This page lists part of the collection. The page links below work without JavaScript. With JavaScript, a filter searches everything and pages the matches.
Like a hyperlink, a remote item points to the thing where it lives instead of recreating it, so the metadata never goes stale or gets mangled.
it's just saying, hey, there's a thing over here. And rather than me recreating it in it's good, just go get it from the original source. You're telling you that way. You don't have to worry about the metadata going stale or getting screwed up?
Dave calls Dave Winer's 'Rules for Standards-Makers' the developmental bible of Podcasting 2.0. One key rule: when developers ignore part of a spec, update the spec to match reality.
Reading from rules for standards makers remember which is the develop the developmental bible of podcasting 2.0. If pressed if practice deviates from the spec, change the spec
The chapters 'toc=false' flag solved an immediate problem, but Dave calls it a hammer: a short-term fix that needs a proper long-term solution (a type parameter) that respects both creators and listeners.
the because the table of contents is a hammer, you you are you are saying as the creator of the chapters, you are saying, I don't want this visible, but some people might want it visible.
Every specification starts with the problems in front of you. Try to envision everything that could come in the future and you go too far. When new use cases show up, respond by extending the spec responsibly rather than ripping it out and starting over.
this happens to all specifications. You you start where you know, you, you know, you start where you know, but it's like, Okay, here's the things we have in front of us. And if you go too far, and try to envision everything that could possibly come in the future,
Big tech isn't worthless to creators, but the deal is lopsided: you give the platform all the value of your content, and if you monetize enough, it lets you keep a little.
Yeah, you give us all the value there. You let us monetize your show. And if you monetize it enough, we'll let you keep a little bit.
Quoting Captain Picard while describing how hard it is to track changing remote items across database runs: sometimes doing everything right still doesn't work out.
It's a quote, to quote Captain Picard. From next generation. You can make all the right decisions and still lose.
Many software projects (Dave's example is Nostr) feel they must reinvent everything, pulling attention and development resources away from things that already work. Searching for a killer feature by rebuilding podcasting could burn years of work.
noster suffers from, from what happens to a lot of software projects where they feel like they have to, they have they're going to reinvent everything and in this is right it is it is sucking a lot of attention and
Someone who writes a 9,000-word email has something to say. Dave would print it and spend a Sunday morning reading it.
I am annoyed when somebody sends me a 9000 word email. I'm like, I will print that thing and spend a Sunday morning reading it. Because I mean, that is a person who has something to say, and you better listen.
Dave's 'big red button' idea: once a control system like a censorship list exists, it will get used, and eventually on everyone, including the people who built it for 'the other guy'. Thinking otherwise is ignorance of history.
when you build when you build it, it's there. And it compels it's compels its own use. You don't build something just for the other guy. It when you build it, it's it's going to be used again.
Before criticizing Wavlake's limited splits rollout, Dave puts himself in the developer's shoes and asks whether the community would judge another company (like Bandcamp adding RSS) the same way.
I just want to make sure there's not a double standard because like if, if ban it, let's say Bandcamp tomorrow, said that they're gonna start doing RSS feeds for all their for all their music,
Dave draws motivation from Podcasting 2.0 developers who plug away every day. Everyone gets discouraged; one measure of character is that you keep doing the thing anyway and don't let it overwhelm you.
One measure of character is you get discouraged, and you get down. But you still keep doing the thing. You don't stop. You don't let it overwhelm you.
Category systems come down to a community-curated master list or freeform tags; both have pros and cons. Dave came around to freeform hashtag-style tags you can aggregate after the fact, because a master list needs someone to curate it and can't anticipate edge cases.
But because when you have a master list somebody has to curate the list in some and then you have you're gonna have
Even as the Index pushes toward decentralization, Dave puts it on the table first: doing decentralization well while keeping an acceptable level of service is very hard. Mature projects like Tor show it takes many people over many years.
