3 ms·
Gee, they could have sold those for a lot of money. Well done researchers, you have a strong moral compass.
by contingencies 13y ago
Gee, they could have sold those for a lot of money. Well done researchers, you have a strong moral compass.
- dguido 13y agoYes, all of those vulnerabilities in non-privileged applications like 'ls' are worth millionssssss.
- jerf 13y agoYes, and then they combine them with a kernel exploit and, uh, suddenly your "unprivileged process" turns out to not be so unprivileged after all. Resist the temptation to marginalize security bugs. They don't exist in isolation.
- tptacek 13y agoYou're right, but so is the parent commenter; there is a popular meme that security bugs in general are worth large amounts of money, but in reality only a small subset of bugs command real money.
- lawnchair_larry 13y agoYou could be a better ambassador of our industry by explaining the misconception instead of being condescending. And we wonder why developers think we are just arrogant jerks.
- dguido 13y agoYou're welcome to step in and explain, but I think it's pretty obvious why local vulnerabilities in non-setuid binaries aren't exactly "critical."
- betterunix 13y agohttps://xkcd.com/1200/ https://xkcd.com/1200/
- dguido 13y agoThat would only be accurate if you could trigger one of these bugs remotely.
- betterunix 13y ago...or if you can trick the user into opening a malicious file. An exploit in ls? Send them a zip, tarball, etc. with a maliciously named file.
- contingencies 13y agoIMHO anyone who contributes to the state of the art with regards to public security tooling is doing the world a favour. Even if the majority of these applications are not well privileged, and therefore of dubious value, applying the same technical knowledge elsewhere could earn them cash. I still applaud their efforts.
- bradleyland 13y agoThe value of a vulnerability is linked to its usefulness in accomplishing the goals held by those paying for the exploit. Based on that information, it should become obvious that the only bugs worth large sums of money are bugs that result in escalated privileges on the system or arbitrary code execution. Bugs that crash services also have value for DDoS attacks, but they're not as valuable as the two previously mentioned types. While 1.2k bugs sounds like a lot, we can't make an inference about the severity of the bugs based solely on the number of them. Also keep in mind that the Debian repository contains thousands of packages. From the Debian 7.0 Wheezy release page: "* more than 36,000 other ready-to-use software packages, built from nearly 17,500 source packages" http://www.debian.org/News/2013/20130504 http://www.debian.org/News/2013/20130504 The best we can infer that given 1.2k bugs, there is some statistical probability that any number of them may be severe. Without studying the data, we can't say what that probability might be though.