We need a light switch for static hosting | Hacker News Reader

We need a light switch for static hosting

willfennel.com

13 pointsharryvederci14 comments

It's too easy to lose a domain, though. I had a decent one with email forwarding set up just in case something happens to my gmail, gave that email address to everyone...and lost it when i neglected to pay stuff for months cause of health problems. Now I have to beg some squatter to get it back.

Oh man, that sucks. But with people randomly getting cancelled on these platforms though, I guess it's still less risky to get a domain?

Depends. We have GDPR here so at least I can request all my data from these platforms. If I screw something up on my server or domain, noone can help me.

Someone can still try to sue you over trademark or copyright infringement. That said, it's quite harder for the one using it against you compared to sheer AI system flagging your account, as it requires paying lawyers a going to a court.

> 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"
      ]
     }

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.

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?

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#blogrol...

Before that there web web rings

https://en.m.wikipedia.org/wiki/Webring

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.

Atom feeds also had the concept of a "via" link which would be used to give credit to the person you discovered it via.

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."

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/

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.

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.

Reply on news.ycombinator.com