Okay, Right To Repair folks. Can you please stop implying that Apple is deliberately sabotaging repair every time you run into a calibration problem?

You are seriously embarrassing yourselves and making it impossible for me to support your narrative.

Yes, Apple have had some instances of deliberate parts pairing (often with at least some justification, e.g. security for Face/Touch ID stuff). They also have increasingly detailed per-device and per-part calibration (their attention to detail and quality is not a myth).

Please, please stop pretending like some evil engineer at Apple put in a deliberate sabotage that makes things work badly when you swap random parts. That is completely ridiculous and absolutely not how any of this works. Deliberate pairing makes things stop working (like the aforementioned Face/Touch ID case) or throw up errors or warnings. If stuff is working, just badly, that's not "pairing". That is calibration. You have swapped a part that relies on calibration stored on another part, and with the mismatched calibration, the part won't function to original standards.

I cringe every time I hear stuff like "Apple are anti repair and pairing their Mac screens so that if you swap them the backlight is uneven" or "Apple are anti repair and pairing their iPad screens so if you swap them your Apple Pencil doesn't draw straight lines any more".

You need to stop pushing these ridiculous conspiracy theories and instead focus on reality: these machines are complex, their production is complex, their repair is complex, and just swapping parts around willy nilly may not result in a quality result, and that is normal. Advocate for Apple to provide access to their calibration re-provisioning processes instead, so you can actually get things set up properly and working as intended by the manufacturer. Them not providing those tools sucks and is anti-repair. The product engineering that requires those tools for a proper outcome is not.

And accept the fact that, for some of the stuff you want to do, you're never going to get OEM quality results, because that's just how it goes. If Apple's brand new screen replacement process involves loading factory calibrated backlight uniformity data from a central database, and aging information is only stored locally on device, and you swap a screen with a used one and don't have the aging information any more to transfer over, tough luck. That's engineering. You are running into complex engineering your repair process can't deal with and Apple can't help you with. Not anti repair.

And also, cut it off with random ideas like similar parts not working across different machines being deliberate pairing. I've heard that one about M1/M2 Mac trackpads. Of course those are not interchangeable, they moved the entire controller on-die for M2 and, if they're even truly physically identical at all, they run wildly different onboard firmware! That's not pairing, and claiming it is just means you have no idea how these things work. To clarify: you can replace the trackpads, with the same model. Just not across models.

@marcan drink every time Forbes posts another clickbait headline about how everyone should panic and uninstall something

@marcan Okay, okay. But can you explain to me the reason for having specific screen controllers designed with megabits of EEPROM accessible via SPI, but using a separate chip for calibration parameters?

@babouille No, because I don't know how the specifics are designed. It's possible they provision factory calibration data on device global data storage and only store aging info on the screen. That might be done partially for security reasons (Apple are big fans of centralizing firmware/config storage and ensuring it is signed and verified to prevent that data from becoming an exploit vector, among other things).

Maybe the design is kinda dumb, maybe it isn't. We'd have to investigate to understand what is going on.

But just shouting "sabotage! parts pairing!" just makes you look like a total fool.

@marcan Strange then how other manufacturers of similarly complex tech don't seem to have the same problems... 🤔

@kishcom @marcan it’s been well-documented that Apple’s attention to detail on a lot of these things is far and away above the rest. Just read some of marcan’s toots and you’ll see the level of effort they go to. What other android tablet manufacturer even manufactures something like the Apple Pencil, which is refreshing at 120hz across the screen, wirelessly? Who else is offering P3 wide color screens that are impressively accurate in their color reproduction?

@kishcom Funny how no other manufacturer is as insistent as Apple in centralizing firmware and config storage and ensuring that everything is signed for security.

There's a reason I keep saying Apple Silicon Macs are *the* most secure open boot consumer platform you can buy today. The difference is real. Apple put in a ton of attention to detail that is leaps and bounds above most other manufacturers. This has been evident, time and time again, as I've worked to understand these machines for the past two and a half years.

"Why doesn't this happen with other manufacturers?"

Two reasons:

- Apple cares more. Seriously, some of their engineering is unheard of elsewhere in the industry. And they have a specific focus on security that is miles ahead of everyone else.
- Apple is vertically integrated. They control their platform throughout and make engineering decisions that intertwine parts as a consequence, because it makes sense when they can do that for various engineering reasons. For example, their LED matrix backlights are controlled by an in-SoC microcontroller, and on M2s and above the trackpad touch processing algorithms also run in-SoC. This allows them, among other things, to ship those firmwares per-OS version and not have to bother with keeping backwards compat on the interfaces, which lets them move faster and improve more easily. Practically no other manufacturer does this to that extent, usually because they outsource complete sub-modules which naturally results in simpler dividing lines that then means easier part swaps. E.g. I bet any other platform doing LED matrix backlights will have the on-screen TCON handle that processing instead.

@gracicot Then why are people making up conspiracy theories about parts pairing and deliberate sabotage instead or focusing on the unavailability of those tools? That's what gets me.

I fully support advocating for access to repair and provisioning tooling.

I do not support crazy conspiratorial claims about parts pairing and deliberate anti-repair engineering.

@butternut list proprietary Apple tech and and ask "who else makes these"? 🙄

@marcan Do you think that "calibrated pairing" somehow falls outside of the jurisdiction of right to repair? That it's somehow different enough of an impediment to label it a non-hostile action by a company who has a vast history of hostile actions?
Further, how are you so sure of anyone else's current attention to detail if Apple Silicon are the machines you've worked to understand for the past two and a half years?

@kishcom @butternut
Access to calibration/provisioning tooling absolutely should fall under right to repair. That doesn't mean you get to make up conspiracy theories that make no engineering sense. Withholding the tools is hostile. The design isn't.

Why do I know Apple is special? Because Apple Silicon machines aren't the first machines I've ever looked at, lol. I'm not new to this game. And all you have to do is tear apart an Android ROM and poke around the firmwares and see it works *very* differently.

@marcan you're obviously incredibly dedicated, but this take is cursed. Why can we switch out components in many other manufacturers equipment?

Apple are licensing others tech, they aren't inventing any foundational tech. It's all assembly and firmware. If they were secure that their software was so good, they'd be open to others making software for their devices. Instead they gate keep.

And they've always gate kept the other side too. Easier to seem amazing if you don't let others play.

