3 ms·

This is a remarkable deep-dive into the history of RSS and the ideas and vision behind it, and how those camps diverged. But despite beginning the article with

This is a remarkable deep-dive into the history of RSS and the ideas and vision behind it, and how those camps diverged. But despite beginning the article with Kevin Werbach's prediction that syndication was going to become a dominant business model on the web, this never quite came to pass, and RSS took on the role of moving media between aggregators and consumers, rather than between authors and aggregators as originally intended. With the shift in usage to the edge, RSS found itself in the company of adhoc webpages that disseminated the same content but supported scripted adtech to enable just-in-time display advertising -- advertising which, over time, became a significant source of publisher revenue. RSS was often surfaced in a different user-agent and unable to clearly accommodate scripted ads, so first-party RSS feeds were practically giving away the content for free. With time, this greatly contributed to the waning appetite of commercial publishers for RSS, despite it remaining popular in circles were the content was meant to be spread wide at no cost. Although many news organizations still offer RSS feeds, they're not really marketed prominently, and remain a loss leader to make savvy readers happy. Since alternate business models never achieved the same uptake, efforts like Google's AMP (and its offshoots like webpackage [1]) are a practical take on the syndication concept again, where the usage of adtech is assumed from the start, and the surfacer of the content becomes decoupled from its author. [1] https://github.com/WICG/webpackage https://github.com/WICG/webpackage

toyg 8y ago
> syndication was going to become a dominant business model on the web, this never quite came to pass This has happened in at least one media segment: podcasting. Ads are simply inserted in the medium proper, or the feed gets paywalled, or money is generated while producing the content (live shows). It is entirely possible to reproduce this achievement in other media sectors, it just needs to be done wisely: 1. a catchy name ("rss" or "atom" are too geeky, "feeds" or "syndication" are too generic, "newscasting" is another thing). 2. leave clients and formats independent, so the immediate needs of users prevail over broadcasters' ones. Broadcasters are good at telling people what to think, but terrible at telling them how to use or consume anything. It also removes the petty fighting like "we won't rely on a platform built by $competitor". 3. Leave ad-management and general money-making schemes up to publishers. They can license feeds to specific clients only, they can do paywalls, micropayments, embedded ads, whatever. The client should assume nothing, simply move content from A to B on a schedule. All this can easily be implemented on HTTP codes, they are all there already. 4. The real point of feeds is exactly this: move stuff from A to B on a schedule. Metadata is almost irrelevant. Content display is a side-show. What podcasting delivers is control: it gets the content a user wants, as soon as it's available, and makes it ready to be consumed when the user needs it, in the way s/he prefers. This concept is what should move the effort, not dreams of semantic sheeps. 5. Because it optimizes pre-fetching, the format shines where on-demand fetching is time or money-expensive. Niches where it would (or does already) work well, ae things like auto-downloading TV episodes, delivering big images (something for visual artists to cut off Pixiv and friends... ?), delivering big binary updates and so on. With the rise of single-page-applications, one could even devise a format that delivers an entire website, ready to be consumed instantaneously without requiring lots of mobile data usage. Bootstrapping it in "stingy" markets (Southern EMEA etc) could help. 6. Good directories for clients to use. Podcasting really took fire when good directories emerged. (on a sidenote: AMP is evil and publishers are shooting themselves in the foot. It seriosly risks to become an "iTunes" moment, where an entire industry relinquishes control of their market because they don't understand long-term consequences.)
- whateveruser 8y ago
  Rest of your suggestions are quite thoughtful, but > They can license feeds to specific clients only this is a bad idea.

toyg 8y ago

> syndication was going to become a dominant business model on the web, this never quite came to pass This has happened in at least one media segment: podcasting. Ads are simply inserted in the medium proper, or the feed gets paywalled, or money is generated while producing the content (live shows). It is entirely possible to reproduce this achievement in other media sectors, it just needs to be done wisely: 1. a catchy name ("rss" or "atom" are too geeky, "feeds" or "syndication" are too generic, "newscasting" is another thing). 2. leave clients and formats independent, so the immediate needs of users prevail over broadcasters' ones. Broadcasters are good at telling people what to think, but terrible at telling them how to use or consume anything. It also removes the petty fighting like "we won't rely on a platform built by $competitor". 3. Leave ad-management and general money-making schemes up to publishers. They can license feeds to specific clients only, they can do paywalls, micropayments, embedded ads, whatever. The client should assume nothing, simply move content from A to B on a schedule. All this can easily be implemented on HTTP codes, they are all there already. 4. The real point of feeds is exactly this: move stuff from A to B on a schedule. Metadata is almost irrelevant. Content display is a side-show. What podcasting delivers is control: it gets the content a user wants, as soon as it's available, and makes it ready to be consumed when the user needs it, in the way s/he prefers. This concept is what should move the effort, not dreams of semantic sheeps. 5. Because it optimizes pre-fetching, the format shines where on-demand fetching is time or money-expensive. Niches where it would (or does already) work well, ae things like auto-downloading TV episodes, delivering big images (something for visual artists to cut off Pixiv and friends... ?), delivering big binary updates and so on. With the rise of single-page-applications, one could even devise a format that delivers an entire website, ready to be consumed instantaneously without requiring lots of mobile data usage. Bootstrapping it in "stingy" markets (Southern EMEA etc) could help. 6. Good directories for clients to use. Podcasting really took fire when good directories emerged. (on a sidenote: AMP is evil and publishers are shooting themselves in the foot. It seriosly risks to become an "iTunes" moment, where an entire industry relinquishes control of their market because they don't understand long-term consequences.)

whateveruser 8y ago
Rest of your suggestions are quite thoughtful, but > They can license feeds to specific clients only this is a bad idea.