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 apps should always download an episode before playing it, as Overcast went back to doing. Downloading ahead for later playback, whether three seconds or three days later, is what podcasting was meant to be. Streaming crept in years later and made things messy, which is why downloads are still the metric podcasting is measured by.
podcast apps should always download the episode before playing it. I think what Marco did on the recent release of overcast is exactly the correct right thing to do.
Many podcast apps are run by a single developer who handles support email, code and marketing alone. Dave imagines the day you can offload some of that to a second person must be a joy.
When you make the move from one to two people, it feels like that's probably I'm just imagining that's such a relief to be able to split some of the workload,
V4V isn't an alien model. It's how restaurants already work in the US: if nobody tipped, restaurants would go out of business. The tip is the gravy on top that keeps the place existing, and podcasters have to learn to remind people of that.
Like when you go to a restaurant, if you if nobody ever tipped at a restaurant, if nobody ever gave a tip, and everybody just paid what was on the on the on the receipt, the restaurant would go out of business.
Software vendors blanket-disclaim all responsibility in their EULAs, but just because you write it into a licence agreement doesn't make it true. It still has to be legal, and that has yet to be seriously challenged in court.
just because you write it into an end user license agreement doesn't mean that that that's true. I mean, it has still has to be legal.
Developers want to build things, but there's always a tension with reality: is it realistic, what's step two, and how does it connect to what we already do? Those questions temper the excitement, and that's healthy.
there's always that tension between when you're when you're, when you're a developer or builder coder, you want to build things. But then there's a tension with is this. Is it realistic? Oh, like, what? Step two?
Live is one of the most important parts of Podcasting 2.0, even if few people use it yet, because it shows RSS can be as powerful a distribution mechanism as YouTube. It puts RSS on par with every proprietary platform.
to me live is one of the most important things about podcasting 2.0 Even if not a whole lot of people use it right now. Because it it shows that RSS can be the can be just as powerful the distribution mechanism as YouTube or anything else.
Testing new things is valuable because once people actually try them, they realize it's not as scary as they thought. Dave was scared to do a live show until Adam pushed him. It turned out not to be that bad, and after a couple of times you learn not to be bothered by mistakes.
when people do it, they realize it's not as scary maybe as they is they think, yeah, I was scared. I was scared to do a live show until until Adam, you know, you brought it up and it's really not that bad.
For a podcasting e-cash mint, Dave would rather not ask people to trust just him and Adam (people trust them almost too much). He'd want the mint held in trust by many stakeholders, such as hosting companies and apps, because 'trust but verify' means other people are involved in the verifying.
I think they do that people do trust us, but it's almost too much. I don't want to expect that out of people. You know, I'd rather have somebody a few other people that are involved
Any number that stays static (CPMs, wages) is effectively going down with inflation. Dave's grown kids each work two jobs because $14 an hour is no longer a livable wage.
any any of these numbers that stay static like this, you're you're losing money.
After years of open source work, you build everything so anyone can run it themselves. But people often just use your hosted instance instead, as with the Podcast Index ActivityPub bridge. Dave doesn't want the Index to become the thing everybody depends on, but you can't force others to run their own. It has to happen organically.
I don't want to become the thing that everybody depends on for the for for data, you know, but but there's that it makes, you know, but you can't force that. I mean, it, it just, it has to be an organic thing.
Dave admits he was ready to give up after learning that Lightning wallet changes meant re-architecting a completely different V4V solution. Frustration kills momentum, which is why bridging such gaps quickly matters.
I was ready to give up this past week. I mean, I was really serious, seriously, I mean, that just the idea of having to go back and re architect and completely different solution. And all this stuff was right. I mean, it, it's it's frustrating. You know, it's frustrating is a momentum killer.
The tech industry sells futurism: the claim that a technology is inevitable, world-changing, and has emergent powers no one can stop. Recognize the pattern.
this is all about the tech industry is all about is futurism. It's all about the myth of inevitability.
Podcasting 2.0 can't build everything. Fountain and Wavlake have built self-contained islands that work, but that isn't healthy, even for them. If Fountain becomes the only place V4V happens, Fountain will be less successful. Everyone benefits from a broad ecosystem of independent Lightning and podcasting infrastructure.
if fountain becomes the only place this stuff is happening, fountain will be less successful. That's not going to help them either.