@lewiscowles1986 I answered that already. Apple absolutely *are* doing a ton of innovation and novel design no other manufacturer is. https://social.treehouse.systems/@marcan/110803450257024343

@marcan
It’s way easier to assume that Apple is trying everything it can to suck more money out of you.

@jrod3737 They already do that more than enough for RAM/SSD upgrades, they don't need to do it to troll people doing DIY screen swaps too :P.

@marcan
Agreed!!

@marcan @jrod3737 because Apple isn't a trillion dollar company that would really love to use every opportunity to milk your wallet, that would be absurd amirite

They trolled people for fucking wheels on their Mac Pros, don't tell me they can't do shenanigans at all

@kentindell @revk Honestly I think modern cars have more of a deliberate pairing problem than Apple (though I could be wrong).

@marcan

> Apple cares more

Seriously? 😂 Let's glue more MacBook parts into place...

> to ship those firmwares per-OS version and not have to bother with keeping backwards compat on the interfaces

That you see this as a positive thing, especially doing the work you do, is wild.

@kishcom

1. I am not commenting on Apple's mechanical design here. Though I hear they're getting better at that.

2. It's a positive for them. It's hell for us but we manage :P

And we absolutely would never claim it's anti-Asahi, because it isn't. It's just what makes sense for them. They don't care about us, which is *very* different from saying they are deliberately sabotaging us.

And to add a bit of nuanced discussion around pairing:

- Pairing Touch ID is absolutely warranted for security. I hear a ton of work went into that. You don't want someone replacing the sensor with a fake module that just captures and replays your fingerprint.
- Pairing Face ID cameras is warranted for the same reason.
- The Face ID dot projector absolutely requires calibration data, that's just how the tech works.
- The Face ID flood illuminator though? I have no freaking idea why they pair that. It also happens to be a part that needed a swap on an iPhone I repaired once (and none of the other Face ID components) and I absolutely resent them for that one. So yeah, you *can* blame them for some pairing nonsense.

None of this negates the fact that Apple *should* provide third party repairers with the tools to re-pair, of course. That process can be done securely with owner authorization. When they don't provide that or attach strings to how you can do it, that's anti-repair.

The pairing of the components (except that one) though? Totally warranted for security reasons. That's engineering. I guarantee "haha screw third party repair" was nowhere on the requirements list when that got designed.

@marcan they could have used non-resident keys, just like every other manufacturers

@marcan You're yelling against the breeze with criticism this granular (even though its perfectly reasoned), but I really appreciate it.

@marcan @kishcom @butternut and accessibility

@Hawkmoon @kishcom @butternut You did not read any of what I wrote. Try again.

@st3fan @marcan designing such systems is basically my job, but thank you.

@st3fan @marcan when a device has non-volatile storarge (especially when it's FRAM like in this case) it makes no sens to have its data stored externally

@st3fan @marcan also, a biometric sensor never exeposes nor stores the valid fingerprints, but hashes, encrypted using a key provided by the cpu.

It someone replaces your sensor (and its asic) it wont be able to steal your biometric data (there's much easier ways to do that) nor unlock your machine.

Of course, the device will looses its registered fongerprints/whatever, but you'd just have ro re-register them.

Apple's design is totally unnecessary

@dotstdy @st3fan @marcan re-read what I said: it's impossible with traditional systems

@babouille @st3fan That's not how it works. Biometrics aren't "hashed". Fingerprints work by feature extraction, and you can absolutely clone those feature sets. The sensor doesn't store fingerprints, it sends either the raw pixels (yes some of them work like that, I should know, I reverse engineered one ages ago) or the extracted features to the CPU. In either case the matching is done offboard and that's why you need a secure link.

Apple are miles ahead of the rest of the industry in threat modeling. Their threat model isn't "someone steals your device and hacks it". It's "someone evil maids your device and replaces the sensor with one that stores your fingerprints so they can replay them later". That's why they are paired, to guard against rogue sensors.

Source: over a decade of security experience, many years of looking at Apple security on the sidelines before Asahi, oh and some of the people who designed the Touch ID security are long time friends.

@marcan @st3fan i'll need some datasheets to back those claims

@marcan For the face ID camera, what exactly is the imagined attack vector here? Someone builds a special camera module or modifies its firmware to store and replay the image, gains physical possession of the device, replaces the module, returns it, the user unlocks their device, the attacker takes the device back and can now get in?

If you actually expect to be targeted by anything that sophisticated you probably shouldn't be using face ID.

@ids1024 @marcan I don't know how Face ID works, if it transmits the entire image or if it does some hashing on the device, but I figure if it wasn't strict about pairing you'd just inject a device that isn't a camera at all. Something like this fingerprint reader replacement attack on Android phones: https://arstechnica.com/information-technology/2023/05/hackers-can-brute-force-fingerprint-authentication-of-android-devices/

@ids1024 Yes. Seriously, that's the threat model level Apple designs their systems for. I know for a fact, some of the people on their security teams are or were long time friends.

I am not joking when I say they beat the pants off of the rest of the industry in this area.

@aburka @ids1024 It sends the entire image. I believe neural networks are involved in the matching; there's a whole story around switching the ANE to a secure mode for SEP to use during Face ID.

(Plus the face ID depth camera is usable as a plain depth camera, they have public APIs for that)

@marcan I’d say some of these behaviours are anti-repair. If a part stores its aging information externally, then it won’t work properly when used as a spare part.

I’m not saying that there’s an evil plot to prevent people re-using parts for repairing devices. I’m saying that their engineering process entirely ignores the fact that people want to repair devices or that people want parts to work when moved from one device to another. That’s anti-repair too.

@whynothugo They don't ignore repair, they just engineer specifically for the repair processes they themselves plan to offer, and not third parties. You can call that anti-repair in some sense, but it's still very different from deliberate sabotage. That would be like us claiming Apple is anti-Asahi just because some of their engineering choices make our life difficult. That's not how it works. Not designing for certain use cases is very different from sabotaging them.

@marcan If Apple uses this for pure engineering reasons it is their duty to provide everything possible to avoid calibration issues. With these corporations the intent behind something is not seperatable from the result. Onle the latter matters, and if that makes repair impossible without functionality loss it is their fault.

@assegaia What, exactly, makes intent magically irrelevant when a corporation is involved?

You can (and should) blame them for not providing repair tools all you want, but again, that is *very* different from making up conspiracy theories about deliberate parts pairing repair sabotage. That is just wrong.