Decentralization is incredibly difficult. Yes, doing it well, and maintaining a level of a level of service that is acceptable stuff that works.
Attribution schemes that try to link a download's IP address to a later purchase on some retailer's website are flimsy. Dave's blunt view: nobody can attribute correctly.
All of this is just a completely flimsy House of Cards. Nobody. Nobody can attribute correctly. It's all a lie. Every bit of it.
Podcasting is secretive about downloads and revenue. When everything is open, there are no question marks: you know what's working and what isn't.
When everything is open, there's no question marks, you just know what's working, or it's not. And it just is, it's real, it's either reality, or is not, you're living in the truth,
When someone is cut off on accusation alone, the accusation itself becomes the punishment. Dave says it's not about whether the person is guilty; it's that the process punishes before anything is proven.
the process is the punishment. This is the process is the punishment because you get just do Is the accusation of a thing? Not, you know, I know nothing about Russell Brand that he may be completely guilty. But that's not the point.
Discussing PodcastOne/LiveOne's roll-up 'scoop and score' plan, Dave points out it has been tried about a hundred times. The 'next big thing' pitches are often the old thing with a new name: it didn't work before, but supposedly it'll work now.
the the strategy only problem is the strategy is probably one that is not going to work, because it's been tried about 100 times
Podcasting 2.0 now builds for music and audiobooks, not just spoken word. Not every feature must apply everywhere, but when building for one medium, ask whether it could also apply to the others, and add a small attribute tweak if that makes it work.
it doesn't mean that every feature has to be equally applicable to all the others. But I think if if if you're building something for one, and then you say, Okay, well, could this apply to also music and audiobooks?
Listeners can only afford a handful of subscriptions and soon have to choose which to drop, so paywalls limit reach and choice rather than reliably producing income.
I'm already having to make decisions already. And I've only got, you know, four or five of these things. And I'm already like, Okay, I'll get like, what if I want to replace one with another, I have to make a decision, like paywalls paywalls, don't do what people think they do.
Below the few established companies is a large tier of startups that will say anything and produce any metrics needed to get acquired. Much industry cheerleading is just exit strategy.
And then you have a much larger, like tier under that of startups, that that'll say anything necessary and produce any metrics necessary in order to be acquired by that high by that desk here, yes. And achieve an exit strategy
Industry reports say everything is great while successful shows with big audiences are hurting. Spend enough time in the podcast or ad-tech conference bubble and you start believing the reports over what you see with your own eyes.
How in God's name does this bullcrap fit with what I see with my eyeballs. But if you live in the podcast, conference bubble, or the divergent digital advertising conference bubble, you start over time to think that's actually what it really what reality?
If podcasters tie themselves to the digital advertising ecosystem, they get dragged wherever that wind blows and have no control over it.
you're just gonna You were just going to plug yourself in to the digital advertising ecosystem in whichever way that wind blows, that's going to drag you right along with it. And you're going to have no more control over it. None.
Like rappers with millions of streams who can't sell concert tickets, some podcasts have paper popularity: big numbers in a database but no real audience that cares.
you might call it like paper popularity or their sequel, their sequel, famous, you know, they're just famous in a database somewhere. They're not they're not actually.
Treating podcasting as an industry that depends on digital ad revenue lets people redefine what a podcast is to follow the ad money. Dave's view is that there is no podcast industry in that sense.
Yeah. And if that's the case, then I'm sort of I'm sort of of the opinion that there actually is no podcast industry.
A chart built on boosts and streams charts what has already happened: by the time a song is on it, people have played it, enjoyed it, and paid for it. Gaming is possible, but nobody pays for a bad song.
You don't get more money? Or more SATs for having a shitty song. That's number one. No, no, because nobody's going to nobody's going to listen to that and pay you for that.
If stopping people from gaming systems were easy, most of the internet's problems wouldn't exist. Plan accordingly.
