It's horrifying to see that our browsers give up anything at all. I feel like we need a new model where we just get served the content, and we don't serve up anything to the content provider. I feel ill.
How would you prevent that? If you want to use a feature, you need to know if that feature is available. So you add methods/functions to ensure that you can fail gracefully when something is not available. That availability is the fingerprint. I don't blame the browsers as much as asshats using the tool for malicious purposes. We're not going to ban hammers because someone uses them for things other than driving nails.
Another one is "draw text, read back pixels" which is simultaneously a really effective fingerprint, and also cuts to the core of what we want a web browser to do. We can't refuse to draw text.
And what about SWE from outside whatever jurisdiction that body issues those licenses? You cannot regulate your way out of tech problem if everyone is not subject to said regulations. Why people keep making comments like this is beyond me.
In the old days, we just sent the content and let the client figure out what to do with it. If you navigated to homestarrunner.com and you didn't have flash installed, you just got a message saying to install it, or worst case, a broken icon in a frame.
Right, but the client side did a query to see what was available and gracefully displayed a message. People are now sending the truthy response of the feature's availability back. You could be doing that as a legit bit of logging, or you could be doing that to be an asshat and deanonymize the user. The browser doesn't know the intent of being asked about a feature. And yes, it has been figured out just fine as a way for making someone money by selling information about you
Cool tool, but the language is painfully characteristic of AI. Maybe we are collectively getting over caring about that, but if not, it's worth a pass of "make this sound a little less like AI."
The APIs it's relying on to show you that information do implement some anti-fingerprinting mechanisms - things like available memory will be put into rough buckets rather than returning the exact amount you have.
It's worth noting that you need both uniqueness and some form of stability. If you consistently show up as a diffent fingerprint every time you visit or for every different site, then that is a form of privacy as well.
I just downloaded a fresh install of LibreWolf to test this and it did indeed have a stable UUID for me at 1/62E6. Better than the 1/680E6 from regular Firefox, but still upsetting. Fonts and Audio fingerprinting seems to be where a lot of the entropy is coming from... Even after I enabled some of the flags to resist font enumeration it still is stable and high entropy. I'm not sure what to do to be honest about this, it is very discouraging.
Fun fact: some malware won't execute if it sees a Russian/Ukrainian keyboard layout, because the authors avoid hitting victims in their own jurisdiction.
Not usually, if you're comparing it to IPv4. Many ISPs give each customer a persistent IPv4 address that doesn't tend to change over time. Because most people use NAT (internal IP address != external IP address), a server you connect to can't identify the specific computer or device.
With IPv6, it's true that in many cases servers will see you as connecting from the specific IPv6 address of your computer/device. However, IPv6 addresses contain many parts, like:
2001:db8:1234:5600 : abcd:1234:5678:9abc
where the left part is basically the equivalent of the unchanging IPv4 address you got from the ISP, and the right part identifies your specific device. But in most OSes the right part is randomly generated and changes on some cadence (e.g. daily).
So in many cases, the ability for a server to individually fingerprint your device solely based on the IP address is roughly equivalent between IPv4 and IPv6. (This isn't the case if the ISP does CGNAT on IPv4, where you're sharing your IPv4 address with many customers - this would be similar in magnitude to a VPN in terms of the ability for a server to fingerprint you based on IP.)
Fascinating visualization , as someone building productivity extensions, it is always sobering to see exactly how much surface area the browser exposes.
I built this after a thread here about Alibaba using an audio-context trick to
fingerprint visitors. I knew a fair number of fingerprinting methods but not that
one, and I wanted to see all of them in one place, running against my own browser.
GlassBox runs ~31 probes (canvas, WebGL/WebGPU, audio, fonts, the WASM feature
set, math/engine quirks, WebRTC IP, timezone/locale, the permission and API
matrices, an incognito heuristic, cross-site login-state, and so on) and shows the
raw values plus an estimate of how identifiable you are.
A few deliberate choices:
- One static HTML file, no dependencies, no build step. Everything runs
client-side and nothing is sent, with one opt-out exception: IP geolocation,
which calls a public API. I didn't want a privacy tool that phones home.