@marcan Setting aside claims about hardware compatibility issues, Apple (like its rivals) does have a documented record of opposing repair laws:

https://www.vice.com/en/article/nz85y7/apple-is-lobbying-against-your-right-to-repair-iphones-new-york-state-records-confirm

https://www.theverge.com/2019/5/1/18525542/apple-right-to-repair-bill-california-lobbyist-comptia

And part of the reason the company is willing to invest so much in hardware engineering is the knowledge that its practices happen to present competitive barriers.

@thomasclaburn Part 1: Sure. I'm not saying Apple is pro-repair.

Part 2: Nonsense. There is absolutely no say "make repair harder" is a goal in any of this engineering. That's not how it works. Apple have better things to spend engineering hours on than that. It is absolutely nonsensical. If they do anti-repair stuff like use funny screws it's obvious stuff like that. Not "hey let's choose this overall design for a complex piece of hardware over others just because it'll make repair harder". There is absolutely no way that was the priority. It'd be stupid.

@marcan These are valid points but I do have a couple things to say.

Why doesn't the usage information go with the screen?
What makes it so the calibration data is so far off with swapped parts?
It would be nice if they had technical info on what the calibrations do to explain why they are needed.

@gudenau There are indeed many questions which we don't have the answers to. A detailed analysis of exactly how the design works might help clear some of it up.

I'm actually interested in this stuff, and if I didn't have my hands full with Asahi I'd totally spend a while trying to answer some of these questions. But honestly, every time I hear the conspiracy nonsense it puts me off of working on any of these problems and actually helping repair folks.

@Alonely0 I'm not shilling for Apple and telling you to buy their stuff. We all have our priorities. I'm just tired of the conspiracies and misinformation.

I'm very much pro right to repair, but it feels like the repair folks are getting stupider and stupider with this stuff and just shoving their hands up in the air and calling conspiracy, instead of, I don't know, actually researching the systems and figuring out if there are solutions to these problems.

@marcan Man, Apple fanbois really will find a way to justify anything. Sounds entirely like Musk-ovites.

The EU is basically going to push Apple into a "change or leave" situation and I am here for it.

@GradientU0 If the EU forces Apple to regress their engineering, that would be a massive failure of the EU.

What they should do is just force Apple to provide the tooling to third party repairers so they can do things the right way. Not make the products worse.

@marcan @ids1024 I see, well, maybe locking down the camera is less defensible than the fingerprint reader then although there could be other reasons I'm not thinking of

@aburka @ids1024 Actually locking down non Face ID uses of the depth camera if not paired is something I think they should stop doing. I suspect that's just lazy engineering (easier to just disable the entire system rather than go into an insecure degraded mode), but it's really sad that they do that.

@marcan Nuanced arguments on the Internet? Do you know how crazy that sounds? 🙃

@marcan Although I totally agree, No way Apple can't find a way to move this calibration to the display memory.

I think what makes the situation so complex is that Apple does seem to make it intentionally harder, but not (always) by outright just blocking parts, but designing things in a way where they don't function as well when swapped over. (Even through there are solutions for this, and other companies do not suffer from the same problem)

I've been out of the loop for 3 years or so though, I used to do board level repair on MacBooks so I have some personal experience with this.

@Purple @marcan Yeah totally agree here. It may make financial sense for them to design things this way, it may even be simpler from an engineering perspective, but the interests of Apple don't align with that of the environment or rest of society, so sometimes they have to be pushed a little.

Corporations aren't really evil (though I like to call them that purely by them being products of an evil system), they're simply indifferent unless you make their profits depend on it.

@babouille @marcan @st3fan https://support.apple.com/guide/security/face-id-and-touch-id-security-sec067eb0c9e/web

@chucker @marcan @st3fan okay, apple choosed to do the feature extraction on the main soc, which is stupid.

But what I was saying is other phones do that on the sensor's asic itself

@babouille @chucker @st3fan Doesn't matter where you do the feature extraction. You can still copy and replay the data at whatever level it exists between the reader and the main SoC.

@dzwiedziu I've already debunked the soldered RAM argument ages ago. You would need like 16 DIMMs to provide the same bandwidth as Apple's soldered RAM, at much higher power consumption. CAMM only goes so far. Just having a connector at all increases power consumption by requiring increased signal power.

These are engineering tradeoffs. If you choose to remain ignorant about them, that's your choice, but it doesn't magically make them not so.

@marcan @babouille So... You can't explain why they make the objectively very unusual choices they do; choices which do make repair really hard and in turn benefits their sales... But because Apple are awesome, you assume it's some 4D chess us mere mortals just aren't big brained enough to understand - and anyone who disagrees is a conspiracy nut? Because... reasons?

@thriveth @babouille It's not 4D chess, it's engineering. I can easily construct a scenario in my head where following Apple's design requirements leads to this situation, without anti-repair being a consideration at any point. I would've made the same design choices if I were them. I don't know whether such a construction is accurate or not without reverse engineering the system, but I sure as hell know that situation is a hell of a lot more likely than thinking teams across all of Apple are deliberately being instructed to design to sabotage repair.

@marcan

>Please, please stop pretending like some evil engineer at Apple put in a deliberate sabotage that makes things work badly when you swap random parts.

The engineer is not evil. The evil person here is the Project Managers and Executives that told the engineer to do that.

@atatassault *facepalm*
You don't get it. Nobody told the engineer to do that. These nuanced engineering decisions don't come top down. They are a result of engineering teams working towards a bunch of goals that do not involve sabotaging repair (but also don't involve making repair easy).

Believing it's all a big conspiracy to hurt repair is just sad and displays a deep ignorance of how product engineering works and what the engineering tradeoffs involved are.

@marcan there is also the difference between right to repair, and design to repair. Right to repair is one thing, and what I would consider the bare minimum. But there are engineering tradeoffs that make repair harder, even if not intended, and conversely, "design for repairability" is an engineering practice and emphasis that can be made at design time to prioritize repair over replacement. Nothing's free in engineering, everything is trade-offs. But we should encourage design for repair.

@Specialist_Being_677 That's a fair argument, but as you say, it's trade-offs. You're generally giving up something else when you design for repair. That may be the tradeoff that works for you, but possibly not everyone. It's complicated and there's no one answer.

@entikan Of course not. But there's a big difference between them not caring about repair, and claiming they are deliberately sabotaging repair and making up conspiracy theories. The latter creates a constructed adversarial situation and is basically misinformation clickbait, and further discourages actually understanding these systems to solve the calibration problems involved.

