That ai header image is a real turnoff to anything they have to say

Really nice article

By spoofing the fingerprint, developers can make their automated tools impersonate real users more convincingly, thereby bypassing bot detections.

Many OSS projects and personal web servers have bot detection because they would otherwise drown under (AI) scrappers and other bots traffic. Hosting or bandwidth cost is often unsustainable without bot protection.

If you don't want to kill these projects, honor robots.txt by default, use throttling, don't try to circumvent bot blocks. Look if there's a purpose built API available to bots. If they don't want to offert such API, go find something else to do.

I removed the Firefox' version number and half the internet broke. I replaced the whole string with Dillo 3/2.0 and this fixed most sites that "don't work" without JS.

Well, this was before the anti-scraping all-captcha now.

What works for me is to use Firefox with tracking protection set to strict (or Librewolf with resistFingerprinting disabled and WebGL enabled), then install Jshelter and set it like below:

• Time precision: High
• Locally generated images: Little lies
• Locally generated audio: Little lies
• Graphic card information: Little lies for maximum privay or disabled to minimise captchas/shadowbans WebAssembly speed-up: Enabled
• Everything else disabled as Firefox/Librewolf takes care of them

Once you do that, FingerprintJS Pro (fingerprint.com) should give you a different ID every time you clear cookies and change the IP.

Good to see the technical reasons I can't just get a website with curl.

I have to yet to find any fingerprint-evading solution that works on Linux and can actually beat creepjs reliably... besides disabling javascript completely, which puts you in a whole new (much, much smaller) set of people that can still be fingerprinted with non-JS solutions including html/css/header methods as well as TLS fingerprinting like JA3/JA4.