Hacker News Reader

Comment by somethoughts | Hacker News Reader

On one hand - it makes sense to push for more visibility - to test the depths of their commitment to right-to-repair.

On the other hand - being overly critical about a new entrant that is attempting to address the right-to-repair market for not getting everything perfect from the get go might send the wrong signals about the right-to-repair market to the rest of the industry.

It seems like a delicate balance might be wise. Having module level repairability for LCD panels, connectors, drivers, memory is a pretty big step relative to the other OEMs.

Perhaps the best analogy would be Raspberry Pi schematics which are also reduced [1] and IMHO, the RPi has generally been a plus for the open source hardware/software community. Similarly I could definitely see Intel having some trade secrets as you get closer the CPU.

[1] https://datasheets.raspberrypi.com/rpi4/raspberry-pi-4-reduc...

View on HN

Yep, we absolutely understand and respect Louis and community members' requests for full schematics, and we appreciate the folks who extend understanding as we work through constraints (including Louis as noted in that last video). We're continuing to make as much open as we can, and we're adding to that list as we go:

* Mainboard partial schematics, drawings, and pinouts: https://github.com/FrameworkComputer/Mainboard

* Expansion Card reference designs: https://github.com/FrameworkComputer/ExpansionCards

* Embedded Controller firmware: https://github.com/FrameworkComputer/EmbeddedController

On the list of things that we are exploring ways to improve on in the future:

* More of the schematics, including of modules beyond the Mainboards.

* Moving to open UEFI/BIOS solutions.

This is the challenge of trying to do good - no matter how much better you are than the default, for some subset of users you will never be good enough, and they’ll complain just as loudly about you as they do the default (far worse) option.

Unrealistic purity tests ruin all sorts of attempts to make the world a better place. We should celebrate attempts to change the status quo.

The problem I have is that if these requests are unrealistic, they should just say that. As far as I can tell, what they're currently doing requires more effort to vet repair shops and create an NDA than if they just dumped the schematics online.

Now, maybe there's a great reason for this. Maybe they're under NDA and can't reveal part of the schematics. Maybe releasing these schematics would make it super easy for a competitor to completely steal everything. Maybe they are going to release the full schematics and we're waiting on their engineering team and legal team to review them.

The problem here is that Framework hasn't communicated any of this. Every time they're asked they don't seem to give any reason and just dodge around with "we'd love to release the full schematics". Well, why haven't you then? Please tell us.

One might even imagine someone would write something like this in an effort to reinforce the parents the point.

The strongest interpretation of "we'd love to release the full schematics" the subtext: but we can't / don't want to due to "reasons*, none of which are likely to satisfy everyone, so we'll try to avoid that discussion as it would help distract us all rather than getting on with our goals.

I feel its likely the marketing verbiage w.r.t. repairability was targeting the average "prosumer level repairability" - which is probably more closer to an iFixit level of repairability (i.e. more on the torx screw and less on the solder/depop spectrum of repair) [1]. That said in response to these HN and Youtube threads - perhaps they could update it to be more specific about their definition of repairability as well as why schematics will be or can't be released. In the future - if Intel is the issue - it might be smart to design some additional modularity and relegate the proprietary Intel stuff to some sort of system on module (like the RPi4 Compute Module) and increase the amount of open source able main board.

[1] https://www.ifixit.com

It seems likely that the 'module' would just cover the entire space of the board as-is if they did this with a modern intel system anyways, especially when there's also SODIMM module slots, an m.2 slot, and 4 full tb4 ports on the board.

All those high speed bus interfaces are probably tied directly into the CPU, the chipset, or both, and so are largely gonna be covered by any confidentiality agreement intel imposes on vendors of those parts.

The Pi4 CM is a much more modest device by comparison. It has no ram expansion and only a few lanes of PCIe, USB3, and some other fairly standard IO interfaces otherwise.

Realistically, the frame.work mainboard is the compute module. You plug it into the equivalent of a CM carrier board with the TB4 ports.

It's not just about the reason, but other details too. Does "we'd love to release them" mean "we just mean more time", "there's a 50/50 chance we'll be able to release them", or "we'd love to but it's a long shot". If they are going to come out, will it be in weeks, months, or years?

Them not mentioning it, despite mentioning it making their optics better, pretty much seem to point to the NDA not even letting them discuss the stuff you stated.

> due to "reasons*, none of which are likely to satisfy everyone, so we'll try to avoid that discussion

Being so vague and listing no reasons doesn't seem helpful to me.

The strongest and most likely interpretation is that they simply don't own the IP, and it's based on a licensed reference design. It makes zero sense for a startup like this to have done all of it from scratch - just copy everything possible and innovate the actually interesting parts. If they don't own the design, they can't say that due to NDAs, and the best they can do is a maximally broad interpretation of "repair shop".

IMHO, the best they can do is being honest about what they can't say due to an NDA. For example, if they can't say that they don't own the design, they might say "due to an NDA we cannot say who owns the design". This goes a long way to earn trust.

It depends if the NDA allows them to discuss it at all!