@the_q Imagine failing to read what I wrote so badly.

I never said a solution could not be created. I explicitly said Apple *should* provide calibration tools.

I'm just saying this isn't a big conspiracy to hurt repair every time a product team adds more complex calibration. Trillion dollar companies are made of people, and these calibration based designs aren't some top down mandate from the C suite to hurt repair. If you want to find stuff to call anti-repair, maybe talk about using nonstandard screws.

@marcan I guess with soldering the SSD to the motherboard they are also miles ahead of the competition just not in the direction I or the environment would appreciate.

@andrekuehne No, that one is stupid. They should put SSDs on modules, like they do on their more expensive machines.

@Purple Apple have a big incentive to *not* have this data scattered around, because it means they can reduce persistency attack vectors. Their DFU model is designed around the idea that the machines should be bootstrappable entirely from hard mask ROM and download all factory calibration data from their servers in the process. Your Mac doesn't even have it's MAC (heh) addresses in permanent memory, nor the WiFi calibration etc. When you do a DFU restore, all that gets pulled from their servers. This design makes Macs uniquely brick-proof (can always recover even if you destroy that data) and end-user-securely-resettable. No other manufacturer offers that. If you accidentally nuke the calibration data partition (or just the bootloader even) on an Android phone, it's a brick.

@marcan I agree this one sounds ridiculous, but I won't cry for Apple when they disable devices running non-genuine accessories or requiring a special chip for their power cables. They are clearly a profit-generating machine, more than all it's competitors.

@dermoth Accessory lock-in is a very different story from this calibration stuff. And in general Lightning was a stupid idea (entirely on engineering grounds too, it's poorly designed) that should've died long ago and I'm glad the EU is killing it finally.

Please don't assume I'm some Apple shill or think everything they do is genius and not evil. I'm just tired of people generalizing and thinking every little engineering decision is part of some master plan to screw over independent repair.

@marcan @whynothugo It assumes the device will need repair in a time & place where they still offer that repair, else you're out of luck.

That's planned obsolescence and given both repair concerns & environmental concerns of e-waste, it's unacceptable.

(It's also anti-competitive.)

@lispi314 @whynothugo That's not planned obsolescence, where are you even getting that? Apple have some of the longest software support cycles in the industry, if anything they are doing better than everyone else in that area. That does a whole lot more to curb ewaste and obsolescence than not bothering with repair tooling does to hurt them.

And their flashing/etc tooling is even longer lived. As far as I know you can still restore original iPhones and they keep all the server side infra for that alive. Which means that for open platforms (Macs) with third party OSes (exactly what I work on) you can have infinite software support cycles now, including brick repair via DFU.

@Alonely0 @marcan

i think some people in hack different discovered a lot of the pairing/calibration/serialization stuff was under this system called "FDR" (Factory Data Recovery i've heard it called), to the best of my knowledge it's sealed in a similar fashion to image4 objects? (the device basically through an FDR proxy asks the Apple server for a copy of the "trust object" then seals it to the device during restore?)

https://github.com/hack-different/apple-knowledge/blob/main/_docs/factory_data.md

for more info on this, tho note the document was last updated last year and I am not the expert on this lol

@Alonely0 @marcan

I will say as a reverse engineer that's delved into these communities, the issue seems to really be that large swaths of the R2R community have become cynical, because of how major right to repair legislation has continually died either formally or in spirit, either in the legislative process or right at the end when it was about to become law (this does happen, it happened a few months ago iirc in new york). there's a lot of lobbying money and effort in the US at least to kill right to repair bills when they pop up in state legislatures so it can get difficult at points for people to not see apple as adversarial

whatever the case, i think the fix here is not just "Apple needs to provide their calibration software" it's also "they need to provide the actual parts they themselves use in assembling iPhones" (from what i understand based on my observations of the community, big issue here is that when independent repair shops want to order parts from apple's suppliers, they're barred via an exclusivity contract so either functionally equivalent parts, or parts scraped from a donor are required here (lack of recalibration tools make this difficult, and if calibration is only a one and done deal (as in it can't be run again), then parts availability is also going to make calibration difficult)

i'm definitely with you on the whole "Apple sabotaging repair" narrative being misleading at best for sure though, i just think that the r2r community's cynicism is reaching a fever pitch so to speak, and to be fair i'm absolutely not *excusing* it, i just think it's worth considering that as a factor in the conspiracies gaining traction, even if it's the wrong headspace to be in.

@marcan

The pencil that comes with an iPad will not draw straight lines if you use a 3rd party replacement screen. Apple DOES make top down decisions to fuck you over if you don't go through them to repair an apple product.

@atatassault That's not a top down decision and not deliberate sabotage. That is a quality/calibration issue with the third party screen. If you think otherwise you are pushing misinformation and conspiracy nonsense. Again, that is not how any of this engineering works.

@marcan @whynothugo That's good, will they keep providing repairs for the newer ones that cannot be repaired by others due to part pairing for anywhere near as long?

In fact, do they still provide repairs for the original or is it simply that others are able to repair them due to no such shenanigans?

In general as far as repair and maintenance goes, it's a good idea to assume one's creations will outlive oneself.

@lispi314 @whynothugo No manufacturer provides spare parts and repair services for all their past products in perpetuity. That is just not practical. At some point the only viable option becomes to get a new product. This is the case in every industry. Life cycles vary but infinite support is not possible.

Apple already provide re-pairing processes for their official third party repair flows. Yes, they should open that up more to unofficial part swaps, and I would expect them to keep such services running as long as they keep DFU restore running for those devices (so far, it's been forever).

@marcan yeah it's all tradeoffs all the way down. In my mind we should at least be requiring right to repair and incentivizing design to repair, given the fact we only have one planet....

But! I think some people say "anti repair" or "anti-right-to-repair" when what they mean is "not designed for repair", is what I think my thesis that I forgot to mention was 😜

@Specialist_Being_677 Honestly, I think the environmental concerns around repairability are overblown. We aren't going to save the planet by making devices have longer lifetimes in every way. People already keep buying new shiny stuff to get better specs periodically anyway.

I get behind wanting easily replaceable batteries (a consumable), which can easily extend the life of a device for an average user. Same with SSDs, which are the most useful upgrade option for laptops and can extend the life of desktops used in non-primary-computing use cases. But everything else? Honestly, it doesn't matter that much.

