5 ms·

Inlinking third party resources is literally in the original hypertext specification of the web. An HTML site might want to play a video hosted by a third part

Inlinking third party resources is literally in the original hypertext specification of the web. An HTML site might want to play a video hosted by a third party, or load a shared resource. Tags like <img src> even at the start of the web was intended to let you load images from any host. I would argue a browser that doesn't access resources on other domains or IPs is not a web browser at all, but instead some weird sandbox that lets you run a self contained app.

inetknght 4y ago
> Inlinking third party resources is literally in the original hypertext specification of the web. Indeed, but it was intended for real content. That intent is being abused by large corporations for tracking purposes and necessitates a heavy-handed block to it. Plus, including third party resources often ends up causing bitrot. It's much better to host your own copy of the things your site wants to load. You want to put in some 5 megabyte of javascript trash? Then you should pay the bandwidth costs for that too instead of letting someone else, with different motives, handle it for you. If you don't want to handle updates to that javascript library, if you don't want to pay for the bandwidth, if you want to offload all of that cost to someone else who doesn't care about the end user's privacy and functionality, then you are literally one of the problems with the "modern" web. > An HTML site might want to play a video hosted by a third party, or load a shared resource. There's no situation where I want website A to load resources from website B. Want to share a youtube video? Then give me a youtube link. I don't want to open some potentially malicious third party website which just loads a youtube link. If I don't want to use youtube _at all_ then your site will break too. If I get blocked or banned from youtube, then your site will break anyway and that's still a _you_ problem and not a _me_ problem. It's the same for images, audio, javascript, whatever. If you want it to load on my machine, then you should host it on your machine accessible from the exact same domain that serves your website. > I would argue a browser that doesn't access resources on other domains or IPs is not a web browser at all, but instead some weird sandbox that lets you run a self contained app. That is exactly what we have right now anyway. And I fully agree: browsers have turned into semi-sandboxes to let people run self-contained "apps". I don't want or like it and I argue that that very "model" of the web is not conducive to users' privacy.
- metacritic12 4y ago
  Well I think we mostly agree on the facts and the terms then. That large companies have co-opted a model of early web enthusiasts sharing. That the original and current model of the web has loading third party resources as a primary feature, and is not conducive to user privacy. > and that's still a _you_ problem and not a _me_ problem. Objectively, it's both a server and a client problem if both you and the site intended it to function. I don't see how you not getting a working site isn't a problem for you.

inetknght 4y ago

> Inlinking third party resources is literally in the original hypertext specification of the web. Indeed, but it was intended for real content. That intent is being abused by large corporations for tracking purposes and necessitates a heavy-handed block to it. Plus, including third party resources often ends up causing bitrot. It's much better to host your own copy of the things your site wants to load. You want to put in some 5 megabyte of javascript trash? Then you should pay the bandwidth costs for that too instead of letting someone else, with different motives, handle it for you. If you don't want to handle updates to that javascript library, if you don't want to pay for the bandwidth, if you want to offload all of that cost to someone else who doesn't care about the end user's privacy and functionality, then you are literally one of the problems with the "modern" web. > An HTML site might want to play a video hosted by a third party, or load a shared resource. There's no situation where I want website A to load resources from website B. Want to share a youtube video? Then give me a youtube link. I don't want to open some potentially malicious third party website which just loads a youtube link. If I don't want to use youtube _at all_ then your site will break too. If I get blocked or banned from youtube, then your site will break anyway and that's still a _you_ problem and not a _me_ problem. It's the same for images, audio, javascript, whatever. If you want it to load on my machine, then you should host it on your machine accessible from the exact same domain that serves your website. > I would argue a browser that doesn't access resources on other domains or IPs is not a web browser at all, but instead some weird sandbox that lets you run a self contained app. That is exactly what we have right now anyway. And I fully agree: browsers have turned into semi-sandboxes to let people run self-contained "apps". I don't want or like it and I argue that that very "model" of the web is not conducive to users' privacy.

metacritic12 4y ago
Well I think we mostly agree on the facts and the terms then. That large companies have co-opted a model of early web enthusiasts sharing. That the original and current model of the web has loading third party resources as a primary feature, and is not conducive to user privacy. > and that's still a _you_ problem and not a _me_ problem. Objectively, it's both a server and a client problem if both you and the site intended it to function. I don't see how you not getting a working site isn't a problem for you.