This page lists part of the section. The page links below work without JavaScript. With JavaScript, a filter searches the whole section and pages the matches.
Bitcoiners and podcasters see things differently. Before building something or proposing a standard, slow down, get consensus, and fully understand the industry and what its audience needs.
you need to understand that industry fully before you dive in and start building something and trying to propose a standard standards aren't bad, but you have to at least understand what the audience needs before you make it.
Open source, and podcasting especially, moves slowly. Dave gives heads-ups about progress, even coyly, so people see momentum and don't lose heart.
things in the in open source projects. Prior and then podcasting, especially things move slow. So I like to let people know that there is momentum and things are happening and don't get discouraged.
His case for lightning, in order of importance: payments go directly from listener to creator, fees are tiny, and you can run the whole thing yourself. In the fiat world you don't even have that option.
No middlemen, the payment goes from directly from the listener directly to you. There's nobody in between. And the fees are small, very tiny networks, you run it yourself.
Under value-for-value, all that matters is the quality of the content. Make a good product and you get value back.
All that matters is the is the quality of the content. Yeah, yeah. If you make a good product, you get value. That's there's no other. There's no other secret sauce.
Dynamic ad insertion that shifts chapter, transcript and soundbite timing is like a banner ad that wrecks a website's layout. Nobody would tolerate that on the web, and choosing it means picking easy money over content quality.
Dynamic ad insertion into your audio that results in screwing up the timing of your other metadata points is like inserting dynamic banner ads into your website which randomly screw up your CSS and change the way your website layout
If you hold your own keys, losing a wallet's infrastructure only costs you functionality. You can restore the keys elsewhere and the funds are still there.
if you have control of your own keys, then even if the infrastructure underlying your wallet goes away, you can go and spin up a wallet in a different location. reapply your keys and your funds are still there like you, you don't your your funds live on the chain, they don't live in the wallet.
On ActivityPub comments for podcast episodes, Dave says the most important design choice is the 'show comments' button. Like a Mastodon content warning, putting comments behind a user action removes 90% of creators' worry and shifts the job of policing what the user sees onto the user, not the creator.
The comments don't show unless you go and ask for them. To me, that's critical.
People underestimate how hard good podcast search is. Google had a clear quality signal in incoming links. Podcasting has nothing like that: download counts are known only to the podcaster, each app sees only its own slice, and on-device apps like Castamatic don't see even that. Dave says you can mine for topics, but you'll never get a real 'best' ranking.
they have a clear signal, the PageRank would signal that link based on incoming links. That was a clear signal, that of quality, that this this thing rises to the top of the stack. As you have incoming links, there's no, there's nothing like that in podcasting.
Dynamic ad insertion keeps blocking features like synced transcripts. Dave says it's very solvable: the ad slots could come back as markers in the ID3 tag or in download headers, so an app gets a list of insertion points and lengths and can adjust its offsets.
And to me, this just seems like an eminently solvable problem.
Bitcoin culture's problem has long been that people don't spend it, which feeds Gresham's Law. Dave says anything that gets people to actually transact (boostagrams, zaps) helps form a market and enable price discovery. Bitcoin isn't financialized yet, and the first step is getting people comfortable spending it.
the problem with Bitcoin for a long time with the culture of Bitcoin has been people are not people don't spend the money. So you're sort of like feeding Gresham's Law by by not doing anything.
Lightning's growing pains look scary because they're playing out in public. Dave says sysadmins deal with this all the time: lots of technology is just this fragile, but it's hidden in a datacenter, and any new technology has funky problems that get worked out over time.
The fragility that you see in a lot of these things, there's a in like lightning and this sort of, there's a lot of stuff, a lot of technology that is just that is easily this fragile.
Not in the audio. Guest episode (Nathan Gathright, 0.35 threshold). Adam's cue directly before ('Dave Jones?') hands the floor to Dave, which supports attribution.
Dave is comfortable with KeePass's open-source code itself. What makes him uneasy is installing it on a phone, where the app store offers a pile of third-party KeePass apps of unknown provenance.
It's when I go to install it on my phone from the App Store. And it's like, there's all these ones that are there, like KeePass apps,
Dave likes how the Lightning ecosystem is evolving, starting with LDK: split a node into its component parts, swap in a different database or an LSP, and you still have a functional node. Breez uses this deconstructed node as a service-provider model.
you can mix and match these component parts. And what you end up with is still a functional node, right? But it's just kind of like, it's like deconstructed, and you're running it in different ways.