I dropped my Pixel 4a 5G a few months ago and smashed the screen. I could've tried to repair it, but it also needed a new battery and power button. Adding up all those repair costs (at iFixit prices) made it only slightly cheaper than just getting a secondhand "never used, essentially new" 5a, which would basically give me water resistance over the old phone (which I wanted), so I got that instead. Repair just... doesn't make all that much sense any more, much of the time, for better or worse.

@marcan I generally agree with the argument you're making, but how would an owner know if they should authorize a new part?

(This is a supply chain question, more than a usability question.)

How would I, as an engineer who knows something about security, decide if I should authorize that new part?

Are you arguing that supply chain risks are low enough that apple should make a different tradeoff?

@adamshostack In general, what repair shops are trying to do is swap parts around. So the security critical components are genuine and authorized, and just need to be re-paired. Apple already have processes for this available for new part swaps through their official third party repair flow. So it's not really any different from that, since it's always genuine parts (presumably provisioned with an authentication keypair/cert at the factory).

That's different from, say, allowing third party Touch ID sensors. They absolutely should not and that is a valid security decision.

@marcan Is this really a meaningful difference to consumers? However we justify it technically, the practical outcome is: Apple released a product where the calibration is either not properly bound to the calibrated element, or not complete enough for replacements. It's their decision and with their history of anti-repair it may or may not have been a conscious business decision. It's not a cheap product either - they have decent enough margins to not do that.

@viraptor It's actually a deliberate, legitimate design choice with major benefits. Apple devices are designed to be recoverable from complete loss of volatile memory. This makes them uniquely brick-proof and secure, since you can always do a full DFU wipe and know the device is back to a known, secure, uncompromised state.

If you lose your calibration data partition on an Android phone (or even just the bootloader), it's a brick forever. While if that happens on an iPhone, you just plug it in and restore it with the tools Apple makes available to everyone.

The trade-off of this design is the devices download all their complex calibration information during DFU restore from Apple's servers, which means you need phone home tooling to update it when you swap parts.

@marcan

https://arstechnica.com/gadgets/2023/07/apple-pencils-cant-draw-straight-on-third-party-replacement-ipad-screens/

Apple uses a proprietary chip connected to their screens, that if the pencil doesn't detect, it doesn't send accurate data. That is DELIBERATE SABOTAGE.

@atatassault That is complete bullshit and not how it works at all. That's exactly what I'm calling out. The R2R people are increasingly making up these legitimate-sounding (to non engineers) conspiracy theories while having minimal to no understanding of how the systems are designed and why it might end up with these results.

It's garbage clickbait conspiracy nonsense designed to get people mad at Apple instead of actively seeking solutions to the calibration issues. I read that article and it was *immediately* obvious that the person making those claims has no idea whatsoever about any of this, and just made it up.

@rayaneb Phones from other manufacturers will brick themselves if they lose their calibration data (or even bootloader) with no recourse for repair.

If that happens to an iPhone, you just plug it in and do a DFU restore with tools available to everyone on the App Store.

Tradeoffs, tradeoffs.

@marcan 99% agreed
One that I dislike is the locking down of nand boards that don’t even have a controller inside.

Granted I see the arguments such as “daughter board could be inserted between the slot and the actual nand board to intercept data” and co, but those arguments are thin at best

@fclc I don't think they do that? If you're referring to the news cycle around Mac Studio flash modules, I never saw a serious trustworthy analysis of that and how it works. Some YouTuber just tried it once, it didn't work, then made a giant news cycle out of it. It could be they store info on module types, not lock things to specific modules.

This is a repeating theme here. Someone without the engineering background to figure this stuff out tries it, it doesn't work, and immediately blames pairing/locking and starts a news cycle instead of actually doing useful research.

@marcan @Hawkmoon @kishcom @butternut so many folks replying didn’t appear to read what you wrote … WTF

@stepheneb @Hawkmoon @kishcom @butternut And blocked me after my reply. It's like Mastodon is turning into Twitter again.

@marcan if it’s not to much to ask, would you be able to test that for real at some point? I’d love to be able to put that discussion to bed one way or the other

@fclc I only have one Mac Studio unfortunately... That's part of the issue, it costs money to do this reaearch.

@rayaneb It's not about being "closed", it's a deliberate design choice. Apple have gone to great lengths to make their devices recoverable from complete loss of volatile storage. This makes them way more robust than anyone else's devices against bricking, as well as much more trustable (recoverable security: even if your machine is compromised, you know a DFU restore will bring it back to a secure, trustable state).

The way this works is they download calibration data from their server during DFU restore, instead of only storing it locally. And the tradeoff there is that if you swap parts, you need their help to update that central database with the new calibration info.

People brick their Android phones forever all the time, just read some threads on XDA. On paper there are secret manufacturing tools that you could use to try to recover them, but they won't recover calibration stuff even if you got your hands on them. They just don't have a process for end user full reset at all. That is something only Apple has.

I am very, very grateful that Apple designed things this way so I can confidently tell people that no, you can't brick your Mac by installing Asahi Linux no matter what you do or how much you screw up.

@marcan If they attached ROMs with id/calibration data to the relevant elements, they'd have the same capabilities to restore though.
Sure, their way is completely legitimate and you're 100% technically correct... and it doesn't change the end result for the replacements.

@viraptor ROMs aren't really a thing; it would have to be Flash/EEPROM and then that adds extra cost and potential for corruption. But not just that: by keeping this stuff in a central, signed data store, they reduce the attack surface vs having a bunch of different calibration tables stored with components.

Again, tradeoffs. Yes it is possible to design devices to seamlessly support part swaps, but you gain some you lose some. Apple's design has legitimate reasons to be the way it is, and it's disingenuous to claim it's a grand conspiracy to hurt repair.

@kitsastro @thriveth @babouille Correct.

@rayaneb And this is why I said in my original post that R2R people should advocate for accessible tooling to do repairs, instead of making up conspiracy theories about anti-repair design.

@marcan @rayaneb I don’t know that you care, but it’s last jab right there that’s unnecessary and unbecoming. One doesn’t require a conspiracy when a profit motive will suffice.

@kitsastro @babouille That's a valid take, but claiming deliberate sabotage (which the R2R people increasingly are) is not.

@marcan counterpoint: Why does the trackpad controller being on-die prevent replacement of the trackpad? I see no reason for it to be that way. The engineering shouldn't be so complicated that a trackpad can't be replaced. That is bad engineering and I question why they made such a decision. Calibration shouldn't be overly complicated either for simple parts of the device such as the screen. I can understand parts that are security related due to the need to keep the device secure.

