Melancholy wrote:
malmon73 wrote:
RX14 wrote:
Hexd wrote:
I don't understand how you guys would install this on your computer.
Because it's not a keylogger? it logs statistics about your keys instead of the keys themselves lol.
To be fair, it could be a keylogger. It sends pulse data encrypted, so we can't really tell what's being sent. I guess one way to check if it's behaving like a keylogger would be to write an absolute fuckload, and see if the encrypted data length is significantly larger.
if they were storing more than just keypresses and stuff, they'd need massive servers considering they have like over 20k people actively submitting data; some of these people with around 200m keypresses.
Just ran a quick calculation: A standard US QWERTY keyboard usually has a maximum of 107 different keys; a keypress could be transmitted as a sequence of 7-bit binary numbers.
7 bits, times 200 million, gives us one-billion-and-four-hundred-million bits, which, with a little help from Google, translates into... .175 gigabytes. This, times its 20,000 active users, gives us 3,500 gigabytes; otherwise known as a measly 3.5 terabytes. Coincidentally, that's actually just a little bit more than the combined storage space of all my external hard disks.
So, yeah: Assuming an algorithm combed through the data at the end of the day and purged whatever wasn't deemed useful, I'd say it's completely possible for WhatPulse to be a keylogger.
malmon73 wrote:
Melancholy wrote:
malmon73 wrote:
RX14 wrote:
Because it's not a keylogger? it logs statistics about your keys instead of the keys themselves lol.
To be fair, it could be a keylogger. It sends pulse data encrypted, so we can't really tell what's being sent. I guess one way to check if it's behaving like a keylogger would be to write an absolute fuckload, and see if the encrypted data length is significantly larger.
if they were storing more than just keypresses and stuff, they'd need massive servers considering they have like over 20k people actively submitting data; some of these people with around 200m keypresses.
I guess they could probably filter it down to keypresses that could be usernames and passwords, e.g. "<something> TAB <something> ENTER"
Malmon has the right idea in determining this, however: Simply run a script that simulates the rapid pressing of many keys, and compare it with a control.
If they've got time, someone should totally check this out!
