fay59

My understanding is that the NCMEC actually has the pictures, and they’re basically the only people in the US who can legally hold onto them.

The government will have to overcome the same issues compelling NCMEC to add pictures to its data set that it would have to overcome with Apple.

Additionally, by compelling the NCMEC, the government cannot increase the scope of searching or lower the match threshold.

> One of the most powerful arguments in Apple’s favor in the 2016 San Bernardino case is that the company didn’t even have the means to break into the iPhone in question, and that to build the capability would open the company up to a multitude of requests that were far less pressing in nature, and weaken the company’s ability to stand up to foreign governments.

I think that a crucial missing piece is the FBI's argument: Apple is fully capable of developing and signing a "law enforcement iBoot". IIRC, the FBI was even willing to have someone else develop the software and only ask Apple to sign it–which they definitely have the capability to do, and only policy of not signing other people's software stood in the way.

If we agree that Apple was right in 2016, it stands to reason that Apple cannot be compelled to modify its CSAM filter to capture arbitrary contents, or report it at lower thresholds, or expand it beyond iCloud Photos (like to Messages itself). The amount of work they would have to do for it seems like it would be even higher. There are whole infrastructure pieces that just don't exist. What am I missing?

Given that enforcement is only possible on photos uploaded to iCloud, the fact that classification happens on device seems like a distinction without a difference.

Has anyone ever publicly talked about controls on user data access at Google? There’s several folks defending Google here, but employees’ goodwill above sont number of people who have access will almost necessarily fail at some point.

Mac OS 9 has virtual memory (it uses virtual address translation and it can swap pages to disk), but it still has a single address space that is shared between every application.

You could have saved yourself a fair amount of embarrassment with just a little bit of reading. Apple can’t decrypt beacon locations. https://support.apple.com/en-ca/guide/security/sece994d0126/...

Genetic determinism is pseudoscience that is generally associated with eugenism and scientific racism. Giving this much importance to genes is a fringe belief.

Given the extremely short supply of telescopes deserving naming, it’s absurd to think that anyone who doesn’t have a telescope named after themselves has been erased from history.

What you’re defending is James Webb’s immaculate memory. Please think long and hard about whether it’s worth defending the immaculate memory of anyone.

These attacks are reported to have long-lasting consequences. That would be (I think?) incompatible with mass hysteria or nocebo effects.

It’s in a very precise midpoint between Notepad and an IDE that preciously few other text editors stand in.

Security fixes are backported with some frequency, but new mitigations are usually not. If a bug is made inexploitable by a mitigation in the latest OS, older OSes that have the same vulnerable code can probably be exploited, even though they still get security updates.

No, you’ve missed the point. It’s still much better to have a trap than “any other imposed behaviour” because the compiler can reason that this condition is impossible on the happy path and remove code accordingly.

Removing code is both among the most sensitive and the most effective optimizations that compilers do, and they use UB as another signal for “unreachable”. The current model of “assume things never happen and don’t plan anything in particular if it does happen” does no one a service.

You could also take a measured approach and, like rust, just define signed overflow to wrap. UB and its current interpretation is not worth whatever small sliver of performance the current implementation of LLVM appears to give you.

That's not a good take. Compilers love UB because it signals unreachable code, and unreachable code can be removed and/or its surroundings can be simplified. If UB was instead defined to trap, most optimizations that are made possible by UB would still be possible, but without the footgun part.

Specifically with this example, if the rust community cared, it could implement pattern detection for this operation and reduce it to a constant-time operation too; the reality is just that nobody cares and this optimization was probably only added to make Clang look better on a specific benchmark.

The other problem with UB is how capricious the specification is. There is no contemporary reason for why signed integer arithmetic is UB and unsigned integer arithmetic isn't. It's just a wrinkle from the past that somehow, some people prefer to praise than to fix.

Have there yet been reports of the AI writing code that has security bugs? Is that something folks are on the lookout for?

