It seems like you're implying they'd be too good to ever get caught, but... they got caught. The trouble is, making a backdoor less obvious makes it more likely that if they try it 10 times they don't get caught all 10 times, more likely it gets into production before they get caught, more likely that it stays in production for a year instead of a month, etc.

▲

tptacek 18 hours ago | parent [-]

Who got caught? The Juniper hackers? Obviously yes. They're not NSA.

Also, "never getting caught" isn't what NOBUS means.

▲

AnthonyMouse 16 hours ago | parent [-]

I mean, didn't the NSA also get caught by Snowden? They intended it to be a secret.

But the Juniper hackers are the NOBUS failure because changing the locks on a backdoor that somebody else had installed is easier than getting one installed yourself.

▲

tptacek 15 hours ago | parent [-]

I don't think you're following. "NOBUS" doesn't mean "nobody but us can ever find out about the backdoor"; it means "nobody but us can actually use the backdoor". Ironically, the Juniper PKRNG backdoor --- I assume it was Chinese --- is also a NOBUS backdoor!

▲

AnthonyMouse 11 hours ago | parent [-]

> it means "nobody but us can actually use the backdoor". Ironically, the Juniper PKRNG backdoor --- I assume it was Chinese --- is also a NOBUS backdoor!

Except that it was intended to be "nobody but the us (i.e. the NSA)" and now you've got China using it.

	▲	tptacek 11 hours ago \| parent [-]
		No, we don't. Respectfully, I don't think you're working from an accurate notion of what "NOBUS" means, and I don't think you have your head fully around the Juniper hack. The Juniper hack replaced the existing backdoor; it didn't break it. NOBUS or not, if your adversary controls your source tree, you're boned. Here, the adversary replaced "our" NOBUS backdoor with theirs. Two different backdoors, different keys, same structure.