@maxinehayes It does not at all. The claim here was "part pairing" because you can't swap trackpads between M1 and M2 machines... Because they aren't the same trackpad. You absolutely can between the same generation.

It makes no sense but the R2R folks can't stop themselves from seeing every engineering decision as some kind of evil thing and it's genuinely exhausting.

@cyastis @dermoth Mostly the USB 2 thing (right when USB3 was coming out) and exposed contacts which are prone to corrosion from contamination. I've seen that way too many times to count. Also having the springs/compliant part inside the phone instead of on the cable, which is bad for durability.

The epitome of silly is that instead of designing a better connector that can pass video, they designed a dongle that takes compressed video tunneled over USB2 and decompresses it right back into HDMI. That one always blew my mind with how dumb it was. Worse video quality and a whole freaking SoC running XNU in a dongle just so they can keep using the inferior Lighting connector.

@absolutejank Of course you can replace those things. The problem is these people are doing so, finding out the replacements are out of calibration and don't work as well, and instead of lobbying for calibration tools or trying to figure out how to fix it on their own, they come up with ridiculous conspiracy theories about how Apple is putting evil pairing chips into their screens that will make your Apple Pencil draw squiggly lines as punishment for swapping your screen.

@marcan @babouille this is way too much text, I ain't reading all this.

@maartenpelgrim @marcan @babouille Twitter and its consequences

@maartenpelgrim is there a reason you wanted us to know that?

@mawhrin sorry, got carried away, shouldn've posted that.

@absolutejank do come back when you succeed in a project allowing to run linux on latest apple hardware – but thanks for your professional opinion, i guess.

@absolutejank @maartenpelgrim@mastodon.nl @mawhrin You, like so many other reply guys here did not read what I said at all. Try again 🤡

@absolutejank
…oh? that's rich: a reply human who went with snide “pls notice me tim cook” complains about getting a not entirely gracious reply?

my apologies, sir, i will make my best to tell you to fuck off in less uncertain words next time.

@marcan @kitsastro @babouille "you can't call it planned obsolescence because there isn't a killswitch timer whose only goal is to kill your battery after 24 months"

You don't need to explicitly and actively sabotage R2R to be very much against R2R. When 99.99% people can't replace a cracked screen because "something something vertical integration and security and complex tech", it IS anti R2R.

@0xSim @kitsastro @babouille Again, try reading what I said again. Reading comprehension is hard I guess.

Hint: I very explicitly said not providing the tools to deal with this stuff is anti R2R.

@marcan @isotopp My Palm Pilot had an integrated calibration program for the touchscreen.

And I really don't know, why soldering the NVMe on the logic-board of my Mac Mini M1 improved the quality. With the broken NVMe it is now just trash. Not even Apple can repair it - they would have to replace the whole logic board (which is so expensive after the warranty period, that buying a new one would make no big difference). I will keep this box as a reminder, to not buy a Mac again.

@treibholz @isotopp Your Palm Pilot had a resistive touchscreen that can be 4-point calibrated with ease. I guarantee the Apple Pencil calibration process is way more involved and probably involves custom tooling. We're talking about a capacitive array touchscreen and multiple degrees of freedom for the pencil. As technology evolves, it becomes more complicated to service and repair. This is inevitable.

I fully agree that they should not solder down the SSDs. There are things you can hold against apple. That's one of them. Parts calibration requirements isn't.

@marcan @lispi314 The infinite software support and brick repair is theoretical and substantially nerfed by activation lock.

I have a M2 MacBook Pro and it’s only useful to hold papers on a shelf and prevent them from flying off when there is a breeze. There’s a ridiculous amount of 1 or 2 year old devices that are already unusable.

@whynothugo @lispi314 Activation lock is... Complicated. I understand why they implemented that (rampant theft) but I don't agree with it.

@marcan @lispi314

I think that the situation would greatly improve with relatively simple solutions.

Like, if I walk into an Apple Store and ask them to unlock it, they could ping the owner and unlock of if the owner doesn’t reply in 6 months.

It’s honestly hard to ignore the link between this poor implementation and increased demand/sales at the price of more digital waste.

@whynothugo @lispi314 Yeah, I think they could definitely do better by providing reasonable recovery flows when the original owner can't be contacted.

@dzwiedziu That's a decision for individual purchasers to make. I very much prefer the performance/watt of these machines over replaceable RAM, thank you very much. I don't even remember the last time I upgraded the RAM on a laptop mid cycle.

@amarioguy @Alonely0 @marcan Indeed. It's mess like the CD3215 that makes people assume it's intentional.

There's also the thing that if it wasn't intentional but makes them more money it isn't like they're going to correct the situation.

@malwareminigun @amarioguy @Alonely0 You mean them using a custom part there? They do something funny with the bootup and firmware update sequences. Pretty sure it's roughly the same part as the TI equivalents, but with a custom mask ROM. This is part of their firmware integrity design; the telltale is those chips are running from the boot ROM only until the OS tells them to actually start up. We have the code for that in m1n1. Pretty sure the intent there is so that they can guarantee a clean flash/reset even if the flash memory is compromised, since then DFU mode would not go through normal boot but rather directly into reflash mode. And then I'm pretty sure they have some custom firmware signing stuff in there too.

Apple do this a lot: they get their suppliers to fix their broken (security or reliability wise) designs by making slight part variations. Unfortunately that results in them using custom parts. They probably wouldn't need to if manufacturers had good designs in these areas to begin with, but as I said, only Apple cares this much about this stuff... So they end up having to ask for custom stuff to meet their standards. Most suppliers are, as a baseline, incompetent when it comes to security (at least relatively).

@marcan just my same thoughts. At last I see someone saying it out loud. Also, their engineering solution, while not anti-repair per se, could be improved to make it more repair-friendly. For example, asking for screen calibration (like the old days), asking the user to confirm that the fingerprint reader was replaced, etc. It may make things a little bit more expensive to produce, bit we're not talking about low margin devices anyway.

@sergioaguayogr Their design actually requires phoning home for part re-pairing, for very good reasons. They try as hard as possible to not rely on volatile memory and instead download all that from a server during DFU restores. This makes Apple stuff uniquely unbrickable and securely restorable to a known good state (no other manufacturer does this). You can bootstrap all the way from the secure boot ROM in the SoC to a fully configured and working system, and know it hasn't been tampered with.

