This page lists part of the collection. The page links below work without JavaScript. With JavaScript, a filter searches everything and pages the matches.
Fans who really love a show don't want the podcaster barely scraping by every episode; they want to see that it's worth the creator's time and financially rewarding. Being open about support numbers helps listeners see that.
if you if you really love a podcast, you want you don't want the podcaster to be struggling and barely scraping by every every episode. You want them to see that it is worth their time. And it's a rewarding experience financially
The best way to develop a feature is when you want it yourself, because then you'll build it well. Running a host or making a podcast also teaches developers things they'd never otherwise learn.
That's the best way to develop any features when you the one there when when you want it for yourself, because they know build it? Well.
Dave predicts the companies that make it through 'podcast winter' will be small teams (Blubrry, RSS.com, Buzzsprout, RedCircle size and smaller), not the big bloated players.
I think the companies that make it through podcast winter are going to be small teams.
Free trials are a huge spam problem for podcast hosts (rss.com and Buzzsprout have struggled with it). Dave asked a new host whether requiring a credit card at sign-up was meant to keep spammers out; the host said yes.
So because that's a huge problem is such a huge problem with free trials. You know, that, you know, I know that rss.com and Buzzsprout. And I know that they've, they've really just struggled
The Podcast Index's boilerplate terms of service accidentally banned building databases off its database, which is the opposite of its mission. Dave wants terms that are a clear list of don'ts, not a 12-page monstrosity nobody understands.
I really wish that we could have a terms of service that was just like, you know, don't do this. Don't do this. Don't do this. But basically just a clear, it's not like some 12 Page monstrosity, that nobody knows what's in it.
If someone puts money into promoting the value-for-value ecosystem, don't spend it on Podcasting 2.0 itself; market the creators who are already there making content people want to hear.
market market the people that are already there that are producing the content people want to hear.
The ActivityPub bridge's notification thread died because Dave force-unwrapped values in Rust to skip error-checking logic until 'later'. The fix: properly check every unwrap, and watch threads and respawn them when they crash.
I should be properly checking all unwraps which I'm almost certainly not doing as as Alex likes to say unwrapping things for Christmas. And, and also checking to see if the thread crashes and bringing it back to life spawning a new thread.
Some developers are addicted to interoperability: the thrill of getting two computers to talk for the first time. That addiction to interop is what created the internet, and it's why people keep building open software even when it doesn't pay.
that that was being addicted to interrup was what created the internet it's It's why people like Dave Weiner got so addicted to to internet technology is because it's it's all it's just all an expression of of interop and it's so much fun
After a week stuck on ActivityPub HTTP signatures, Dave admitted he'd looked at the code too long to see his own mistakes, and decided to post it publicly so someone else could spot the missing newline or bracket.
I'm at the point where I've been looking at it for too long. And I can't see my own mistakes at this point
Industry surveys use forced-choice questions instead of open-ended ones not because it's more accurate but because the alternative is hard to analyze. Keep that in mind when 'the research' is quoted at you.
that's because it's hard. That's why they didn't do it. Because it's hard. And it makes it really difficult. Yeah. So when you when you when you're talking about surveys, in research and research from these from these agencies,
With some collaborators (Dave's example is Todd Cochran on alternate enclosure), discussion is a negotiation: 'no' doesn't really mean no; it means they aren't ready yet, and you need to change tactics.
when he says no, that doesn't really mean no. it mean it means I'm just not ready yet. Like you can use it means you need to change tactics.
Many people miss parts of value for value. When everyone involved (even the lighting designer) can share in the revenue through splits, it's no longer people helping just out of kindness, and that changes the game.
when you can have the split where everybody can share in the in in the revenue. It changes the game. Now it's no now it's no longer people just doing it simply out of the out of the kindness of their
Earlier podcast namespaces failed by tying themselves to one protocol. Tags should accept any protocol so you can rip and replace when the next big thing arrives.
we try first and foremost to make the tag except future proof. So they accept any protocol. Yeah. The problem with namespaces in the past with with with podcasting is they, they they tied themselves to a specific protocol
A PWA with no server back end can't hold keys or secrets, because everything ships to the browser and is visible. Plan a back end for anything that needs secrets (like wallet SDK keys).
if you're a web only app, and you don't have a server back end, you you can't, you've really essentially can't have any secrets. No.
Podcasters instinctively avoid shady sponsorship schemes because it would be a one-shot, take-the-money-and-run deal, and then your audience hates you.
we sort of inherently know in the back of our mind that if we did that, it would be a one shot deal. And then your audience hates you. Yeah. Take the Money and Run option.
Advertising has a permanent baseline of brand-awareness spending that never goes away and only fluctuates a bit, separate from specific push campaigns.
there's that back sort of background radiation level of brand awareness advertising, where it's just like, Okay, we, we just got to do this always.
When interest rates rise, US Treasuries paying 5% become more attractive than risky bets, so investors pull capital out of speculative ventures, including money-losing podcast and tech companies.
when interest rates go up. Basically, US Treasuries become more attractive than risky bets, because you have you have US Treasuries paying in specifically treasuries, paying 5%.
Dave's answer to the 'what is a podcast?' debate: a YouTube video is something that shows up on YouTube; a podcast is something that shows up in your podcast app.
So YouTube. Yeah. As a YouTube video, some shows up on YouTube. A podcast is something shows up in my podcasts Exactly.
