the market goes under, that money is gone. Hearing the same Squarespace ad today as eight years ago is another sign the market is immature and unstable.
what I see is a basically a table with one with with one leg that's almost broken in half. If you have that VC VC money is not dependable. evaporate with the slightest amount of economic, you know, disruption.
Developers who build for the love of it, like Marco, Xavier, Martin, Mitch and John Spurlock, are motivated most by creating something cool and hearing people talk about the great experience. It isn't selfishness or narcissism. To rally them around a common cause, let them build something that makes other people happy and gets talked about.
if you want to get developers to all join in and on a common cause, the way you do it, is by letting them letting them have that experience by joining in, in developing something that makes other people happy.
Building Podcast Wallet, Dave found a threshold. You keep making onboarding simpler until you hit reality: a monumental effort buys an extra 5 percent ease of use, and the result is a user who doesn't understand what they've done, which opens a new set of support problems.
we can spend a monumental amount of effort to get an extra 5% ease of use here. But once you get that, then what you what your end, what you end up with, is that you've you've on boarded a user who doesn't understand what they've just done.
Asked about NFTs for podcasts, Dave says the complexity reminds him of DeFi: ten ways to crash and burn. Streaming value straight to an address is simple, with a limited set of failure scenarios. DeFi and NFT constructs open up a hundred pathways to failure, from fraud to technical problems. Simplicity puts a moat around itself.
The the whole the construct of things like defy and NFT open up 100 different pathways of failure, where from fraud, you know, all the way down to technical problems.
Dave says people should understand that the Apple lookup API, which the open ecosystem relies on for feed URLs, was never benevolence. The Apple Podcasts app aggregates feeds itself and needed an open way to call home. Because the API wasn't built for others, it can be switched off, like the episode-level API that went down and never came back.
the apple lookup API only exists for one. It exists as an accident.
Only about a third of the 2 million known podcasts still publish. Dave says polling them all equally wastes time, money and bandwidth. The Index's aggregators back off when a feed stops posting. If a show that's been silent for six months takes a day to pick up a new episode, that's fine; checking ramps back up once it's active.
it's a complete waste of time, money, bandwidth, everything to be pulling 2 million feeds, when only a third of them are actually posting content.
Planning phase three of the namespace, Dave wanted a batch that balanced immediately useful things, like trailers (requested by a host), with more aspirational ones, like recommendations and location in chapters.
I like there to be some sort of balance between things that we can use right now. And things that are more aspirational, looking down the road, Moon shots, whatever you want to call them.
Laying out plans to split the Index database dump into hashed chunks for distributed aggregators, Dave admits he's wary of going too fast. Decentralization has to evolve in pieces, because nobody will sign on to 'let's move podcasting to the blockchain' all at once, even if that's where it ends up.
it has to evolve in in pieces, because you don't if you nobody's going to jump on board with everybody, let's move podcasting to the blockchain.
Each podcast app has its own personality: Podverse's clips, Podfriend's ease of use and feature completeness, Hypercatcher's chapters and new UI. You can see by looking at an app what its developer loved first. Together they each hold a piece of the ecosystem, and open source lets everyone contribute to something greater than the whole.
each application has a personality of its own, where you can tell what the difference is, you can tell by looking at it, what the developers first love was.
When an author of a book or article quotes or praises other books, Dave writes them down to buy or research later. He recently bought a book Glenn Greenwald mentioned in an article. This habit is what the podcast recommendations spec brings to shows: the host pointing to other podcasts, books and videos.
One of the things I always do and I'm reading a book is, as the or an article is, as the author is, is quoting other people from other books, um, I'm always writing it down,
Dave loves that open source lets a project outgrow what its creators could do alone. Otherwise he'd have to do everything, which he can't, and there are things he doesn't know. For example, a documentation specialist took it on himself to redesign the Index's docs.
open source allows you to grow beyond your own capacity. Yeah, that's only good for a project.
We picture price as a line from free to very expensive, but Dave thinks it's a circle where free and priceless meet. Neither has a price you can put on it, so they're hard to tell apart mentally. Social media and freemium services operate there: give you something free and you'll hand over things that are priceless. Attaching a cost helps you place a value on it.
you have free, you know, free on the left and priceless on the right. And those things are actually closer together,
Not in the audio. Guest episode; quotes come from 15- and 8-cue Dave runs. The fuller explanation (free and priceless aren't opposite ends) partly falls on Unknown-labelled cues and isn't quoted.
Dave finds it striking that so many podcast companies and apps that champion progressive causes and minority groups haven't adopted the transcript tag. It immediately helps people with disabilities and people learning a language.