The tradeoff is you rely on them for updating their database when you swap parts, which is why they have that whole phone home process for their official repair guides now. They absolutely should open that up more and make it more accessible to independent repair, and that's where we need legislation since they obviously don't have much incentive to do so themselves.

@marcan @sergioaguayogr So technically they own your device after you bought it?

@lobingera @sergioaguayogr How exactly do they "own" it? Nothing stops you from just never bricking your device and never having to use DFU mode. If it were an Android your device would be an irrecoverable brick if you get it into that situation. They are *adding* options, not taking them away.

@marcan absolutely agreed! 100% on point!

@marcan @lobingera @sergioaguayogr On android devices there’s ways to unbrick with DFU mode but it usually involves a questionably sourced DFU flashing utility and a questionably sourced stock firmware

@charlotte @lobingera @sergioaguayogr And if you've lost your calibration data, good luck, since there's no magic to get that back.

Personally I've only been able to find those questionably sourced tools some of the time, definitely not all, and they only work sometimes, and usually only run on Windows.

Meanwhile Apple just provides the tool that magically works on the App Store for everyone to use, with documentation. And the community has reverse engineered it into idevicerestore, which works on Windows and Linux.

@marcan @lobingera @sergioaguayogr On my 3DS i played around with a dev tool and miscalibrated the lower LCD and it caused image retention lmao

@charlotte @lobingera @sergioaguayogr Yeah, LCDs need careful bias voltage calibration to prevent this. Even screwing up power sequencing can lead to long lasting retention.

@marcan @lobingera @sergioaguayogr I find it interesting that most common current and previous display technologies have this issue or similar issues

@marcan @dzwiedziu There are people who are not in your privileged position who do care about upgrading RAM.
A rising tide lifts all boats.
The fact that I may not want or need upgradability or right to repair doesn't stop me championing those rights for people who do need them.

@sleepyfox @dzwiedziu Then those people should buy machines that cater to their needs. It is completely illogical to say one engineering tradeoff is the only valid tradeoff. If you want a machine with replaceable RAM and lower performance, they exist. I don't, I want the higher performance. Both are valid options. There is no having your cake and eating it too. Modular RAM at the densities and interface widths Apple uses is just completely impractical.

@Kahanis @dzwiedziu SSDs, unlike RAM, can be made modular without any trade-off (other than mechanical, but they can definitely fit SSD modules in today's Mac lineup). So yes, in *that* case I would very much like for Apple to switch to modules. The engineering does work out in that case.

@marcan @amarioguy @Alonely0 🤷 it's also a part electrically exposed to the outside world, and handles enough power to accelerate corrosion, so they die. To replace in a MacBook, repair shops have to buy an iPhone battery case, desolder that chip, and dispose of the rest of the battery case as ewaste due to exclusivity contract.

@malwareminigun @amarioguy @Alonely0 Yeah, that's the availability problem. I'm just explaining why it's custom (i.e. not to screw over repair folks).

@marcan @dzwiedziu I'm talking about rights.
You're talking about engineering tradeoffs.

@sleepyfox @dzwiedziu Your right to purchase a more repairable/serviceable device should not make less repairable but more desirable for other reasons options unavailable. That's like saying waterproof phones shouldn't exist because if you take them apart you destroy the seal and therefore they aren't repairable.

@dzwiedziu RAM almost never fails in the field like that, that's a strawman. You'll have a better time with that argument with SSDs, which are actually a consumable. By that token everything should be modular to the tiniest part because what if any random part breaks? Except that kind of design is increasingly inviable the more modular you make it. There's a reason modular smartphones never took off and modular laptops are a tiny niche market.

I don't think anyone buys a Mac and expects the RAM to be replaceable, that's also a silly argument.

@Wowfunhappy Honestly, I'm very unconvinced by sustainability arguments beyond the really obvious stuff (making batteries replaceable, since they're a consumable). Most people buy new devices within a few years anyway, even if they aren't broken. I'd love to see solid data on how much environmental improvement could be had if everything were magically perfectly repairable (taking into account consumer purchase patterns). Most of what people talk about is that there's so much ewaste, but that in no way means it would cease to be ewaste if it were repairable. People throw out working stuff all the time too.

@marcan whilst you sound like you know much more about this stuff than me, from a none engineering user's point of view, apple has been caught out doing stuff to try make more money in a sneaky way, like with the batteries on, I forget which phone it was but finally admitted it was their fault and gave free replacements, after denying for ages, so some of this "apple is doing this deliberately" stuff is probably born from the correct assumption apple can be very sneaky and lie

@marcan @guiltmanager The battery dilemma back then (I believe it was the 6s) also was an engineering challenge with two options:
1. Throttle the processor when the battery gets too low
2. Risk the phone turning off unsafely without any information to the user
Option 2 ist what they did previously and users complained. So they went with number 1. The only fault I see is that they didn’t tell anyone about the change.

@guiltmanager @marcan The solution now is that now you have a toggle in the system settings where you can choose between option 1 or 2 yourself. In my opinion Apple never tried to limit the lifespan of the device but rather increase it, since it won’t be shutting down unexpectedly anymore.

@melgu @guiltmanager Yup, this is exactly right. It's a physical limitation of deteriorated Li-ion batteries. They have a physical maximum continuous power they can provide. If you try to draw more, it just shuts down.

But people conspiracy theorie'd it into a giant mess. Exactly the same kind of nonsense I'm seeing with part pairing.

@fclc @marcan someone else already did and you can upgrade to differently sized nand modules if you know what you’re doing https://forums.macrumors.com/threads/upgrading-mac-studio-storage.2370048/
biggest problem is that the storage modules are not available. With apple’s self repair program you’ll only get modules of the same size as the bought machine has. Buying a second machine just to upgrade the internal SSD kinda defeats the purpose. In most cases buying a USB4 NVMe makes more sense and will be cheaper than Apple’s NAND prices.

@marcan @janne thanks for this!

reading through the posts, I think there’s fair criticism to be laid at the feet of apple for (what seems to be a business decision) not allowing the nand chips to be sold to third parties

Per this post on P2 https://forums.macrumors.com/threads/upgrading-mac-studio-storage.2370048/post-32293975 it *may* be possible to buy a second SSD/nand package?

Pricing is a little Oof, but at least it exists?

Not sure If this can be used for upgrades?