Dave is 'religious' about podcasts being downloadable because streaming fails in real life: stairwells, elevators, going in and out of coverage.
I will get sort of religious on the downloadable part because I fight a lot with streaming. Some there's some aspect of my life where I walk through stairwells and elevator elevator elevators and I go in and out of
Advertising worked until media went digital. As soon as people can control what they see in an automated way (TiVo, ad blockers), interruptive ads collapse, and you can't stop it, just like you can't stop spam filters.
as soon as you can control in an automated way. What the what the content is? It's just it's your it's a complete collapse. Yeah. And you can't stop it because it just the same way that you can't stop.
When companies replace report-writing with an LLM, they still have to prove the output is right by actually doing the work, so it becomes double the work.
you're gonna produce something with with the large language model. And then you'll always have to prove it. by actually doing the thing. It's double the word.
Advice to app developers: find a set of podcasters who want a specific thing and build your feature to that. Find the gap, put a rock in someone's shoe, meet the need. Live is a good place to start because live podcasters are hardcore and willing to switch.
when you're a developer build to a build to what somebody wants. Find out what your what if your podcast app, find out what find a set of podcasts podcasters that want a certain thing, build your app feature to the thing that they are wanting.
Dave thinks people forget 2019 was already recessionary. COVID-era money printing created a boom, and podcasting is now reverting to the mean.
A lot of people don't realize that 2019 was a recession. We were in an economic recession in 2019. Then COVID happened. Screwed out or you coincidence. We printed trillions of dollars. Yeah. And then had this boom. And now what we're doing is we're reverting after the party's over reverting to the mean.
Implementing a protocol is always reading plus reverse engineering. The spec alone is never enough, so step two is always testing against reality.
implementing a protocol is always an exercise I in reading and reverse engineering, you have to read the spec to understand it enough to begin the process of doing the thing, but that's not enough. It'll never be enough. You'll never blindly write a rhotic write things to a spec. Step two is always go see if it works.
You can't syllogism people into seeing your vision, because they don't share your frame of reference. You have to show it, or you'll just talk past each other.
What I'm trying to show is the thing that I see in my head and that's not is the same as a logical conclusion. It's a it's a vision. It's not an argument. And so you can argue all day long, and you're just talking past each other.
Claims that a protocol 'doesn't require a server' miss the point: an RSS feed isn't written on your front door for everyone to see. Every protocol (RSS, ActivityPub, Nostr relays) depends on some computer serving it.
You don't you don't write your RSS feed. You know, you don't write your your in pub on your front door and everybody sees it. It has to be served by a computer.
People in podcasting should use podcast apps to do their thing (live shows, boosts, promotion) instead of jumping over to YouTube. Use the apps in your own ecosystem.
podcast apps to do the stuff. Don't you know, instead of jumping over to, to YouTube or something else, yes, use the apps use the apps in your own ecosystem.
Formal standards bodies (like the RSS/podcast standards groups) are supposed to move at glacial speed; for them moving fast is a liability. For Podcasting 2.0, speed is a feature. Dave has no problem with either.
moving fast is a is a is a liability for a group like that whereas it's a feature for us
Spec proposals that nobody picks up don't need to be killed. If one is going to die, it will. Some take years to catch on: ActivityPub languished for about ten years before Mastodon made it huge.
if it's gonna die, it'll just die. I mean, it doesn't have to be it. I mean, look at activity pub. It took 10 years for anybody to do much of anything with that spec.
After building a working ActivityPub proof of concept, Dave deliberately paused to discuss open questions publicly before going so far down a path that it would have to be unrolled later.
I get this far. And I'm like, okay, all right, stop. Let's discuss publicly and come up with some things before I get too far down a path that we're gonna have to unroll, you know,
Dave sees fortifying the Podcast Index (so podcasting's data is durable and decentralized) as a years-long life mission. There's no hall of fame for that kind of infrastructure work. It's just the thing you do, and then you die.
get no recognition for that. Like, that's not you're not in any Hall of Fame's for that. That's just the thing you do and then you die.
After a dinner where his kids were mortified by a joke, Dave observes that young people are afraid of humor and argues that joking about cultural things is not disrespect.
jokes about culture. cultural things are not the same thing as being mean.
When a new feature (like adding a video alternate to an audio feed) means more work for creators, ask whether software can take that work away by doing it automatically while they're already doing the related task.
if adding a video source to the audio feed, is is more work? Could you take away the work? By doing it for them when they do when they're already doing the work?
People get freaked out by in-depth protocols, but if you sit down and bang it out, they're not that hard. Truly complicated designs (like Web Monetization) just don't get adopted.
I feel like people get really freaked out over over sort of a more a more in depth protocols like that. But they're really at the end of the day if you if you sit down and really bang it out. It's not that hard, right?
People hunt for new protocols to get decentralization and forget the web is the original decentralized system: DNS, URLs, links, everyone hosting their own thing. Podcasting, spread across thousands of hosts, is already decentralized; new work is just an overlay.
we forget that the web is the the original decentralized. Yes. Like, there's so many times where we look at these other protocols. And we're like, you know, we think, Okay, well, this, this protocol is going to meet my needs for decentralization.
Building a real podcast audience is a commitment, but it pays off. Trying to shortcut by going YouTube and digital advertising just cheapens your own product.
in the commitment eventually does pay off, if you try to shortcut and just say I'm going to do YouTube, yeah. You're you're just cheapening your own product.