This [0] is a great set of configurations for the Firefox browser that remove many fingerprinting vectors. It turns on several fingerprint-homogenization features that the TBB project upstreamed into Firefox. Check it out!
(By using this, G--gle will hound you with impossible capchas and privacy-forward search engines will think that you're a bot.)
Most of these fingerprinting vectors are from the browser and the scourge of JS. I wonder about the footprint left by a user who doesn't use a browser, or instead uses some kind of parsing client that just fetches HTTP or data from API endpoint. Aside from the IP address, there are other ways to fingerprint a user based on network requests from various protocol leaks, many are presented here [1][2]. Are there any leak vectors missing from this list?
>By using this, G--gle will hound you with impossible capchas and privacy-forward search engines will think that you're a bot
A friend of mine had an alternative theory: by making it very very hard to track you, you have shown to be somewhat intelligent and you have shown that you know more about computers than the average Joe. Goolag, knowing you're a somewhat intelligent human, gives you the very hard captchas it needs to train.
If Google has a way to verify correctness of an answer, which they seemingly would have to - then would it matter? Smart person or smart bot - if they are given a tough question and answer correctly Google gains confidence in the answer _(or however it works haha)_.
Because really i don't think Google cares at all about bots. They just want data. And the captcha system is an impressive system to pull training data out of intelligent beings/code.
It's a tough call. You're right that blending in is better but also turning off JS cuts off access to a lot of fingerprinting vectors. That lack of extra information may result in you looking like many others.
(By using this, G--gle will hound you with impossible capchas and privacy-forward search engines will think that you're a bot.)
Most of these fingerprinting vectors are from the browser and the scourge of JS. I wonder about the footprint left by a user who doesn't use a browser, or instead uses some kind of parsing client that just fetches HTTP or data from API endpoint. Aside from the IP address, there are other ways to fingerprint a user based on network requests from various protocol leaks, many are presented here [1][2]. Are there any leak vectors missing from this list?
[0]: https://github.com/pyllyukko/user.js
[1]: https://www.whonix.org/wiki/Protocol-Leak-Protection_and_Fin...
[2]: https://www.whonix.org/wiki/Data_Collection_Techniques