The fact that this so few apps have adopted the transcript tag, which is immediately helps people with disabilities and people learning trying to learn
Dave expects a mini ice age followed by someone who brings all the 2.0 elements together into one killer piece of content. Picture a documentary with chapter art, transcripts, value, automatic audio/video switching when you drive, and locations on an interactive map.
the mini ice age, and then you have somebody brings all the elements together into a killer product.
Dave wants the 'Podcasting 2.0' badge to link to a guidance page. Adding the namespace isn't enough: publishers should review their feeds for HTTPS, mixed content and CORS problems, because web podcast players and PWAs can't work properly without them.
part of cast part of podcasting 2.0 is that web podcast players are important. And the web, the pw a guys cannot function properly. If your webs if your feed rep has mixed content problems or in court issues
Podcast ad revenue isn't even at a billion dollars. Dave notes that only about 10 percent of podcasts, or fewer, have enough audience to carry ads. So an ad-dominant model can only grow revenue by putting more ads into that same small slice of shows.
the advertising dominant model, the only, the only option it has in order to get more revenue out out of the out of this thing is to put more ads in, right,
Dave proposed changing value-block splits from percentages that must total 100 to integer shares of the total. The numbers can look much the same, but no recalculation is needed. People will make mistakes: a five-recipient value block won't add up to 100, and it's too fragile to depend on that.
people are going to screw up people are gonna say, if somebody puts a value block in with five people in it, they're gonna make a mistake, and it's not going to add up to 100.
Malcolm X never incited violence. He warned the country about what was boiling over and argued the Black community should defend itself because the government wasn't. Yet he was accused of incitement for years. Dave argues that if Twitter and Facebook had existed in 1962, moderation would have cancelled him immediately.
The the whole idea of moderation of having a moderator, somebody who will silence speech. Mm hmm. Malcolm, if Twitter and Facebook had existed in 1962, Malcolm X would have been cancelled immediately.
When Facebook blocked news in Australia, Dave called it a terrible mistake for a dominant company. People simply went back to old habits, like setting their browser homepage to a newspaper. He hasn't logged into Facebook in over a year and life is fine; what happens on social media isn't reality.
You don't ever want to show people that they can live without you. That's not a good decision for somebody who is in the market position of dominance that they are.
Dave rejects the business cliché that young people are tech-savvy. They have lots of devices and apps, but you can do a lot without knowing how anything works. In the IRQ-conflict, config.sys era, people were forced to learn pieces of the system underneath; today fewer people understand even what a browser is.
They have lots of devices. And they have lots of apps. But that is not the same thing as being tech savvy.
Dave's definition of success for value for value: a year out, around 200 podcasters running their own Lightning nodes. That's a fully funded payment network for podcasting that chat relays, media distribution and even feeds could build on. At that critical mass, hosts need to bring their easy authoring tools to the decentralized side.
let's just say you have 200 podcasters that each have a, a big lightning node. And over, you know, at that point, what you have is a fully funded payment network for podcasting.
The value block in XML is the correct, open, transparent way to say who should be paid. Turning that into API calls between app and wallet service is where it gets sketchy. If an app holding broad payment powers is compromised, it could drain everyone's wallets to an unknown address.
the danger, the danger is within is giving the app too much power.
Dave doesn't believe any network, Bitcoin included, is untouchable. Governments could block it at internet backbone borders. But blocking everything requires a globally coordinated effort, and the only truly global coordinated efforts have been world wars; at that point there are bigger problems.
if you want to get everything can be blocked, ultimately there has to be some realization of that. But then you also have to think well, that requires globe that requires a global coordinated effort.
After Spotify delisted Pete Evans for COVID misinformation, Dave searched Spotify. Evans was gone, but the New York Times' Caliphate series, which had been thoroughly debunked, was still there. 'False information' gets enforced selectively.
I type in New York Times caliphate. And it's still on there. Even though that even though that entire series has been completely debunked every it was all a lie.
Dave's core conviction: restrict even small amounts of freedom of speech or of the press and equality immediately drops. He believes the fight from here on will be preserving equality through freedom, not trading freedom away to achieve it.
people are the most equal when they're the most free. And when you if you take, if you restrict freedom, even a slight amount of freedom of speech for you know, freedom of the press, even as small quantities, immediately, equality, equality goes way down.
Sphinx's revolutionary idea was charging a tiny amount per message. It's small enough that you chat freely, but it adds anti-spam protection, and if you post something hateful and it's deleted, you lose your money. Staked, revertible payments let the community moderate itself instead of putting top-down moderation on the platform's dime.
If, if every, if the message is cost you money, and there's a way to stake it and revert it, then the community can police itself?
Developers like Marco have refused countless requests for listener stats. A private company offering money for stats feels gross. Dave thinks an open, transparent listener-behavior standard with privacy built in might win them over, because it helps the podcasting they love rather than a backroom deal.
If a private company comes to you and says, Hey, we want to pay you for stats, that feels a little gross. But then if there was an open standard that said, Okay, here is how we're going to deliver listener behavior stats.
