Obviously you can have source code without reproducible builds, what a ridiculous thing to say?
-
-
Replying to @taviso @dEnergy_dTime and
In practice you *don't* have source code without repro builds. You have approximate source code that differs from the actual source the binary was built from in various ways for various reasons.
2 replies 0 retweets 4 likes -
-
Replying to @wiretapped @RichFelker and
First, you need to explain why you want to use that specific binary so badly? Here is what I do if I don't trust a binary: I compile the source code, and use that binary. What is the attack against this system, which works today, that you're trying to solve?
1 reply 0 retweets 0 likes -
Replying to @taviso @RichFelker and
it's solving the problem that very few people have sufficient resources to build all the binaries for all the software on all of their systems, so in practice nearly everyone must rely on lots of other people's build servers. without repro builds, every build server is a SPOF.
2 replies 0 retweets 0 likes -
Replying to @wiretapped @taviso and
also, for organizations who do have sufficient resources (i.e. companies like your employer) it gives them the opportunity to make it much more difficult for a bad actor to compromise their own builds (even for internal and/or proprietary builds, potentially)
1 reply 0 retweets 0 likes -
Replying to @wiretapped @RichFelker and
Yes, your proposal is that we need more more infrastructure to maintain, more access to audit, and more attack surface for real attacks that actually happen to defend against attacks that don't seem to happen. Do you understand why I think it's not a strong argument?
1 reply 0 retweets 0 likes -
Replying to @taviso @RichFelker and
for closed-source, i agree there's a tradeoff between ensuring build integrity and ensuring source secrecy. for the rest of the world, the only downside is that it is some work; i think it's worth it to eliminate SPOFs and am glad when
@ReproBuilds et al receive funding to do it!1 reply 0 retweets 0 likes -
Replying to @wiretapped @RichFelker and
Huh, that's the first time I've seen a pro-repro person acknowledge literally any flaw in it. Let me ask you this, do you agree that you can eliminate the same SPOFs *today*? I understand the benefits of repro builds, do you understand the problems?
2 replies 0 retweets 0 likes -
Replying to @taviso @RichFelker and
you're saying you can eliminate the SPOF by building software oneself, but i assume you aren't saying it would be practical for everyone to actually do this. i'm saying repro builds enable everyone's phones and laptops to stop relying on build SPOFs in a way that is practical.
1 reply 0 retweets 0 likes
You don't have to build it yourself, you can choose someone you trust to build it. That is required in your system too, how do you imagine users will verify the build can be reproduced? They nominate a trusted party to check.
-
-
Replying to @taviso @RichFelker and
Leif Ryge Retweeted Leif Ryge
The user doesn't need to pick "a" (singular) party; they can pick several independent parties and require build signatures from m of n of them. (But their software update mechanism should come pre-configured to trust several, so actually they do nothing.)https://twitter.com/wiretapped/status/1265028062074241027 …
Leif Ryge added,
2 replies 0 retweets 2 likes -
Replying to @wiretapped @RichFelker and
How will they trust the software update mechanism?
1 reply 0 retweets 0 likes - 5 more replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.