

Something odd happened on Reddit today…
I don’t find that extraordinarily odd at all really. This has been Reddit’s modus operandi for quite a while now. Anything that might pull the curtains back to peep at what/who’s running the show is sternly frowned upon. Usually, they will just shadow ban you which I personally find cowardly. I’d rather you tell me straight out to piss off.
On the topic of browser fingerprinting. I have a more than fair understanding of how it works, however, I am an expert at nothing. What has always struck me as odd is that browser fingerprints change over time, so how do you use a browser fingerprint to source the origin user? Without changing anything, my fingerprint ‘score’ changes daily. Some things that change or affect browser fingerprinting are:
- User-Agent (browser, OS, version)
- Screen resolution & color depth
- Installed fonts
- Plugins & extensions
- Canvas & WebGL rendering
- Timezone & language settings
- HTTP headers (Accept, Do-Not-Track, etc.)
- WebRTC, audio context, hardware info
- Cookies, local storage, caching behavior
About 80% to 90% of all browser fingerprints are unique at any given time. Only 30% to 50% of browser fingerprints change within 1 to 3 months. Users who regularly update, wipe their browser data, or install extensions have the most changes, whereas users who hardly ever update anything, never wipe browser data, or install extensions have the most consistent browser fingerprints that can last months to years. So, in my thinking, a browser fingerprint alone would do little to pinpoint a specific user, if they are regularly maintaining their security envelope. I guess in the case of forensics, a browser fingerprint could be used as a part of complementary evidence.
If they were using a VPN, it could be that their DNS was leaking. However, Reddit usually rejects accounts made with a VPN engaged.
Checking fingerprinting is something I do regularly because I’m very curious. The best I’ve been able to achieve is partial or nearly unique. I also do daily DNS leak tests, which may sound all paranoid, but even with a VPN, and a stand alone pfsense firewall/unbound, and various other obfuscation techniques, VPN IPs change and the IP you had yesterday for a certain locale, might not be the same as today, so it’s worth me taking a minute to check. Not that I have anything to hide. /s
I recommend a daily cleansing with Bleachbit, or Privazer. Schedule task or a cron to run it before shut down.
If someone has expert knowledge of browser fingerprinting, I stand by to be schooled.
30 miles covers a lot of potential users.
What about wired connections? I guess I fail to remember, a lot of people use their phones as a mobile compute platform, which I very rarely do, and certainly not a Reddit app.