Dave thinks privacy fears around listener stats may be overblown. Ad blockers took off less because of tracking than because ads are annoying. He blocks ads on his phone only because junk makes pages slow and unreadable. Most people wouldn't mind opt-in behavior tracking if they were assured it was safe.
the only reason I use an ad blocker on on my phone is because I keep because it makes it slow. And I can't read the page, because there's so much junk all over it.
Ad salespeople need inflated download numbers too. But if Apple's stats and an open stats standard from several apps all point to similar levels, the truth becomes hard to put back in the box. Dave admits money might drop at first. He also notes that VC exit deals like Wondery and Megaphone were driven by phony statistics.
maybe the money goes down initially. But then it ultimately though, isn't that better for everybody involved, that everything's just more accurate,
Dave warns podcasters entering the Bitcoin ecosystem that many people misunderstand it as completely unregulated. Just like ad income, Bitcoin earnings and gains are money and taxable. Check how your local regulations classify it.
there's a lot of people out there that misunderstand that Bitcoin is, is a completely unregulated financial instrument. And it's not.
Lightning wallet developers focus on a good Lightning experience and podcast app developers on a good listening experience, and the two don't mesh well. For a good podcast app, podcast developers should design it, with Lightning plumbing tucked behind, for example by linking to an external wallet like Strike through an API.
lightning wallet developers are not going to necessarily be good podcast app developers. Because podcast app developers are focused on a good podcasting experience.
On whether to label, ban or take down contested claims, Dave sides with Matt Taibbi: banning something is the quickest way to convince people there's a cover-up and that you're trying to suppress it. Opinions also shift fast, as with public views on same-sex marriage since 2008.
If you know, if you're concerned about something is being something being disinformation, banning it, is the quickest way to make people think that there's a cover up, and that you're trying to suppress it.
Move a show from one IAB-compliant host to another with the same audience and the numbers change. Each host makes its own calls within the IAB rules: some count bot downloads with an FFmpeg user agent, others exclude them, and they classify Apple, Core Media and Android traffic differently. NAT offices make many listeners look like one.
each host is making their own decisions about within the confines of the AI based ads, rules. There's lots of leeway to make those decisions.
Recent experience shows podcasting's standards bodies and working groups slow-roll and accomplish little. Dave says there's a reason: when you form a standards body, especially with a lot of money and memberships involved, gatekeeping follows naturally. The established players who sit on the boards are experts at it.
whenever you put together a standards body, and especially when there's money involved, a lot of money involved, and there's memberships and all these sorts of things. There's a natural tendency for gatekeeping
Responding to Rob Greenlee's point that the open ecosystem must keep innovating to compete with Spotify, Amazon and Apple exclusives, Dave agrees and goes further. By the end of namespace phase two, with location, person, named seasons, social links and other new tags, open podcasting would blow away any closed platform.
If you extrapolate out to the end of phase two, and you have potentially 10, new tags that give you all this extra stuff. I mean, like that blows away any closed platform,
Dave says what sets podcasting apart from other media is direct engagement between podcaster and audience. A popular morning radio show might build a tribe, but listeners have limited ways to interact back. Podcasting's interactivity has a depth that doesn't exist elsewhere.
that is the strength of podcasting. And one of the differentiators of podcasting from other forms of media is the direct engagement between the podcast and the podcaster. And the audience.
Dave thanks Fireside for being the first host to auto-submit new feeds to the Index. He won't routinely scrape Apple to find new feeds; that would contradict the goal of not depending on Apple. Independent streams from hosts, republished through a clearinghouse endpoint, build a decentralized index where Apple could someday be just another peer.
I don't want to scrape Apple to try to find new feeds on a contract basis. I just feel like this ideologically opposed to the thing we're trying to do, which is not dependent on them.
Why define a chapters spec when MP3s can already embed chapters? Dave says a separate JSON file lets hosts add chapters without heavy audio work: no re-encoding, no parsing files or writing frame headers.
you don't have to re encode or parse the file and stick them in the frame headers and all this kind of stuff. I mean, it's just so much easier to keep everything in a separate JSON
Talking with Fireside's Dan Benjamin, Dave describes the hosting landscape. The mega-hosts are on autopilot much of the time and don't feel a need to innovate. Smaller, hungry hosts can use new features to differentiate themselves, help their customers and move podcasting forward.
you have these mega large hosting companies, who they're just, you know, a lot of that they're on autopilot most of the time.
Dave built a people search that scraped episode descriptions. James Cridland pointed out that it could slow adoption of the person tag: if search already works, publishers don't feel they need to add the tags. Dave agreed and left it as an interim tool.
it gives people a reason, if you're doing these searches already, it gives people reason to, to not feel like they have to do it.