Why Gibiru attempts to take the canvas image data of my browser?(gibiru.com)
gibiru.com
Why Gibiru attempts to take the canvas image data of my browser?
https://gibiru.com/
12 comments
Perhaps the link for this post being the homepage of a website that the post title is intimating is a security or privacy risk isn't the greatest idea, regardless of the answer to the question?
Firefox has blocked their attempt for canvas, why would they need it while most sites don’t ?
n.b. Gibiru is a supposedly private search website based on Google..
n.b. Gibiru is a supposedly private search website based on Google..
The obvious answer is browser fingerprinting.
You probably already knew that. I'm just mentioning the missing context of this post.
You probably already knew that. I'm just mentioning the missing context of this post.
Because they want you to use their site with Tor Browser exclusively, obviously.
Browser fingerprinting and/or bot detection (ie. unreasonably small windows.)
Why would a bot have an unreasonably small window?
Because then the bot can hide its activity like the Facebook pixel, where it's visibly difficult for human users to detect (while being relatively easy for computers and scripts to detect).
A lot of security is very similar. A camera in the business that is visibly difficult to detect, yet can easily observe movement within the business.
Its also one of the main areas in online interaction lately. Anything that marks an actor as "not a human".
The security arms race has escalated a lot lately, since now bots can write comparably well and pull a fake image to pass face / camera checks, or similar security. You can't do "find the motorcycle" cause now the bots are better than the humans in many cases. The text recognition failed a long time ago, because it quickly devolved into image tests that humans are not actually that great at.
Unfortunately, the entire problem (in my opinion), is trying to secure a realm for human use, that humans are not native to. Computers are better at being computers.
A lot of security is very similar. A camera in the business that is visibly difficult to detect, yet can easily observe movement within the business.
Its also one of the main areas in online interaction lately. Anything that marks an actor as "not a human".
The security arms race has escalated a lot lately, since now bots can write comparably well and pull a fake image to pass face / camera checks, or similar security. You can't do "find the motorcycle" cause now the bots are better than the humans in many cases. The text recognition failed a long time ago, because it quickly devolved into image tests that humans are not actually that great at.
Unfortunately, the entire problem (in my opinion), is trying to secure a realm for human use, that humans are not native to. Computers are better at being computers.
conserving resources, by minimizing things it doesnt need.
just what size is unreasonable comes to mind.
i run with 3 displays one of them is tiled with browser windows. -that makes me curious about how many i can tile up until threshold size and the windows start triggering false +ives
just what size is unreasonable comes to mind.
i run with 3 displays one of them is tiled with browser windows. -that makes me curious about how many i can tile up until threshold size and the windows start triggering false +ives
Because it's running in a $4/month instance w/512MB memory and is rendering its Chrome window in Xvfb, which has been allocated a 640x480 frame buffer.
In one of my projects I used it to detect if browser supports blur filter and if not fall back on stack box blur.