3 ms·

What I would like to see is some kind of "white-list" browser - I think it's our best chance to keep an ad-free environment. Maybe it is a bad idea in general t

What I would like to see is some kind of "white-list" browser - I think it's our best chance to keep an ad-free environment. Maybe it is a bad idea in general to allow a remote page write pixels to your screen: images, videos, text and now applications It is nice that the web is open (at least reasonably), but maybe raw JS, HTML and CSS is too general? The most valuable webpages have simple designs and are liked because of their content, like - wikipedia.org - HN - stackoverflow.com - maps.google.com - ... These pages could implement - Knowledge-Base - Topic-Comment-Community - Answer-Response-Community - Interactive-GIS specifications, that obviously do not exist. E.g. if a newspaper implements a hypothetical "Newspaper-Specification" it would have to provide original, sourced content containing text, images, audio and video via some API (which is often already the case when single page apps fetch content client-side via e.g. JSON). There would be no visual difference between reading "The Guardian" vs. "USA Today" or browsing "OpenStreetMap" vs "Google Maps" as content would be rendered on the client.

krapp 8y ago
>Maybe it is a bad idea in general to allow a remote page write pixels to your screen: images, videos, text and now applications Computers have been doing this since the 60s, at least, in some form or another. I don't think you want to live in a world where the very idea of networks is considered harmful. Where all of the software you consume has to be loaded from physical media and the internet and any network connected device just doesn't exist. You couldn't even have cellphones. We don't need a "whitelist" browser - you simply choosing not to visit sites you don't want to visit is already a sufficient whitelist. Having "specifications" that websites and authors would have to enforce based on content would make the web less free. Websites are software. Software authors have the right and the freedom to write whatever they like, any way they choose, up to and including the UI. The Guardian being able to make their own decisions about how to implement the web versions of a "newspaper" - including design choices which might differentiate them from USA Today, is part and parcel the same sort of freedom that any author of any work, software, literary, music, what have you, enjoys. That HTML, CSS and JS don't step over the line from describing content to making editorial decisions about content is a feature, not a bug.
- protomikron 8y ago
  Maybe I have expressed myself unclear, but I don't want to advocate against networked software and loading resources over a network in general. I would like to have a more "typed" API to networking services (including web-pages) that provides sufficient information to render the application on my client - the term "whitelist" may indeed be a sub-optimal term. > Having "specifications" that websites and authors would have to enforce based on content would make the web less free. We already have enforced specifications (HTML, HTTP, TCP, ...) and they did make the web free, as you could and can participate, when you implement these specifications (at least to a sufficient degree). However we could go further and provide specifications for more high-level tasks. Consider the problem of going from A to B via public transport. Cities with a working public transport infrastructure normally provide a webpage that allows you to search for routes and reporting journey time, cost and additional information. However these webpages are more or less convenient to use and every transportation provider has to implement its own shitty webpage. Hypothetically it could implement an API that provides this information and the "Browser" has its builtin GUI (configurable by the user) to access this information. > Software authors have the right and the freedom to write whatever they like, any way they choose, up to and including the UI. I am more talking about users who should have the possibility to access the information they seek in an uncluttered way. And we can keep stuff like <canvas>, so you are still able to draw your pixels.
onion2k 8y ago
There would be no visual difference between reading "The Guardian" vs. "USA Today" or browsing "OpenStreetMap" vs "Google Maps" as content would be rendered on the client. You're describing Reader Mode, which most browsers already have.

krapp 8y ago

>Maybe it is a bad idea in general to allow a remote page write pixels to your screen: images, videos, text and now applications Computers have been doing this since the 60s, at least, in some form or another. I don't think you want to live in a world where the very idea of networks is considered harmful. Where all of the software you consume has to be loaded from physical media and the internet and any network connected device just doesn't exist. You couldn't even have cellphones. We don't need a "whitelist" browser - you simply choosing not to visit sites you don't want to visit is already a sufficient whitelist. Having "specifications" that websites and authors would have to enforce based on content would make the web less free. Websites are software. Software authors have the right and the freedom to write whatever they like, any way they choose, up to and including the UI. The Guardian being able to make their own decisions about how to implement the web versions of a "newspaper" - including design choices which might differentiate them from USA Today, is part and parcel the same sort of freedom that any author of any work, software, literary, music, what have you, enjoys. That HTML, CSS and JS don't step over the line from describing content to making editorial decisions about content is a feature, not a bug.

protomikron 8y ago
Maybe I have expressed myself unclear, but I don't want to advocate against networked software and loading resources over a network in general. I would like to have a more "typed" API to networking services (including web-pages) that provides sufficient information to render the application on my client - the term "whitelist" may indeed be a sub-optimal term. > Having "specifications" that websites and authors would have to enforce based on content would make the web less free. We already have enforced specifications (HTML, HTTP, TCP, ...) and they did make the web free, as you could and can participate, when you implement these specifications (at least to a sufficient degree). However we could go further and provide specifications for more high-level tasks. Consider the problem of going from A to B via public transport. Cities with a working public transport infrastructure normally provide a webpage that allows you to search for routes and reporting journey time, cost and additional information. However these webpages are more or less convenient to use and every transportation provider has to implement its own shitty webpage. Hypothetically it could implement an API that provides this information and the "Browser" has its builtin GUI (configurable by the user) to access this information. > Software authors have the right and the freedom to write whatever they like, any way they choose, up to and including the UI. I am more talking about users who should have the possibility to access the information they seek in an uncluttered way. And we can keep stuff like <canvas>, so you are still able to draw your pixels.

onion2k 8y ago

There would be no visual difference between reading "The Guardian" vs. "USA Today" or browsing "OpenStreetMap" vs "Google Maps" as content would be rendered on the client. You're describing Reader Mode, which most browsers already have.