- The "identifiability" number is an honest model, not a measurement. It sums
published per-signal entropy (Panopticlick / AmIUnique / Cover Your Tracks),
discounts signals your browser masks, and caps at the ~33 bits needed to single
out one person on Earth. A no-server tool can't compute true rarity against a
live population, so I label it an estimate instead of pretending. For real
population numbers, Cover Your Tracks and AmIUnique have the datasets.
- There's a companion guide on lowering your fingerprint, with the caveat that
uniqueness isn't privacy: blending into a big crowd (Tor at its default size)
beats a bespoke hardened setup that makes you the only one who looks like that.
Also see https://coveryourtracks.eff.org/
It's horrifying to see that our browsers give up anything at all. I feel like we need a new model where we just get served the content, and we don't serve up anything to the content provider. I feel ill.
How would you prevent that? If you want to use a feature, you need to know if that feature is available. So you add methods/functions to ensure that you can fail gracefully when something is not available. That availability is the fingerprint. I don't blame the browsers as much as asshats using the tool for malicious purposes. We're not going to ban hammers because someone uses them for things other than driving nails.
Another one is "draw text, read back pixels" which is simultaneously a really effective fingerprint, and also cuts to the core of what we want a web browser to do. We can't refuse to draw text.
need to license swe just like real engineers and make them have a responsibility to the public
And what about SWE from outside whatever jurisdiction that body issues those licenses? You cannot regulate your way out of tech problem if everyone is not subject to said regulations. Why people keep making comments like this is beyond me.
In the old days, we just sent the content and let the client figure out what to do with it. If you navigated to homestarrunner.com and you didn't have flash installed, you just got a message saying to install it, or worst case, a broken icon in a frame.
We figured it out just fine.
Right, but the client side did a query to see what was available and gracefully displayed a message. People are now sending the truthy response of the feature's availability back. You could be doing that as a legit bit of logging, or you could be doing that to be an asshat and deanonymize the user. The browser doesn't know the intent of being asked about a feature. And yes, it has been figured out just fine as a way for making someone money by selling information about you
Pretty sure someone else also promoted Claude to make something like this and posted it a few weeks back.
Recently discussed prompts:
https://news.ycombinator.com/item?id=48727059 [flagged] What any website can see about you
https://news.ycombinator.com/item?id=48062178 A web page that shows you everything the browser told it without asking
https://news.ycombinator.com/item?id=49182352 Show HN: Modern Browsers Don't Need the Cookie Anymore
https://news.ycombinator.com/item?id=46517654 [flagged] Show HN: DevicePrint – device fingerprinting without cookies
https://news.ycombinator.com/item?id=46037031 Show HN: Explore what the browser exposes about you
https://news.ycombinator.com/item?id=45169286 Browser Fingerprint Detector
https://news.ycombinator.com/item?id=49359153 A live browser fingerprinting and bot detection playground
https://news.ycombinator.com/item?id=46674460 Show HN: Anti-Detect.com – Browser Fingerprint Scanner
There are definitely more! Could this be HN's favourite prompt?
Haha! Well done for putting in the effort to find those. I was sure I’d seen it more than once but didn’t want to overrreach!
The trick is that in most cases, someone links https://coveryourtracks.eff.org/
Cool tool, but the language is painfully characteristic of AI. Maybe we are collectively getting over caring about that, but if not, it's worth a pass of "make this sound a little less like AI."
E.g., the guide page ends with
> The honest bottom line. Perfect anonymity [...]
Not just the language. Its like all LLM front end design is working off one of three templates.
* You're absolutely right! ;-)
This doesn’t seem right. 1 in 6.2 billion for Firefox on an iphone, not including ip/network uniqueness?
Seems more plausible that anti-fingerprinting is throwing it off?
Some of the data seems straight-up wrong.
I know this computer has more RAM than reported and it's definitely not on a 4G network either.
The APIs it's relying on to show you that information do implement some anti-fingerprinting mechanisms - things like available memory will be put into rough buckets rather than returning the exact amount you have.
