What advantage memory safety

mzman

Can someone explain how memory safe coding helps make 1Password better able to protect secrets in memory? I don't want to email Support about this. I'd like to get a better understanding through participation in a public forum.

Please don't close this new thread. You have other threads asking supplemental questions indirectly related to your recent KB entry on related topics. I'm trying to participate in the same way.

AGAlumB

I don't think there's a benefit to further repetition. This has already been discussed at length, and you were a participant in those discussions. As mentioned previously, you can email us if you have specific questions, but whether or not you take us up on that offer is entirely up to you. Take care.

AGAlumB

I'm sorry. I wasn't under the impression that you were legitimately asking, since you yourself touched on this previously:

Re-code in a language that will result in more vulnerabilities and crashes. Do whatever it takes, because the vulnerability here is the greatest imaginable one.

You seemed to be suggesting that we knowingly eschew the benefits of using memory-safe languages in spite of those risks, because you believe the current risk (that someone else in control of your PC can read memory) to be greater. And, as I myself mentioned earlier in that discussion:

giving up memory safety has its own risks. When a game crashes, it's bad for the experience and can allow for exploitation. When security software crashes, there's a lot more at stake.

As did Goldberg in his very first reply. I'm not even going to quote from it, because pretty much the whole thing was about this. But the linked article about how 70% of Microsoft security bugs over the past 12 years have been memory safety bugs is a good jumping off point, I'd say.

So I thought we were on the same page with this already.

Ultimately, expounding on the concept of memory safety is really well beyond the scope of 1Password, as it applies to software in general. There are plenty of resources available out there if you're interested to learn more about that or any other computer science concepts, and I think you'd get a much better understanding of this concept from people who teach the subject for a living. :)

Ben

In short, because "every crash is a potential security vulnerability" - Raymond Chen, Microsoft. Just because someone hasn't figured out how to exploit a crash yet doesn't mean it isn't exploitable. As such crashes are treated as high priorities, and every opportunity to avoid crashes (such as using memory-safe languages) are taken advantage of. Consider that reading secrets from memory isn't the only possible threat. I appreciate your passion for this situation but we're not going to stop using memory-safe languages because of it (as Jeff says: the cure is worse than the disease).

Our security team may be willing/able to discuss further if you'd like to reach out to them directly, at support+security@1password.com, but there isn't anything further any of us are going to be able to contribute here in the forum.

Ben

Ben

You're very welcome. :+1:

Ben