[staltz]

Fun experiment to try out on your computer and experience the gradual death of the internet: block all domains and subdomains from Google and Facebook in your `/etc/hosts` file, and see how many other websites fail to function.

I've had this enabled for 6 months and I've seen websites stopped working as they should, e.g.:

- I couldn't reset my password in Twitch, it required (for some reason) access to both Facebook and Google servers. I had to temporarily disable the `/etc/hosts` restrictions

- National Geographic site fetches jQuery from Google and the website is unusable

- No more embedded Google Maps on many small websites (like local restaurants etc)

- Say bye to embedded YouTube videos

- Many small websites just display in blank

- Other websites have default ugly fonts but are still usable

So Google and Facebook's reach is not just related to direct visits to their sites. Essentially, without them, it's a very crippled internet. The one site that works great is Wikipedia, even its videos are self hosted.

[savanaly]

Analyzing how websites work without Google or Facebook given that they were designed with the availability of those in mind is pointless. To truly see what it would be like without Google or Facebook we'd have to test how websites work without Google or Facebook given that they were designed without the availability of those in mind.

The saying "if things were different, things would be different" comes to mind.

[staltz]

It's not about how websites depend on Google or Facebook being available. It's about the leverage Google and Facebook gain over you through data collection once they are nearly omnipresent on the web, and the growing inability of opting-out from those two tech giants without crippling your basic web experience.

[LambdaComplex]

The point is, they were designed that way. And that's the problem.

[659087]

> - National Geographic site fetches jQuery from Google and the website is unusable

Try this. It might help with a lot of the issues you're experiencing.

https://decentraleyes.org/

[solotronics]

I wonder if anyone has tried decentralized peer to peer CDN for things like JS.. code seems a perfect fit for a distributed CDN.

[dexterdog]

It would help to just cache assets based on the integrity hash. I know there is a potential security hole, but the browser can remember what the sourced URL was and verify/validate. At the very least it gives you a way to handle a downed CDN.