The Podcast Index promised an API free for anyone's use forever. To keep that promise, any surplus goes into the bank until it covers years of hosting, so the API keeps running if someone gets sick or the project is derailed, even with no income. That means keeping expenses low enough to keep saving.
we're going to stash that money in the bank for as long as it takes to build up some number x number of years of hosting fees, where if something happened, and one of us got sick, or something happened, and the project got derailed, we would have an X amount of years of runway
Asked for a primer on table locking, Dave explains atomic operations. A transaction must be sure the data won't change in the middle of the change; otherwise parallel updates and deletes corrupt the data. The query plan locks the rows that could conflict, and that locking is critical to safety.
you have to be sure that when you put when you when you do a transaction to a database, you have to be sure that the data is not going to change in the middle of, of the change you're trying to make.
When email clients suggest autocompleted replies, Dave rejects them, even when the suggestion is exactly what he was about to say. He doesn't want the other person to think he just hit a button without thinking, or to come across as rude.
Yeah, and I would, I would reject it. Because even though it's what I was about to say, I don't want the other person to think that I just hit a button and didn't actually think about what I was saying. Like, I don't want to give them a canned response.
Responding to a Brookings proposal to flag podcast 'misinformation,' Dave asks whether George W. Bush's State Department would have hesitated to flag podcasts doubting WMDs in Iraq, a claim that proved bogus. Each side builds silencing systems while it's in charge, then the power flips.
if that if the State Department under George Bush had access to some sort of system, that would, they could instantly send out a bat signal to flag all the podcasts that were created?
Mastodon is better than Twitter, but Dave says that doesn't mean it can't be worse under the right circumstances. An open reputation system assigning social scores could have instances auto-blocking users. The same could happen in podcasting if interested parties transcribe everything and use AI to tag 'misinformation,' regardless of whether it later proves true.
That the things decentralization doesn't always protect you from the bad stuff because
Responding to calls for podcast apps to publish recommendation-algorithm guidelines, Dave describes what the algorithm actually is in most apps: grep the request logs or subscriber counts and recommend whatever is at the top.
the algorithm get in most podcast apps, it goes like this GREP the request logs and find out which which at which podcast is being recommended, like requested the most?
P2P live streaming struggles because it's hard to get everyone to seed and participate. Like DHT networks, once something such as IPFS reaches critical mass you're good to go. Dave says the initial hump of getting enough peers online is the problem, and live streams would face it every time.
just, it's hard to get everybody to, to seed and to participate. Right?
People balk when a Bitcoin app asks for a driver's license. Dave says Apple subscriptions feel seamless only because you did KYC for your financial accounts years ago. Bitcoin onboarding hits you with it all at once, but it's no more onerous than getting a credit card or bank account. It's actually less.
It's funny, because it's really no different than anything else we do. It's just that you're hit with it all at once.
Agreeing with Todd Cochrane that podcasting can't have the plug pulled on it, Dave says Podcasting 2.0 has tried from day one to make sure everything can continue to exist without Adam and him.
Todd because we have like that's what we've tried to do from day one is to make it where we don't to where all this stuff continues can continue to exist without us.
Agreeing with Alex Gates, Dave says universal keys as universal identities scare him a bit. There's value in leaving a comment on a WordPress site without exposing yourself fully on every service, the same way you can generate a fresh Bitcoin address for each transaction for a measure of anonymity.
I'm gonna have to agree with Alex on this one, though, though, in this regard, that the universal keys that serve as universal identities scare me a bit.
Not in the audio. Alex's surname (Gates) and the Nostr tag are inferred from context and the show's regular cast; the audio says only 'Alex' and 'universal keys'.
The open web has collapsed to Safari and Chrome, with no diversity left in browser engines. Dave says podcasting almost stands alone in having listeners use dozens of completely different apps to reach the same content, and that's something you don't want to give up.
What happened to the open web where you've basically got, you know, everything's come down to Safari and Chrome.
Podcast apps are personal: each person has a set of features they care deeply about, and sometimes it's just that an app feels more native. Dave says all it takes is one solid feature someone can't live without. For him that's live: he needs an app that pushes a notification when a show goes live.
all it really takes is one solid feature that a person can't live without. And then you've got to you've got as an app developer, you have a customer like that.
Not in the audio. Guest episode (Todd Cochrane, Mike Dell; 0.35 threshold). The surrounding passage is first-person about Dave's own love of live items, which supports attribution.
Amid podcast industry layoffs and turmoil, Dave says companies with great business models (a good product for a good price that provides value) will be around and stay strong. There are plenty of them in podcasting. It's the Silicon Valley creepers who won't last.
the companies with great with great business models with that did just sell a good product for a good price and
Praised for the Podcast Index API's richness, Dave says his real job is keeping it from becoming an unwieldy monster. As features pile up you have to maintain a structure that makes sense, like every web developer who promises that this time the stylesheet will stay clean.
