4 ms·

Let’s not let better be the enemy of good either. Better than terrible is still bad and is nowhere near good. It is frankly ridiculous that anybody should beli

by Veserv 4y ago

Let’s not let better be the enemy of good either. Better than terrible is still bad and is nowhere near good. It is frankly ridiculous that anybody should believe Apple when they claim to provide even minimal resistance to well-funded determined attackers. Protecting against well-funded determined attackers has been the holy grail of software security since forever and everybody in software security at least claims to be working toward that. Despite that, the prevailing state of “best-in-class” “best-practices” commercial software security is objectively terrible including Apple circa 1 year ago. Are we supposed to believe that Apple, despite abject failure over the last few decades until as recently as the last time they announced security updates to the iPhone, has finally this time, for sure, pinky swear its true, jumped from terrible to the holy grail, or even good, because they said so? No, this is absolute, utter, unequivocal garbage. Their claims are completely unsupported and they should be excoriated for spewing unsubstantiated bullshit that muddies the waters of the actual state of software security and misleads people into believing they are getting a meaningful degree of protection or software security. If they want to make such claims, they should put their money where there mouth is and, instead of certifying iOS to EAL1+ and AVA_VAN.1 as they currently do, they should certify it in “Lockdown Mode” to EAL6-7 and AVA_VAN.5 which actually does certify protection against “high attack potential” attackers such as large organized crime and state-sponsored attackers. At the very least they could certify it to EAL5 and AVA_VAN.4 which certifies protection against “moderate attack potential” attackers. Until they do that, their claims to protect against state-sponsored attackers are complete unverifiable bullshit.

donw 4y ago
Especially as Apple is often the "well-funded attacker".
- Bud 4y ago
  Citation, please?
saagarjha 4y ago
First off, calm down. This feature came out today. It's not really clear yet how well it will fare. Second, this feature is a step in the direction of Apple accepting that defending against a well-funded attacker is difficult when providing general-purpose software, so this is still a step in the right direction.
- Veserv 4y ago
  It came out today, which means it should be assumed insecure against state sponsored actors until proven otherwise with overwhelming evidence, not we should give it the benefit of the doubt because maybe they really did it the 57th time after 56 total failures. For that matter, it is not like they could not provide such evidence even though it came out today. It has presumably been in development for some time, so if they did actually provide verifiable protection against state sponsored attackers they could just release their formal proofs of security to that effect and be done with it or at least preliminary certification evidence demonstrating protection against high attack potential attackers as outlined in the international Common Criteria standard via AVA_VAN.5. iOS is already certified according to Common Criteria as their only advertised security certification, just at the lowest possible level, and it already has a certification for high attack potential attackers, so doing this would be consistent with their existing certification regime and provide clear evidence supporting their claims. Absent that, I see no independent verifiable evidence of any of their claims, endless precedent to dispute their claims, and not even a token effort to provide even a sliver of objective backing for their claims. So why should I or anybody else reject the standard wisdom of “you are screwed if state sponsored attackers are interested in you and there is no product that can help you” and instead believe Apple’s marketing that they can?
  - saagarjha 4y ago
    Security is not black-and-white, it's shades of gray. This feature aims to make exploitation harder. Formal proofs and certifications are nice but what I just said remains true even in the face of such things. iOS is regularly tested in the real world against highly resourceful attackers, and the results there are far more indicative of how well its security fares than anything else could be.

by Veserv 4y ago

donw 4y ago

Especially as Apple is often the "well-funded attacker".

Bud 4y ago
Citation, please?

saagarjha 4y ago

First off, calm down. This feature came out today. It's not really clear yet how well it will fare. Second, this feature is a step in the direction of Apple accepting that defending against a well-funded attacker is difficult when providing general-purpose software, so this is still a step in the right direction.

Veserv 4y ago
It came out today, which means it should be assumed insecure against state sponsored actors until proven otherwise with overwhelming evidence, not we should give it the benefit of the doubt because maybe they really did it the 57th time after 56 total failures. For that matter, it is not like they could not provide such evidence even though it came out today. It has presumably been in development for some time, so if they did actually provide verifiable protection against state sponsored attackers they could just release their formal proofs of security to that effect and be done with it or at least preliminary certification evidence demonstrating protection against high attack potential attackers as outlined in the international Common Criteria standard via AVA_VAN.5. iOS is already certified according to Common Criteria as their only advertised security certification, just at the lowest possible level, and it already has a certification for high attack potential attackers, so doing this would be consistent with their existing certification regime and provide clear evidence supporting their claims. Absent that, I see no independent verifiable evidence of any of their claims, endless precedent to dispute their claims, and not even a token effort to provide even a sliver of objective backing for their claims. So why should I or anybody else reject the standard wisdom of “you are screwed if state sponsored attackers are interested in you and there is no product that can help you” and instead believe Apple’s marketing that they can?
- saagarjha 4y ago
  Security is not black-and-white, it's shades of gray. This feature aims to make exploitation harder. Formal proofs and certifications are nice but what I just said remains true even in the face of such things. iOS is regularly tested in the real world against highly resourceful attackers, and the results there are far more indicative of how well its security fares than anything else could be.