It got facebook pixel wrong. I see zero DNS requests to facebook.com on NextDNS
Yeah seems like it doesn't account for anti-fingerprinting at all.
It's worth noting that you need both uniqueness and some form of stability. If you consistently show up as a diffent fingerprint every time you visit or for every different site, then that is a form of privacy as well.
I just downloaded a fresh install of LibreWolf to test this and it did indeed have a stable UUID for me at 1/62E6. Better than the 1/680E6 from regular Firefox, but still upsetting. Fonts and Audio fingerprinting seems to be where a lot of the entropy is coming from... Even after I enabled some of the flags to resist font enumeration it still is stable and high entropy. I'm not sure what to do to be honest about this, it is very discouraging.
Shocking, this site assigned me a UUID and I found it listed here! https://everyuuid.com/
Interesting work! Would randomnize the least frequent used fonts, apis and settings be an way to anti-fingerprinting?
Fun fact: some malware won't execute if it sees a Russian/Ukrainian keyboard layout, because the authors avoid hitting victims in their own jurisdiction.
https://krebsonsecurity.com/2021/05/try-this-one-weird-trick...
If your ISP issues an IPv6 address, isn't that pretty much game over for anonymity/uniqueness?
Not usually, if you're comparing it to IPv4. Many ISPs give each customer a persistent IPv4 address that doesn't tend to change over time. Because most people use NAT (internal IP address != external IP address), a server you connect to can't identify the specific computer or device.
With IPv6, it's true that in many cases servers will see you as connecting from the specific IPv6 address of your computer/device. However, IPv6 addresses contain many parts, like:
where the left part is basically the equivalent of the unchanging IPv4 address you got from the ISP, and the right part identifies your specific device. But in most OSes the right part is randomly generated and changes on some cadence (e.g. daily).So in many cases, the ability for a server to individually fingerprint your device solely based on the IP address is roughly equivalent between IPv4 and IPv6. (This isn't the case if the ISP does CGNAT on IPv4, where you're sharing your IPv4 address with many customers - this would be similar in magnitude to a VPN in terms of the ability for a server to fingerprint you based on IP.)
A VPN or Tor swaps it out entirely, which your browser fingerprint survives. It is a strong identifier, not a permanent one.
This does not account for anti fingerprinting. Also, are there any new tricks that are not in fingerprintjs?
Fascinating visualization , as someone building productivity extensions, it is always sobering to see exactly how much surface area the browser exposes.
The "Hardware & Environment" part which supposedly links me across different browsers doesn't work — it differs between Safari, Firefox and Chromium.
In a private Safari window, it's not even stable across reloads.
I built this after a thread here about Alibaba using an audio-context trick to fingerprint visitors. I knew a fair number of fingerprinting methods but not that one, and I wanted to see all of them in one place, running against my own browser.
GlassBox runs ~31 probes (canvas, WebGL/WebGPU, audio, fonts, the WASM feature set, math/engine quirks, WebRTC IP, timezone/locale, the permission and API matrices, an incognito heuristic, cross-site login-state, and so on) and shows the raw values plus an estimate of how identifiable you are.
A few deliberate choices:
- One static HTML file, no dependencies, no build step. Everything runs client-side and nothing is sent, with one opt-out exception: IP geolocation, which calls a public API. I didn't want a privacy tool that phones home.
- The "identifiability" number is an honest model, not a measurement. It sums published per-signal entropy (Panopticlick / AmIUnique / Cover Your Tracks), discounts signals your browser masks, and caps at the ~33 bits needed to single out one person on Earth. A no-server tool can't compute true rarity against a live population, so I label it an estimate instead of pretending. For real population numbers, Cover Your Tracks and AmIUnique have the datasets.
- There's a companion guide on lowering your fingerprint, with the caveat that uniqueness isn't privacy: blending into a big crowd (Tor at its default size) beats a bespoke hardened setup that makes you the only one who looks like that.
Source (MIT): https://github.com/HotStartLabs/glassbox
I'd genuinely like to know which vectors I'm missing, especially from the anti-fraud / detection side.