3 ms·

Signal spent months and years designing a system for backups that had no opt-in or opt-out feature. They deployed it over a period of months, ignoring a lot of

Signal spent months and years designing a system for backups that had no opt-in or opt-out feature. They deployed it over a period of months, ignoring a lot of vocal criticism from people who said "this is fine, but why aren't you letting me opt out?" Finally, at the 11th hour, they hacked in an opt-out feature. Note that the opt-out feature doesn't even stop data from going to Signal's servers -- preventing this upload appears to be technically infeasible! Rather, the hack is to send the data anyway, but encrypt it under a random key. It does not bode well for Signal's security engineering and design philosophy that "don't upload data to a server, even when users understand and accept the consequences of that" is not something Signal even considered supporting, and their design is completely incompatible with the idea. If you want to think of this article as a post-mortem on a thing that went wrong at Signal, that's fine with me. Claiming that it shouldn't be written is just silly.

tptacek 6y ago
I don't believe anyone is really reading this thread and so it's probably not worth our time to further litigate this, but the criticism you're making here is of a kind with what I described in my comment: that I'm supposed to believe that Signal is either working in bad faith or is negligent with user data, when all the evidence to date strongly suggests the opposite thing: that no other project of comparable popularity is as diligent as Signal is about not exposing user data, even at significant cost to UX and uptake. I don't understand "claiming it shouldn't be written", by the way. That's not my claim. My claim is that it is written weirdly.
- matthewdgreen 6y ago
  If my post indicated that Signal is working in bad faith, please let me know. That wasn't my intent and I went to a lot of trouble to signal that. Even people who have done a good job in the past can make mistakes, and this post was written to highlight what I think was a mistake. ETA: I'm not writing here because I want to have a public argument with you. That's what Twitter is for. I'm curious because you seem to have strong feelings and I respect your opinion, even if I often disagree with it.
  - tptacek 6y ago
    As I said in first comment I wrote here, because of the way you wrote the piece, with an addendum that basically refuted most of what you had to say in the long windup of microarchitectural side channel explanation and getting beers with Daniel Gruss, indications of bad faith from Signal were essentially all you communicated.
    - matthewdgreen 6y ago
      I re-read the addendum. Could you point me to the part that refutes the idea that microarchitectural side channel exploitation represents a threat to SVR? Cause I think you and I both agree that it does. As far as I can tell, all the addendum says is "I complained a lot and as a result of a direct conversation with Moxie he hacked in a really jury-rigged opt-out to the system, also he disagrees with me that a compromise of this system would be a big deal."
      - tptacek 6y ago
        How many paragraphs into your post are we before you acknowledge that this feature, which is universally present in other secure messengers, can simply be disabled? Is it before or after you discuss how hard it is to disable message backups in other messengers, a feature that Signal doesn't even have? Before or after your beers with Daniel Gruss? (You asked, which is why I'm pointing this out; I'm comfortable with how my original comment reads, and that comment didn't suggest that nothing like this should have been written, or accuse you of bad faith).
        matthewdgreen 6y ago
        The first 60% of the post is a technical explainer of SVR, and SGX which it relies on. I've been meaning to write this post for a while, and it doesn't become less relevant because Moxie hacked in an opt-out -- SVR is still present in Signal, and therefore it's still technically interesting. It will likely become more technically interesting over time. The first paragraph that I think you could actually consider a criticism of Signal is this one: "My major issue with SVR is that it’s something I basically don’t want, and don’t trust. I’m happy with Signal offering it as an option to users, as long as users are allowed to choose not to use it. Unfortunately, up until this week, Signal was not giving users that choice." (I added the emphasis here, just to make clear you don't have to read to the addendum for this information.) I guess we'll have to disagree on this one. It seems like you view this as an attack on Signal, and I view it as a tech post that contains a mild criticism.

tptacek 6y ago

I don't believe anyone is really reading this thread and so it's probably not worth our time to further litigate this, but the criticism you're making here is of a kind with what I described in my comment: that I'm supposed to believe that Signal is either working in bad faith or is negligent with user data, when all the evidence to date strongly suggests the opposite thing: that no other project of comparable popularity is as diligent as Signal is about not exposing user data, even at significant cost to UX and uptake. I don't understand "claiming it shouldn't be written", by the way. That's not my claim. My claim is that it is written weirdly.

matthewdgreen 6y ago
If my post indicated that Signal is working in bad faith, please let me know. That wasn't my intent and I went to a lot of trouble to signal that. Even people who have done a good job in the past can make mistakes, and this post was written to highlight what I think was a mistake. ETA: I'm not writing here because I want to have a public argument with you. That's what Twitter is for. I'm curious because you seem to have strong feelings and I respect your opinion, even if I often disagree with it.
- tptacek 6y ago
  As I said in first comment I wrote here, because of the way you wrote the piece, with an addendum that basically refuted most of what you had to say in the long windup of microarchitectural side channel explanation and getting beers with Daniel Gruss, indications of bad faith from Signal were essentially all you communicated.
  - matthewdgreen 6y ago
    I re-read the addendum. Could you point me to the part that refutes the idea that microarchitectural side channel exploitation represents a threat to SVR? Cause I think you and I both agree that it does. As far as I can tell, all the addendum says is "I complained a lot and as a result of a direct conversation with Moxie he hacked in a really jury-rigged opt-out to the system, also he disagrees with me that a compromise of this system would be a big deal."
    - tptacek 6y ago
      How many paragraphs into your post are we before you acknowledge that this feature, which is universally present in other secure messengers, can simply be disabled? Is it before or after you discuss how hard it is to disable message backups in other messengers, a feature that Signal doesn't even have? Before or after your beers with Daniel Gruss? (You asked, which is why I'm pointing this out; I'm comfortable with how my original comment reads, and that comment didn't suggest that nothing like this should have been written, or accuse you of bad faith).
      - matthewdgreen 6y ago
        The first 60% of the post is a technical explainer of SVR, and SGX which it relies on. I've been meaning to write this post for a while, and it doesn't become less relevant because Moxie hacked in an opt-out -- SVR is still present in Signal, and therefore it's still technically interesting. It will likely become more technically interesting over time. The first paragraph that I think you could actually consider a criticism of Signal is this one: "My major issue with SVR is that it’s something I basically don’t want, and don’t trust. I’m happy with Signal offering it as an option to users, as long as users are allowed to choose not to use it. Unfortunately, up until this week, Signal was not giving users that choice." (I added the emphasis here, just to make clear you don't have to read to the addendum for this information.) I guess we'll have to disagree on this one. It seems like you view this as an attack on Signal, and I view it as a tech post that contains a mild criticism.