They have to load it before first unlock by definition, and at this point I don't believe that it would be possible to establish persistence. Restart your phone when it's handed back to you by people you don't trust.

It’s always been known that your carrier has access to your unencrypted cell traffic (including voice and text) and that carriers are slimy. You’re also protected by using secure services over IP. I think that the set of people for whom this will cause a threat model change is really small.

One of the points brought up by privacy folks in review of Apple’s plan to have your ID in your digital wallet is that the mere convenience of allowing access to ID may create ID requirements for users where none existed before, which is a loss for privacy. Do you think that Identity is going to create such new requirements?

I have no idea what people use to mine cryptocurrency these days and quite frankly I would rather it all stop, so no, I’m not saying that.

I absolutely despise cryptocurrency, so this may not be accurate, but my understanding is that they’re only reliable if many independent people mine. If NVIDIA kept all its GPUs and mined Bitcoin instead of selling them, it could control more than 50% of the hashing capacity. This could allow NVIDIA to double-spend bitcoins, the possibility of which would probably crash confidence in the asset.

The problem is that not having the library usually means you also don’t have the headers, so you can’t compile the C code that would use it. What you’re talking about isn’t impossible, but it probably requires either a binary distribution built from a system aware of these libraries (this is the Darwin+availability attributes approach) or some massive header repository that you can pull from to build things.

I’m assuming that instructions to disable the AirTag are basically instructions for how to take off the battery.

There's a good point to make that Patreon's donation tiers mean patrons can be rewarded for their donations. I would hope that the digital equivalent of a tip jar is a different category since it's not a purchase.

I completely disagree. The developer offers the same version of their game for free to everyone. Donations are solicited, but people who donate don't receive anything in return.

In-app purchases have value to developers to verify payment validity and function as DRM to prove the user did purchase a thing, and they have value to the user to provide legitimacy to the transaction and assure that if they don't get what they paid for, there's reliable customer service to help them. None of these things are relevant when users donate.

So what's Google's role in this? If they're OK not getting a cut of subscription services paid for outside the app, what's the problem with not getting a cut of transactions users get nothing out of?

I don’t remember the exact controversy, but I remember some controversy during the Obama years where a prominent US company complied with some censorious Chinese requirements. Obama said something along the lines of “I wish you’d come to us before agreeing to this. We would have helped.”

As these the list starts in 2017, I wonder how much of this is the result of the US’s decreased willingness to export its values abroad during the Trump years.

You're right that in the end, the FSF does not have to produce anything of value. However, that is literally the rock bottom of standards to hold an incorporated entity that receives donations.

The problem with the FSF being stagnant and useless is that it still governs a (decreasing) share of software that many people rely on and hamper progress on these projects.

To say that the FSF failing is different from people not caring enough, you have to define the goals of the FSF in a way that it can succeed regardless of whether people care about it or not. I don't think that this is possible, and therefore the FSF failing is indistinguishable from people not caring enough.

Even people who aren't very good with email or don't understand the difference between WhatsApp and SMS can do Zoom/FaceTime calls or use E2E encrypted messaging platforms while there is no sanctioned FSF platform to do either. Maybe the public would be concerned with the FSF's ideals would the FSF do the smallest modicum of effort to keep up with the public's use of software.

And, of course, the FSF failing is not an indication that the open source movement failed. Open source arguably won. However, free software is currently grasping at the threads of relevance.

This definitely riled up a lot of people. The comments point out many flaws, but I’m not seeing rebuttals to what I find to be the core arguments:

* The overlap between what the FSF thinks a majority of users want and what they actually want is minuscule * The FSF failed to inspire new leaders * The FSF, a political foundation, has no governance agenda * The FSF missed the boat on a plurality of software revolutions, causing free software to capture only a small portion of the explosive growth of all FLOSS software

Is any of this incorrect?

What is non-sketchy offensive cyber-security to you?