When V4V infrastructure gets messy (Alby leaving custodial wallets), go back to why it started. Programmable money allows tiny payments without crazy fees, but the bigger reason is a direct, unblockable relationship from listener to podcaster with zero intermediaries, so that if advertisers bail over a controversial stance, you still have an income stream.
if you go back to the beginning, and you look at why we did this to begin with, it was for two reasons. Because the progress programmable money allows you to send people small, micro small amounts of money A without the exhort without the crazy fees.
If Podcast Index funding ever dried up, Dave would cut everything to the bone. Core services can run for about $300 a month, which he could pay out of his own wallet, so he isn't worried. The pressure to keep a company from dying is where all the bad decisions start.
the day that that does happen, I cut everything down to the bone. We can run, we can run just we can run core services for about $300 a month.
Not in the audio. The 'survive as a company is where bad stuff starts' point is the follow-on turn at 01:34:20. The cost figure is Dave's 2024 estimate.
After his daughter's new band had a rough gig, Dave told her that's part of it: she'd only been with the band a couple of months, and this is how you learn, by going out there and failing. Giving up control is part of what makes playing live exciting.
you're playing with a band, you've only been playing for the band for a couple months. This is part of this is how you learn by failing.
The ad model is only a winning strategy if you're essentially a monopoly like Google or Facebook. Podcasting is too diverse for anyone to dominate. Even someone who listens to three hours of Rogan a week has 17 other subscriptions. So advertising will never be what supports podcasters' livelihoods.
podcasting is by definition, anti monopoly because it is so diverse. Nobody will ever dominate. In this in this arena. Even people who have large audiences, there still only one show in a person's podcast app.
Talking about Apple's crushing ad and a widespread tech fatigue, Dave says technology hasn't been the great liberator it promised to be. In many ways it has become a set of shackles.
technology is not it's not. It has not been the great liberator. Now, in a lot of sense, it has been a set of shackles.
For Podcasting 2.0 to become the trusted home for podcast standards, maybe even one day stewarding what the iTunes namespace does, it has to stick around long enough for people to see it's a trusted community that will carry things forward.
we we have to have staying power is gonna, you know, we got to stick around long enough for for people to see that. That podcasting. 2.0 is the right place to house something like this is a trusted in is a trusted community
When an iTunes tag isn't good enough (block, or NPR wanting named seasons), you can beg Apple for ten years with near-zero odds, or duplicate the tag, add the feature, and keep it backwards compatible. If there's no real improvement to offer, don't duplicate it. Do this enough and the podcast namespace could one day become a drop-in replacement.
we could have begged Apple to do it for 10 years. And just hoped that one day they listened and did it. Yeah. Or we could just duplicate that tag, make sure it's backwards compatible with the iTunes version, and then modify it ourselves while maintaining the backwards compatibility, which is what we did.
Anything critical to how podcasting works should change through communities of people, not private companies meeting in private boardrooms. Apple controls the iTunes namespace and nobody else gets a seat at the table. The best anyone can do is beg.
If there's something at this critical to podcasting, infrastructure or technology, we as podcasting 2.0. Want one Shane changes to that thing to be done by communities of people, not private companies, meeting in private boardrooms.
People who built RSS software through Google Reader's rise and fall saw Google pull everyone into its ecosystem, kill competitors over years, then rug-pull everyone. It nearly killed RSS. That experience means you should get very nervous when a big company gains a dominant position over something that used to be open, as YouTube now threatens with podcasting.
it makes you not trust Google in any way. Well, when they start to get it when they start to get a dominant position over and open over something that used to be open. You have to get very nervous.
The Kindle was supposed to kill books. Instead, worn-out Kindles went in drawers and paper bookstores came back. Hype about a medium dying usually overshoots.
as most people's Kindles, right, get get worn down, as the battery starts having problems or the screen gets wonky. They just toss them in the trash, and they don't buy a new one.
Australian guitarist Alan Gogoll had much of his music removed from Apple Music over unproven fake-stream accusations, with no one to call and half his income gone. Dave says this is why open, V4V, RSS-based distribution matters for real people making real money, and why creators need a plan B.
