What the Hack Day 2.5
I started the morning with a talk about reverse-engineering XDA devices with WinCE. I don't own a handheld or a smartphone, but I don't want to see the engineering group that decided that it'd be a good idea to implement an AT command access arbitrary memory regions. This includes the position to unlock the devices safety mode as well...
The noon talk was excellent; a group of engineers implemented a 500km wifi link to bring internet connectivity to rural Peru. They're operating on some wave bands that get reflected in the Ionosphere layer and successfully implemented a stable 3 kbps link.
Later on I listened to a talk from some kernel security guy that evaluated attacks on /proc/pid/mem, which exposes adress segments into user space. The attack on Linux is somewhat theoretical, as the attacker would need to call execve() 2^32 times, but it was very interesting nonetheless.
The next talk about "Client security" was not too good and I left earlier for talk about forging fingerprints to fool biometric scanners. While fingerprints are publically conceived as secure measures, they've actually proven to be rather weak. The lecturer was able to make a working copy in abou ten minutes, really impressive. The followup talk about defeating biometric systems was hampered by the referent's extreme nerviosity and insufficient English, so I left early to drink Flens at the tents. Unfortunately iit started to rain again, but the evening
talks look promising.
The noon talk was excellent; a group of engineers implemented a 500km wifi link to bring internet connectivity to rural Peru. They're operating on some wave bands that get reflected in the Ionosphere layer and successfully implemented a stable 3 kbps link.
Later on I listened to a talk from some kernel security guy that evaluated attacks on /proc/pid/mem, which exposes adress segments into user space. The attack on Linux is somewhat theoretical, as the attacker would need to call execve() 2^32 times, but it was very interesting nonetheless.
The next talk about "Client security" was not too good and I left earlier for talk about forging fingerprints to fool biometric scanners. While fingerprints are publically conceived as secure measures, they've actually proven to be rather weak. The lecturer was able to make a working copy in abou ten minutes, really impressive. The followup talk about defeating biometric systems was hampered by the referent's extreme nerviosity and insufficient English, so I left early to drink Flens at the tents. Unfortunately iit started to rain again, but the evening
talks look promising.
posted by Moritz Muehlenhoff at 7:38 PM
<< Home