Most people don’t know how to go to a podcast’s website, copy the link to the RSS feed, and then paste it into their podcast app. They just search for the podcast in their podcast app. So, if you produce a podcast, you need to be in the major podcast directories to get traction. Thus, in effect, podcasts are highly centralized.
This feels like moving the goalpost to me. If PeerTube succeeded, but most of its videos were being linked and shared on Facebook and Reddit, would you call it centralized?
It's fairly common for me to see people share direct links to a podcast's website if they want to share a specific episode. And even from the point of view where iTunes is a centralized indexer, they're still very clearly pointing at a decentralized service.
The biggest podcast aggregation service in the industry decided to rely on a decentralized model rather than hosting files themselves. That doesn't count for anything?
The fact that every single podcast app is using a distributed protocol (RSS), including Apple, doesn't count for anything?
Note that my original point was not that podcasts are centralized. I merely took issue with somebody else claiming that podcasts are decentralized. They’re decentralized for distribution. But distribution is the easy part on the internet. It’s aggregating attention that is the hard part. And that part of the podcast market is entirely centralized.
> But distribution is the easy part on the internet. It’s aggregating attention that is the hard part.
ATM for video the content hosting, distribution, discovery, indexing, consumption, and commenting are all centralized. Let's say PeerTube brings the video world inline with podcasts and knocks that list down to just discovery and content indexing. That would be a really big win. That would be way more decentralized than what we have now.
If you're right, and solving all of those problems are the easy part, then PeerTube will probably be an incredible improvement to the video ecosystem.
Podcsts may not be perfectly decentralized, but they're pretty stinking close. Services like iTunes are basically card catalogues at this point. The podcast app I use doesn't even include iTunes ratings, reviews, or suggestions, so they've obviously made the decision that these aren't features their users care about.
My experience with podcasts is I get recommendations in a decentralized manner from friends, family, social media, and online articles. Then I go to one of several centralized card catalogues and search for the podcast by the name. Then I add the decentralized source to any podcast app (all of which work with every podcast regardless of who developed them) and use RSS to download a file to my physical device, which makes it easy for me to back up or mirror the file to other devices if the source ever goes down in the future.
If 99% of my experience is decentralized, does the centralized 1% that basically boils down to a list of urls and a regex expression override that? Especially keeping in mind that nearly every podcast app still provides a mechanism for you to bypass that list whenever you want, and any good podcast app allows you to search for a podcast across multiple preloaded sources at the same time?
Maybe this is a better approach: what happens to the podcast market if Apple stops shipping a podcast app?
The priors that come to mind when trying to answer this hypothetical is what happened to independent blogs when Google shuttered Google Reader and what happened to news companies when Facebook tweaked their algorithm to show less 3rd party content. In both cases, distribution was decentralized (blogs host their own RSS feeds and news sites host their own content), but attention was centralized (people accessed blogs through Google Reader and news through Facebook). In both cases, the decentralized distribution failed prevent a massive decrease in traffic to blogs and news sites. And that’s because competition for attention is fierce, so aggregating attention is hard.
Neither of those things are dead though (news is struggling with the adblocking apocalypse, but that's a different category of problem). I kind of get what you're saying, and I agree that Facebook and Google are powerful, and that we should look for ways to distribute that power more evenly.
But at the same time, to me those scenarios kind of look like distributed architectures doing the jobs they're supposed to. I mean, Google Reader's shutdown hurt bloggers less than Live Journal's did, right?
Another way of looking at it: think about what happened when Microsoft bought Github. A bunch of people panicked, but for the most part, it was fine - because Git repos are decentralized. And a bunch of people came out with hot takes that said, "Git's not really decentralized, because Github is the part that matters."
But... no, for the most part it's decentralized, and we saw the benefits of that architecture.
If you have a decentralized core you might interface with or feed off of a few centralized services, but you will be more resilient and better equipped to deal with their failures. You don't have to be perfect to reap most of those benefits.
If iTunes stopped distributing podcasts, that market would suffer. But I'd still be able to directly share episodes on Twitter and Reddit, and there are at least 2 other preloaded sources on my listening app that could be serving the same purpose within a day with zero change to the way I find new podcasts or download them. The big change for people like me would be that when searching, I would click the second button on a list of sources instead of the first one. It would definitely hurt the health of the network (mostly just for iPhone users), and you can make an argument that it would disproportionately hurt the health of the network, but it probably wouldn't kill it.
But suppose Apple or Google stopped distributing an app store. That market would instantly die, for basically everyone, and no one would be able to do anything to save it.