This is potentially important for for real people in the real world making real money shall we? I think having a plan B you know,
ActivityPub is much harder than RSS: an RSS feed is one file, while ActivityPub needs a server that speaks and listens. But Podcasting 2.0 went from zero to pushing Bitcoin around in real time, so it can do hard stuff. Dave isn't scared of doing hard things.
went from zero to pushing Bitcoin around in real time. Yeah. So that we can do hard stuff. You know, that's, that's not it's, it's not I'm not scared of that. I'm not scared of doing hard things.
With RSS, all you need is one file. An ActivityPub relationship with the world needs a server that can both speak and listen. RSS made sense when servers and bandwidth were expensive, and it still does.
if I have an RSS feed, all I need is I need just one file. If I have an activity pub relationship to the world, I need Need a server? Yeah. Yes. And that can Can I need a server that can that can speak and listen both.
Dave's read of NPR: it tried to diversify its audience by changing what it made, alienated the audience it had, and the new audience never showed up. You can't make people care, and that's a recipe for bad times.
Your old audience feels alienated and now you're new, the new audience never showed up. And so that's like a recipe for bad times.
Changing your audience is fine, but ask whether the new audience wants you. Otherwise the old audience feels alienated and the new one never shows up.
changing your audience is fine. If if what you want at the end of the day is a completely different audience. That's fine. There's nothing wrong with that necessarily. But you have to stop and ask if that audience if that different audience wants you.
Looking at Spotify's podcaster numbers, Dave concludes that when you give people a chance to just pay for something, they do, and you make more money. The ad route is selling yourself short.
when you give somebody a chance to just flat out pay for something they do it just make you make more money. Yeah. The ad and the ad route is just selling yourself short.
If everybody is chasing video, the contrarian move may be even less media: text. Dave thinks the best content right now is probably written, and he misses a lot of good writing if he gets sucked entirely into podcasts.
the best content right now? Probably is written. Sure. I mean, there's a there's a lot of people that are writing a lot.
With streaming value-for-value-style ad payments, advertisers pay only for what's actually listened to. They aren't blindly throwing money and waiting for a spreadsheet that says maybe it reached people, so they could spend less to reach more.
You're not throwing money, you're not blindly throwing money and then waiting for a spreadsheet to say that maybe this thing got in front of people, you're not sure. I mean, like, it seems like you would spend less money to reach more people.
In sysadmin work, things are either unnoticeable or very noticeable, with no in-between. A maintenance task almost never takes 10 or 15 minutes. It's either 30 seconds or four hours.
They think things are either either unnoticeable or very noticeable. There's no in between. Correct. Correct. Things don't usually take like 10 minutes. 15 minutes. It's either 30 seconds or it's four hours.
Dave's mental framework for ActivityPub in podcasting: imagine what the world would look like if the thing you're talking about were itself an actor, such as the episode, the listener, the clip or the transcript. Taking the smallest node in your idea and making it an actor is where things get weird and beautiful.
consider what the world would look like? If the thing you're talking about was itself an actor? What if the episode was an actor would if the US listener was an actor with a podcaster?
A value-for-value show can welcome advertisers as participants without becoming 'ad supported'. You stay in control of your content and don't change or censor it. If a business wants to support you as part of value for value, that's welcome.
I'm still in control of my content. And if you want to participate in value for value as an advertiser, you're free to you're welcome to do so.
Dave compares goofy protocol features to 'chair pants', trousers with a built-in frame so you can sit anywhere: do they function? Yes. Are they a good idea? No. He could invent ten features for the ActivityPub bridge in a minute, but he doesn't want a bunch of 'chair pants' that nobody cares about running around Podcasting 2.0.
It's It's like, it's a does it? Does it function? Yes. Is it a good idea? No, it's dumb.
Offloading features onto a third-party service is easy, but you end up with a single point of failure and a lot of dependency. For ActivityPub and podcasting to work, everyone needs to participate fully from the start. Helper services are fine, but they shouldn't become required.
that's kind of a crutch, okay? Because it's easy to just sort of offload some of these things on to other services. And then you end up with you know, with a single point of failure, and a lot of depend a lot of dependency.
The beauty of writing a spec is knowing up front that nobody will do half of what it says. If you write with that in mind, you know what to leave out and what not to get hung up about.
That's the beauty and beautiful thing about writing a spec, you already know that nobody's going to do half of what you say in the spec. So you write if you if you write that, if you write the spec with that in mind, then you know what to leave out and what not to be hung up about.