If if anti gaming of things was was easy, or a solvable problem, most of the problems on the internet would have already gone well who wouldn't exist? This is not an easy
Self-reported metadata (like boost TLV records) can be faked by anyone with a command line, so it can't be trusted for rankings without something verifiable to key on.
So I don't trust the DOB record at all. Got it. Somebody could create a command line and just pump out to vase with all kinds of fake information that is completely possible. There's no way to stop.
When data is incomplete, label stats as the best approximation and state the direction of error: the numbers could be higher, but they can't be lower.
it does make a point you know, it's this represents the best approximation of how many sets are flown to artists and it says the numbers could be higher but they can't be lower
Live items are ephemeral, so boosts sent during a live show can't be attributed later unless the live item has a GUID and that same GUID persists into the published episode.
the live item would have to be have a Gu ID and that Gu ID would have to persist into the published episode as the same Gu ID so that I could go back and know what is what. So I can know that they match.
One person feeds a few bullet points to an AI to write a blog post or press release; the reader feeds it back to an AI to ask what it means. The content in between is just an envelope, like a bank's pneumatic tube.
you've got the guy on the front end say you know, saying write me a blog post and then you got the gal on the other end saying take this blog post and tell me what it means. And basically it's just the the actual blog post is just an envelope.
Like Amazon reviewers who only complain about shipping, feedback gets muddled when delivery problems are mixed with the thing itself. When reviewing Adam's first music show, Dave split it in two: the delivery (Safari kept killing the audio) and the show itself (four out of five stars).
you have an Amazon review where they're supposed to be reviewing the product. Yeah, all they do is bitch about the shipping. Okay, so I for this is there's going to be a two part two part.
Overcast can throw $5,000+ a month of hardware at aggregation; Podcast Index spends about $620 and aggregates four times as many feeds. So instead of costly real-time deletes that lock tables, a lightweight 'sweeper' cleans up later. That's fine 99% of the time, and the rare urgent case gets handled by email.
so you can see the difference there, you can overcome this problem with sheer horsepower. And money. We don't have that. So we have to be wise about about the way that we allocate the things that we do.
People are detaching from social media and big tech because they can't process it anymore. Podcasting 2.0's role is to be there with a catcher's mitt for everyone who exits.
we have to be the we've we've we're here with with a catcher's mitt, we're catching, we're catching it. Everybody who exits the technological system we're catching
Stay in IT long enough and you reach a paradox: you love the problem-solving and creation, yet you're deeply pessimistic and aggravated by it all at the same time. Podcasting 2.0 doesn't get a free pass from these criticisms either.
there's something in you that loves the technology and loves the problem solving of it, and then the innovation and the creation of the thing. But you're also so pessimistic and so aggravated at it all.
Drawing on Ellul, Dave warns that building a tracking-and-optimization framework around podcasting (tracking pixels and the like) will damage the medium.
Because otherwise we create a technological framework around podcasting, that's going to kill it.
Believing in elegant, simple solutions to inherently hard problems (like decentralization or general AI) is self-deception. Even a beautiful three-line macro had to become an ugly static table to handle one odd network card constant.
The easy problems are all in technology, the easy problems are all solved. They were all solved a long time ago.
Technology piles up constraints (security, verification, app-connected appliances), and AI arrives promising to ease the burdens technology itself created, such as Buzzsprout needing AI to keep up with new podcast features. Dave sees it as the end of a phase, the system begging you not to leave.
is the end of innovation is not the beginning of any new phase of innovation. It's the end it's not in the AI is not innovation. It's supplication.
After a week stuck in a vendor's automated identity-verification loop, Dave described a growing awareness that the rule-based techniques we're bound to no longer serve a purpose worth the effort of following them. While technology moves fast, excitement hides its second-order effects; when it slows, those constraints become the main experience.
there's this emerging awareness that the the rules and the techniques that we are bound to in all of these areas of life don't actually serve a purpose for which their usefulness outweighs the effort we have to x put out to adhere to them?