I'm trying to keep it from becoming an unwieldy monster. That's what's really going on here. I want you have to maintain some structure and something that makes sense otherwise you just end up with a big mess.
Spotify's record-label contracts make payouts rise in lockstep with subscribers and revenue, so its margin is stuck. More premium subscribers from podcasts don't help, because they only raise label fees. Dave says the only escape hatch was higher-margin podcast advertising.
their only escape hatch was to make more money through podcast advertising, because this just getting more premium subscribers through podcasts doesn't help you with your main with your main expense, which is record labels.
Dave publishes Podcast Index stats that exclude junk feeds, because he sees his numbers quoted everywhere. Inflated counts (like YouTube's hours uploaded, when little is watched) could mislead someone into building a product the ecosystem can't support. He says he's the perfect person for it because he has no agenda and doesn't even care much about stats.
I don't want that to mislead somebody into thinking that the podcast ecosystem can support a product that they may think otherwise about.
Looking back at 20 years of cloud disruption, Dave says most of the software that disrupted vertical markets did less than the on-prem incumbents. It won by being cloud-based and wrapping complex things in a simple, pretty UI full of charts. Much of it was smoke and mirrors.
99% of the software, that that is, quote, unquote, disrupted the market in any particular vertical market. Most of that software didn't do new things.
Explaining why he thinks Google is in trouble, Dave points to Standard Oil. It held 90%+ of the kerosene market, but when cars made gasoline the thing, it wasn't prepared to transition. He says Google's equivalent is search still meaning 'ask a question, get a list of links.'
But their dominant position was in kerosene. That was their big product. And they owned that market to the tune of 90 plus percent market share in the kerosene market. But then gasoline became the thing when cars, when cars began to be rolled out in mass. They were not prepared to transition.
Not in the audio. Guest episode (Oscar Merry, 0.35 threshold). Adam directs the follow-up question to the same speaker, and the content matches Dave's recurring arguments. The historical claims about Standard Oil are Dave's, not verified here.
Dave wants tools like Helipad to run on people's own infrastructure. But open-sourcing code and curating it so others can run it on their systems commits you to long-term support, a lot of work and bug fixes, so for niche tools you have to ask whether the tradeoff is worth it.
when you curate it and care for it in a way that lets other people run it on their system. Yeah, that's, that's you're signing up for support long term for a lot of work, a lot of work and a lot of bug fixes and things like that.
The value block was designed to be plug-and-play with anything people dream up. Dave will merge any PR that fits the spec and the spirit of decentralized payments, even if it isn't what he'd personally push. His reasoning: if the base protocol is open, decentralized, free and accessible, the layers on top may not be perfect, but they'll inherit that base level of freedom.
if the base protocol itself is open, and decentralized and free and accessible, then what's built on top of it, like layers, 234, whatever, they may not be perfect, but they'll at least maintain that that base level of freedom.
Most of the Podcasting 2.0 namespace makes things better than they were: an evolution. Dave says live in independent podcast apps feels like something genuinely new, turning every app into a radio receiver. He's normally not a live listener because having to sit at a computer and know the schedule was too much mental overhead.
The rest of the stuff in the namespace, yes, it's cool. But they're, they're like, they feel like just sort of making things making things better than they were. They feel like it feels like an evolution of things. This feels like something truly new in a podcast app.
Dave's bold statement: if you're a podcaster who cares about Spotify, it's for monetization, not features like chapters, polls or Q&A. Nobody uses that stuff. Spotify listeners don't pay for podcasts and don't care about the features either.
You don't give a crap about features. Nobody's putting all the stuff that they that they're spending time on. Polls, q&a, nobody uses that stuff. Nobody cares.
Asked about Venmo and other big players entering the space, Dave admits he doesn't really care. Big moves make waves for good and bad, but what matters to him is being able to transact directly, peer to peer. So much is built on big platforms that he just wants to minimize that, while accepting you can't fully escape them.
to me, what matters is that you're able to do things direct, directly, peer to peer. I mean, and, and I don't think we I think so much of so much of everything that we do today is built in these on these big platforms that I just want to minimize that.
On using Mastodon as the 'S3 of social' for podcast comments, Dave says we can all wish things were more generic and higher-level, but that's wishing versus accepting reality. A practical bonus of piggybacking on Mastodon instances is that moderating comments becomes the instance's problem, not the app's.
we can all wish that things were more generic and more sort of higher, higher level but then it's sort of like it's wishing versus sort of accepting reality as it is
We picture podcast listeners as a certain type of person who's geeked out about podcasting. Dave says that's the growth opportunity. A coworker who streams SiriusXM all day could listen in a podcast app if the content were there and the interface felt like a normal player, with live content, country, talk radio and so on.
It's like people have to be into podcasting, to listen to podcasts.