3 ms·
I realize this is convenient for people running Chrome on OS X but... The way I surf is simple: Chrome is installed from the .tgz and certainly not from the rp
by martinced 14y ago
I realize this is convenient for people running Chrome on OS X but...
The way I surf is simple: Chrome is installed from the .tgz and certainly not from the rpm or deb files (simply put: I don't need to be root to install Chrome and Chrome is confined to a single user account).
The account I use for surfing is separated from the account from which I use SSH to access SSH servers.
Should someone "root" Chrome while I surf the Web, he'd still need to find what is called a "local root exploit" to access what's in the other user accounts (for example, but not limited to, SSH keys).
I think there's no way that, for me, moving SSH inside the browser would provide me more security than having Chrome not doing SSH and being in a separate (throwaway) user account. It seems to be, once again, trading conveniency in the name of security. I don't doubt it fits Google's plan to make Chrome the OS that said and so they'll of course tell you: "Nothing to see here, move along".
Another concerning thing, depending on how it's done, it's that the potential to mount a "mocking bird" attack may be strong on this one. Mocking birds attack tend to not work well when you use MagicSysRQ to emit a keypress bypassing X, sending you to a text console, and doing your SSH magic.
I realize security is a pain, but it would be great if people could still criticize quite critical security issues on HN without having everyone downplaying the ones criticizing as "redditers" (is that an insult?), contrarians, etc.
I love HN. I love entrepreneurship. But I do love Bruce Schneier too and I consider security to be something very important.
Yet here everytime an app that does "one master online password to store all your passwords" gets upvoted like mad just because it's a "startup" and just because it "runs on a smartphone".
I understand that some of you may not like the "tone" of people using sarcasm to make a point but... They're still making a point and, very often, it's a valid point.
So it would be great if people could come up with better arguments than simply name-calling the ones pointing out the security implications of apps / plugins / etc.
- arcatek 14y agoI don't see how installing from an archive is safer than using a package, can you develop ? From my understanding, Chrome is obviously not chrooted, so a local root exploit will be required in either case to access to the other accounts (since it will run with your current privileges). And actually it seems easier to corrupt a chrome extracted in the home directory, since an intruder would not require root credentials to inject some code into your chrome (and replacing every bank front page with fishing pages, for example).
- throwaway54-762 14y agoRPMs / dpkgs require root to install to the machine-global database. Both package formats have mechanisms called "triggers" which allow the package to run arbitrary shell script at install or uninstall time. Since install/uninstall are run as root, this means arbitrary root code execution on install. Vs: a non-root user extracting a tgz and then running some file, root is never involved. I'm not in agreement with OP about this being a concern worth his or her mitigation strategy, but the logic is sound.