Raphiel Rollerscaperers @raphielscape

The Funny | Backend Engineer and Data Engineer | Oh nyo. Nightmare Realm Joined July 2018

Tweets

553
Followers

252
Following

76
Likes

276

Raphiel Rollerscaperers @raphielscape

4 months ago

@GrapheneOS @TrueAmPatriot86 Who is the verifier? How do they verified it? Where could one verify it? This is literally the gist of my questions too

0 0 1 169 0

View Details

Raphiel Rollerscaperers @raphielscape

4 months ago

@GrapheneOS @TrueAmPatriot86 Lack of good faith? You're completely evading my questions here, I'm unrelated to OP or SailfishOS, I'm just giving my bread a dip on a coffee, if you say it's completely worthless, how is your inexistent solution be better than that?

0 0 1 295 0

View Details

Raphiel Rollerscaperers @raphielscape

4 months ago

@GrapheneOS @TrueAmPatriot86 Publishing the logs will attest to the build hashes that used to build the images at time, as I said on the other thread, how would one attest reproducibility for builds that has embargoed changes? Logs can be a Proof-of-hash used to build prior when embargoed code slated to GA

0 0 0 62 0

View Details

Raphiel Rollerscaperers @raphielscape

4 months ago

@GrapheneOS @TrueAmPatriot86 And how about the builds that has embargoed changes from your partners?

1 0 0 204 0

View Details

Raphiel Rollerscaperers @raphielscape

4 months ago

@GrapheneOS @TrueAmPatriot86 The scenarios I described is for pre-flash and pre-OTA installation, narrowing the scope a bit, do your Alpha and Beta testers tests the Reproducibility of the build they tested? If so, why don't you provide a way to easily attest to their build reproducibility results?

1 0 0 119 0

View Details

Raphiel Rollerscaperers @raphielscape

4 months ago

@GrapheneOS @TrueAmPatriot86 @TheVancedGamer I'm going to further extend to your collaboration too, you for sure can't apply reproducible build as an attestation for the build reproducibility until the embargo lifted up, how the build can be trusted before flashing until reproducibility can be tested externally?

0 0 0 43 0

View Details

Raphiel Rollerscaperers @raphielscape

4 months ago

@GrapheneOS @TrueAmPatriot86 @TheVancedGamer If users can trust the build from you and that you can attest that by making your build transparent to begin with without telling user to "Just build the OS like we described on grapheneos.org/build#reproduc…", for a pre-flash attest, why are you vehemently refused it?

1 0 0 50 0

View Details

Raphiel Rollerscaperers @raphielscape

4 months ago

@TrueAmPatriot86 @GrapheneOS @TheVancedGamer In a way the compiler can be swapped at build and it would be count as one

1 0 0 59 0

View Details

Raphiel Rollerscaperers @raphielscape

4 months ago

@TrueAmPatriot86 @GrapheneOS @TheVancedGamer I know that the term used strictly for compilers itself, but in this case, it can be slightly broadened to that

1 0 0 45 0

View Details

Raphiel Rollerscaperers @raphielscape

4 months ago

@TrueAmPatriot86 @GrapheneOS @TheVancedGamer Users is unlikely to verify the build directly (and let's be real no one verifies OTA before applying it) so this is a case of a possible Trusting Trust attack

1 0 0 39 0

View Details

Raphiel Rollerscaperers @raphielscape

4 months ago

@TrueAmPatriot86 @GrapheneOS @TheVancedGamer My question is legitimate, if you don't have any inputs here, you're the one who's trolling, in cases where the infiltration happens during build that is clearly opaque (malicious actor within the team itself), it can be hard to be quickly identified when OTA is served

1 0 0 39 0

View Details

Raphiel Rollerscaperers @raphielscape

4 months ago

@GrapheneOS @TrueAmPatriot86 @TheVancedGamer This would also require the user to trust everything too, especially for parts that you ships as prebuilt, how can we verify that you're not vulnerable to Trusting Trust attack?

2 0 0 100 0

View Details

Raphiel Rollerscaperers @raphielscape

4 months ago

@TrueAmPatriot86 @GrapheneOS @TheVancedGamer It does not guarantee that the build can be trusted, though if update engine has a vulnerability (e.g signature parser buf overflow), both cros and android would be in a lot of pain, but there is a possibility of this might happen and unknown to even Google

1 0 0 192 0

View Details

Raphiel Rollerscaperers @raphielscape

4 months ago

@TrueAmPatriot86 @GrapheneOS @TheVancedGamer If you're part of Android Security or anywhere near Device and Services, we're doomed

1 0 0 52 0

View Details

Raphiel Rollerscaperers @raphielscape

4 months ago

@GrapheneOS @TrueAmPatriot86 @TheVancedGamer In cases where the OTA is somehow intercepted (as in one somehow convinced update engine to accept the OTA), this would be your weak point if you didn't let users verify the build beforehand, before they accepted the OTA to be installed

2 0 0 192 0

View Details

Raphiel Rollerscaperers @raphielscape

4 months ago

@GrapheneOS @TrueAmPatriot86 @TheVancedGamer Stock Pixel verifies the OTA integrity too, and I didn't see you deviate far from it based on your changes on the update engine and your OTA app, this also assumes that the hash is a trusty and you verify the hash pre-OTA

1 0 0 182 0

View Details

Raphiel Rollerscaperers @raphielscape

4 months ago

@TrueAmPatriot86 @TheVancedGamer @GrapheneOS Logs are one point of truth, Point-of-time build logging is one of the truth, artifact, and the build itself would need to be verifiable too, as I said on the other thread, if the whole build flow is not verifiable, how would one trust the OTA builds?