3 ms·

a) No one is blaming the user. b) End to end encryption has always meant transport encryption not encryption at rest.

threatofrain 5y ago
I mean, I'm blaming the user. What's wrong with that in this context? Similarly, if your region hates spicy food and thus spicy is dying in your area, I'd be blaming the region too. If you think the word "blame" is too morally hot, then we can always go with causal attribution.
amelius 5y ago
a) this feature supposedly exists for the average user who wants to have their data back if they forget their password. Apple points at the user and says they are the reason for having this backdoor. b) this is just false; whether the data is at rest or not has nothing to do with the definition of the term. And it is misleading in any case.
- threeseed 5y ago
  a) It's not a backdoor but an explicit design tradeoff. It's hypocritical to be calling people out for using terms correctly and then misuse the term backdoor. b) The etymology of the term has been blurred by WhatsApp referring to it as both transport and encryption at rest. But historically it has not included the latter part.
  - amelius 5y ago
    > It's hypocritical to be calling people out for using terms correctly and then misuse the term backdoor. No, it's not because it is another instance. Someone can have one thing right and another thing wrong. And whatever the true meaning of "E2EE", its use in this case is misleading at best, so I'm pretty sure that the profession does not approve of it.
JadeNB 5y ago
> a) No one is blaming the user. The user is being used as an excuse: "we compromise your privacy because that user over there wants it". This is not blame in the obvious sense of "it's that user's fault that they lost their data", but it is blame: "the responsibility for this thing isn't ours, the people who wrote the code, but yours."
amarshall 5y ago
> End to end encryption has always meant transport encryption not encryption at rest. Sure, but if someone in the middle wants to store, then they don’t have anything unencrypted to store. Ultimately it all depends what one defines the “ends” as. In general for backup, both ends are the user. For chat, the sender and recipient. Apple is not an end, but a provider in the middle.
- d1sxeyes 5y ago
  User A messages user B using iMessages. User B screenshots all of those messages and uploads them unencrypted to a random FTP server. Now replace the random FTP server with iCloud. Or Google Drive. Does that change the nature of the messaging between User A and User B? I would say it doesn't. In fact, WhatsApp does basically the same.
  - amarshall 5y ago
    No, but it means the backup is not end to end.
spsful 5y ago
OP is quite literally blaming the user.
foxpurple 5y ago
End to end encryption usually refers to users being the ends and the service provider being the transport. So while it is sitting on apples servers it is kind of in transport but delayed until you make the request to get it back.
beermonster 5y ago
E2ee, transport encryption, encryption at rest and client side encryption are all different things.

I mean, I'm blaming the user. What's wrong with that in this context? Similarly, if your region hates spicy food and thus spicy is dying in your area, I'd be blaming the region too. If you think the word "blame" is too morally hot, then we can always go with causal attribution.

amelius 5y ago

a) this feature supposedly exists for the average user who wants to have their data back if they forget their password. Apple points at the user and says they are the reason for having this backdoor. b) this is just false; whether the data is at rest or not has nothing to do with the definition of the term. And it is misleading in any case.

threeseed 5y ago
a) It's not a backdoor but an explicit design tradeoff. It's hypocritical to be calling people out for using terms correctly and then misuse the term backdoor. b) The etymology of the term has been blurred by WhatsApp referring to it as both transport and encryption at rest. But historically it has not included the latter part.
- amelius 5y ago
  > It's hypocritical to be calling people out for using terms correctly and then misuse the term backdoor. No, it's not because it is another instance. Someone can have one thing right and another thing wrong. And whatever the true meaning of "E2EE", its use in this case is misleading at best, so I'm pretty sure that the profession does not approve of it.

JadeNB 5y ago

> a) No one is blaming the user. The user is being used as an excuse: "we compromise your privacy because that user over there wants it". This is not blame in the obvious sense of "it's that user's fault that they lost their data", but it is blame: "the responsibility for this thing isn't ours, the people who wrote the code, but yours."

amarshall 5y ago

> End to end encryption has always meant transport encryption not encryption at rest. Sure, but if someone in the middle wants to store, then they don’t have anything unencrypted to store. Ultimately it all depends what one defines the “ends” as. In general for backup, both ends are the user. For chat, the sender and recipient. Apple is not an end, but a provider in the middle.

d1sxeyes 5y ago
User A messages user B using iMessages. User B screenshots all of those messages and uploads them unencrypted to a random FTP server. Now replace the random FTP server with iCloud. Or Google Drive. Does that change the nature of the messaging between User A and User B? I would say it doesn't. In fact, WhatsApp does basically the same.
- amarshall 5y ago
  No, but it means the backup is not end to end.

spsful 5y ago

OP is quite literally blaming the user.

foxpurple 5y ago

End to end encryption usually refers to users being the ends and the service provider being the transport. So while it is sitting on apples servers it is kind of in transport but delayed until you make the request to get it back.

beermonster 5y ago

E2ee, transport encryption, encryption at rest and client side encryption are all different things.