5 ms·

> they could have an endpoint “/following.json”, containing an array of urls from people whose stuff they like to read. I'm game. This will work with Gemini t

> they could have an endpoint “/following.json”, containing an array of urls from people whose stuff they like to read. I'm game. This will work with Gemini too. Literally just an array of URLs? Don't you want at least a name for the array? Or does that make it took complex? { "following": [ "https://news.ycombinator.com", "https://aol.com" ] }

dreamlessfate 4y ago
This is just RSS. It was released 23 years ago. It was very popular and a big part of the web. The model that harryvederci is describing is exactly what the internet looked like prior to the rise of Facebook/Twitter, during the dawn of Web 2.0. Everyone had their own blog, on their own site. And you could curate your own feed of people you wanted to following using RSS feeds.
- harryvederci 4y ago
  Thanks for providing some historical context. I'm familiar with RSS in the sense of providing a way to publish my own updates, I do that on willfennel.com as well. I'm also using it (with newsboat) to "subscribe" to RSS feeds myself, but those aren't shared anywhere by me currently. What I'm not aware of is what you're saying, that RSS is a way to let people know what content I'm consuming. > The model that harryvederci is describing is exactly what the internet looked like prior to the rise of Facebook/Twitter, during the dawn of Web 2.0. How was the "following" part of it implemented back then? Where would I find out what blogs you were following, if I looked at your own RSS feed?
  - ZeroGravitas 4y ago
    This was known as a "blogroll" and would just be a list of links of other sites you read/liked. https://en.m.wikipedia.org/wiki/Glossary_of_blogging#blogroll https://en.m.wikipedia.org/wiki/Glossary_of_blogging#blogrol... Before that there web web rings https://en.m.wikipedia.org/wiki/Webring https://en.m.wikipedia.org/wiki/Webring
    - harryvederci 4y ago
      After a quick search I haven't found a good example of a blogroll yet. Was this something that was added as a part of the html on a page? I think an advantage of an endpoint with just the information (regardless of if that's in JSON, newline-separated plain-text, etc) would be that you could query them witout everyone and their mother having to create custom parsers. Think "consuming an API" instead of "surfing the internet in a browser". I believe webrings also fall into the "surfing" part, which is not what I'm going for here.
      - ZeroGravitas 4y ago
        Atom feeds also had the concept of a "via" link which would be used to give credit to the person you discovered it via.
        harryvederci 4y ago
        That's more like sharing a post and including its source. In twitter terms: retweeting. What I intend with the "/following" endpoint example is to say: "Here are some sources that I think publish worthwile content." Not: "Here is an interesting individual post, which I got from $source."
      - dreamlessfate 4y ago
        Here's an example of what ZeroGravitas mentioned. This is a blog from the "old" Web2.0 days, when social media was nascent. This old man still updates his blog almost daily with content that would have him banned on Twitter. He's one of the most insightful and inciteful voices in the world of advertising. https://adscam.typepad.com/ https://adscam.typepad.com/
        harryvederci 4y ago
        Yeah that definitely falls in the "surfing the internet" category, unless I'm missing a link to a bare list of sources he follows, so I don't have to write a parser. I understand it can be annoying when people come up with things that have already been implemented, and for which standards exist. But so far I don't see how what I suggested with a "/following" endpoint is exactly like RSS.
harryvederci 4y ago
Love the enthusiasm! I guess the "following" keyword could be helpful in case additional information would be needed. Maybe "blocking"? :-) On the other hand, then it's already becoming a potentially bloated idea with infinite expandability, which also would make the endpoint name weird or redundant. Both options have pros and cons, I guess. It was mainly meant as an idea, though, just as an example of how this type of data could be shared easily without needing a megacorp hosted platform.

dreamlessfate 4y ago

This is just RSS. It was released 23 years ago. It was very popular and a big part of the web. The model that harryvederci is describing is exactly what the internet looked like prior to the rise of Facebook/Twitter, during the dawn of Web 2.0. Everyone had their own blog, on their own site. And you could curate your own feed of people you wanted to following using RSS feeds.

harryvederci 4y ago
Thanks for providing some historical context. I'm familiar with RSS in the sense of providing a way to publish my own updates, I do that on willfennel.com as well. I'm also using it (with newsboat) to "subscribe" to RSS feeds myself, but those aren't shared anywhere by me currently. What I'm not aware of is what you're saying, that RSS is a way to let people know what content I'm consuming. > The model that harryvederci is describing is exactly what the internet looked like prior to the rise of Facebook/Twitter, during the dawn of Web 2.0. How was the "following" part of it implemented back then? Where would I find out what blogs you were following, if I looked at your own RSS feed?
- ZeroGravitas 4y ago
  This was known as a "blogroll" and would just be a list of links of other sites you read/liked. https://en.m.wikipedia.org/wiki/Glossary_of_blogging#blogroll https://en.m.wikipedia.org/wiki/Glossary_of_blogging#blogrol... Before that there web web rings https://en.m.wikipedia.org/wiki/Webring https://en.m.wikipedia.org/wiki/Webring
  - harryvederci 4y ago
    After a quick search I haven't found a good example of a blogroll yet. Was this something that was added as a part of the html on a page? I think an advantage of an endpoint with just the information (regardless of if that's in JSON, newline-separated plain-text, etc) would be that you could query them witout everyone and their mother having to create custom parsers. Think "consuming an API" instead of "surfing the internet in a browser". I believe webrings also fall into the "surfing" part, which is not what I'm going for here.
    - ZeroGravitas 4y ago
      Atom feeds also had the concept of a "via" link which would be used to give credit to the person you discovered it via.
      - harryvederci 4y ago
        That's more like sharing a post and including its source. In twitter terms: retweeting. What I intend with the "/following" endpoint example is to say: "Here are some sources that I think publish worthwile content." Not: "Here is an interesting individual post, which I got from $source."
    - dreamlessfate 4y ago
      Here's an example of what ZeroGravitas mentioned. This is a blog from the "old" Web2.0 days, when social media was nascent. This old man still updates his blog almost daily with content that would have him banned on Twitter. He's one of the most insightful and inciteful voices in the world of advertising. https://adscam.typepad.com/ https://adscam.typepad.com/
      - harryvederci 4y ago
        Yeah that definitely falls in the "surfing the internet" category, unless I'm missing a link to a bare list of sources he follows, so I don't have to write a parser. I understand it can be annoying when people come up with things that have already been implemented, and for which standards exist. But so far I don't see how what I suggested with a "/following" endpoint is exactly like RSS.

harryvederci 4y ago

Love the enthusiasm! I guess the "following" keyword could be helpful in case additional information would be needed. Maybe "blocking"? :-) On the other hand, then it's already becoming a potentially bloated idea with infinite expandability, which also would make the endpoint name weird or redundant. Both options have pros and cons, I guess. It was mainly meant as an idea, though, just as an example of how this type of data could be shared easily without needing a megacorp hosted platform.