@fclc @janne Note that the NAND chips are Apple proprietary. They literally have Apple-designed controllers with Apple-designed CPUs in them. That's why you can't source them from third parties. It's not a deliberate lock, it's legitimately a completely bespoke custom NAND interface (and general SSD architecture) they invented.

@marcan @WAHa_06x36 The first thing that came to mind with this most recent outrage w/ Pencil is Wacom tablets.

Genuine Wacom tablets have amazing tracking. Poorly-refurbished Wacom, “Powered by Wacom” manufactured by companies other than Wacom, and knockoffs using similar tracking technology, invariably *suck* in the exact same way that the Apple complaint was about.

People complain about genuine Wacom based on price, then willfully ignore the shitty tracking on the cheaper alternatives.

@Loeil_du_tigre Providing the tools takes considerable effort. Not providing the tools is the default.

Apple don't give a damn about third party repair, which is why they don't invest anything into this. But it's not like they could just flip a switch and solve all these issues if they wanted to either.

@marcan well, on a technical perspective, yes.

But the end result when you have to repair these is the same: you can't.

As Louis Rossmann said, Apple decided not to publish the calibration software, when they could have.
Apple decided not to publish schematics and other documentation.

cf. https://www.youtube.com/watch?v=MUifN6abtSI

@mmu_man Louis Rossmann is a kiwifarmer and actively involved and credited in harassment campaigns there (the same people who drove a friend to suicide), so forgive me if I don't have the slightest inclination to care a single iota about what he says.

@marcan@social.treehouse.systems @mmu_man@m.g3l.org well fuck I did not know that

@marcan @mmu_man Woah, he is? I'm trying to find anything on it, but there's nothing I can dredge up. That's fucked up (if true, of course).

@thomholwerda @mmu_man https://archive.is/RHBJc See the very end of the post. He's verified on there even.

@marcan I see where you're coming from, but as far as I can tell the popular narrative(by, for example, Rossman) regarding this is that yes, having calibrated parts by itself is not necessarily the issue, but not having tools or documentation available for repair techs to do the calibration is.

Sure, maybe this needs to be done in "factory conditions", but 3rd party repair folks are crafty, so if you give them the ability to reflash a hinge sensor with calibration data, they will probably figure it out.

Of course there's also an argument regarding complexity, but as you said, there's probably a good reason these things are the way they are which doesn't boil down to some engineer going "let's make this harder on purpose".

@goat @marcan As someone one said, it takes a higher level of competence to solve a problem than it does to create the problem in the first place.

@goat Yea but invariably the narratives I keep hearing are not "Apple need to provide us with the tooling to handle their complex hardware" but rather "Apple are deliberately engineering sabotage into their products" and those are two *very* different things.

@marcan @goat How is not providing the tooling not deliberate sabotage wrt the repair industry?

@qazx @goat Read what I wrote again. The engineering isn't sabotage. I'm literally telling people to advocate for access to the repair tooling.

@marcan I can agree and see where you come from for the most part, but i cant get over your M2 trackpads example.

Why? trackpads have existed for decades now, and they were replacable on 2020 macbooks, and those before then, and have been for as long as they held the reputation of having phenomenal trackpads. why do they NOW need to have controllers on the SoC? other than they can do it now since they own the silicon?

Maybe some cost cutting measure could be brought up as an counterargument, but again; they made them with dedicated controllers all this time, and they would be likely saving cents per unit if it was relocated onto the SoC, if at all.

Though you might still argue that its just a calibration issue, I see no reason for introducing it other than making the replacement harder, if not impossible.

@mlerp You misunderstood my post. They are replaceable. What you can't do is grab a trackpad from an M1 and stick it into an M2 or vice versa, because they're not the same trackpads.

I believe the trackpad calibration is, in fact, stored on-board (even in the new M2 case), so all that works as intended.

But the R2R people went and called that "part pairing" 🤦‍♂️​

@marcan I see what you mean now.

All I can say is, thats public media for you.

Anger inducing articles sell better than those promising hope and righteousness. article telling of how terrible of a company Apple is is more profitable than one telling of how to calibrate their screens. we went over this when journalists used to complain about the ecology and price of shipping rented tools to users to repair the devices.

(as for why they made trackpads on two literally hardware identical machines, just one with slightly better processor, entirely incompatible, though a bit baffling, is beyond me)

(like come on i could take a battery from my t410 thinkpad and fit it in my 2 years newer t430 and it would work just fine, and though i didnt try trackpads, you can swap keyboards on those as well, although with a file and some elbow grease)

@RisenValkyrie @mmu_man @thomholwerda He is literally credited in the harassment thread 🤷‍♂️​

@marcan @mmu_man @thomholwerda which thread?

@RisenValkyrie @mmu_man @thomholwerda The one I linked.

@marcan @mmu_man @thomholwerda linked where? sorry my instance just died trying to load this entire hellthread, i can’t exactly find it

@RisenValkyrie @mmu_man @thomholwerda https://archive.is/RHBJc

@marcan @mmu_man @thomholwerda ah wait the archive link, yeah. that’s what i went off of as well

@RisenValkyrie @mmu_man @marcan Being active on a criminal nazi forum like that one has no ifs and buts for me.

No quarter for nazis.

@thomholwerda @RisenValkyrie @mmu_man Especially after having put out a long video that was a major driving factor in that abuse campaign, even if he wasn't on KF yet then.

*Checks date* actually he was already on KF then. The harassment thread postdates his video... by a couple days. Very, very easy to see how it was a major trigger.

@thomholwerda @mmu_man @marcan he’s not active though…

@marcan @mmu_man @thomholwerda yeah that’s what im looking at as well and i literally cannot see anything that looks like “Rossmann is a harasser”

nor does his profile on KF look the part

@RisenValkyrie @mmu_man @thomholwerda Puts out a video effectively baiting his audience to harass strcat, a couple days later a KF thread pops up and credits him.

Regardless of his level of direct involvement in the KF campaign, this ain't a good look.

@marcan @mmu_man @thomholwerda you are the very first person i see claiming that he’s a harasser because of that

any larger YouTuber’s “callout” video results in harassment towards the target, that’s inevitable, i don’t see how this is any different other than “ooh look they made a bad birdsite thread about the target”

@RisenValkyrie @mmu_man @thomholwerda Maybe large YouTubers shouldn't put out callout videos. Especially not against people who have already been the target of SWATting.

@marcan @mmu_man @thomholwerda that’s a problem larger than just Louis Rossmann

@RisenValkyrie @mmu_man @thomholwerda And he's part of it.