These has to be a limit to this madness, can I get an NDA that does not let you state your name to anyone

There are limits to contracts in the law. There would have to be some form of consideration to that contract. If you gave me a million dollars a year but I just can't use my current legal name in casual conversation, I'd take that contract. If you just wrote a contract that said "you can't use your name, you get nothing" its an unenforceable contract under US (and most of the developed world's) contract laws under the principle of consideration.

https://en.wikipedia.org/wiki/Consideration

I don't know about these things a great deal, but it would seem reasonable to me that an NDA might include limiting what you can say about the NDA?

They already claim to fully own the design.

>Our main manufacturing partner is Compal, who makes laptops for many other large US notebook makers.

>Apart from off the shelf modules like the SSD, RAM, WiFi, and LCD, the Framework Laptop is entirely custom to us, and we own all of the design and tooling.

If you locate Acer Swift 3 SF313-53 diagram it will be 90% same thing, just shuffled around with different mechanical design.

They might be owning the design itself, but I can easily imagine that the NDAs required to get access to specifications and design guidelines for parts like the CPU, GPU and other chipsets include a provision that no material must be released that allows others to clean-room reverse-engineer the protected material.

I wish "innovative, actually interesting parts" would include delivering an actually repair-friendly laptop - that is - one with full schematics; only ICs that are normally in-stock on digi-key/mouser; no soldered-down components made of melty plastic, especially not next to places that are likely to be reworked; soldermask that doesn't obscure traces; all important signals brought to accessible probe points; mechanical design/connectors/cabling allowing for easy power-on outside of chassis, etc...

Sure, QR stickers and a bunch of new/incompatible iteration of good old ExpressCard are nice to have, thank you, but that doesn't make all that big of a difference to repairability.

And no SMD mounted connectors, only hole through connectors. The problem with that is that it will preclude the use of certain kinds of cables because the whole industry has now pretty much moved to SMD where possible to shave off the last couple of percent cost.

External connectors are pretty much the only kind of component that is still routinely being used in (full or partial) through-hole form-factor.

Laptop in question does use partial through-hole connectors for all 4 USB-C ports - notice 6 shiny slots around each such port around 0:25 mark in video. Such connectors are actually harder to rework/replace, but also tougher and less likely to require repair in the first place. So overall, well-done here to framework, even if most everybody else is doing that too.

Yes, but there are plenty of internal connectors as well.

honestly with how much heat the ground plane can soak up on modern electronics, desoldering a beefy through hole ground connector is such a pain in the ass I might rather they be SMD if I have to do something with them in general.

This sounds like a great set of constraints for an industrial controller, or a custom Cyberdeck-type portable.

I’m confident any device following these guidelines would be DOA in the larger market, though. Even _if_ you could source things like LCD assemblies that fit your criteria, the expected density (high) and labor costs (low) people have for modern electronics don’t leave a lot of room for this kind of bespoke design.

I’m extremely impressed with how far Framework got on the path to “full repairability” using off-the-shelf parts without producing something that looks like a late-90s Thinkpad.

Don’t get me wrong: I love the old IBM, Tadpole, Toshiba, et. al. designs, but I don’t think that many people outside this community would be willing to pay 2-3x as much for something 2-3x larger and heavier and several generation behind in terms of performance. (Not to mention wildly incompatible with any non-hobbyist OS.)

OTOH, what you’re describing does sound a lot like why custom, hand-wired audio amplifier builders do. They manage to make a business out of it, mainly by appealing to folks willing to spend effectively unlimited money to get that “special something”.

Not sure what you meant by "full repairability", but for component-level system board repair, framework laptop is not far at all from current-gen lenovo or macbook. That's the whole point of OP's video.

User-replaceable RAM, SSD and keyboard are all steps in the right direction, of course. But it would be quite a stretch to give product full marks on repairability based on that alone.

The veting of repair shops is a ploy.

The watch industry used to vet repair Watch Repair shops on the guise of "Quality Assurance". They would require Watchmakers buy their expensive cleaning machines, and a bunch of other pricy tooling, and requirements; in order to procure parts. 90% of watchmakers couldn't get a parts accout, and the watch companies loved it. it meant more after sales/warranty business.

This was because they didn't want to sell parts to anyone. They wanted to do the repairs

For a few years, certain Watchmakers kept up with the demands from Rolex, and The Swatch Group.

We were just happy to be able to get parts.

About 15 years ago the watch companies just decided not to sell parts to anyone.

That whole Vertical Intregtation consumer ripoff.

I am very leary of any company that wants to vet their repair shops for anything, other than maybe medical equipment.

They will supply the schematics, but they don't want to make it easy.

(I'm pretty sure I'm hellbanned again?)

It is not them though, they clearly just follow rules here that they have to obey. But I think it is a great idea to compare it with the watch market - independent repair shops there have neven seen like a bad idea, and the schematics are pretty much always open - in that the movement is just there, accessible - and the watches are copyable. Which happens a lot though - might not be the most convincing argument if that were the fear.

> Now, maybe there's a great reason for this. Maybe they're under NDA and can't reveal part of the schematics. Maybe releasing these schematics would make it super easy for a competitor to completely steal everything. Maybe they are going to release the full schematics and we're waiting on their engineering team and legal team to review them.

I guess I'd question 'fear of competitors' being a "good" reason, but I'd agree with you otherwise. Regardless of what the need for secrecy is, unless it's "we're forbidden by the state" and there's some kind of national security letter involved or "We're doing something illegal and we'd get sued if it were known to the public", why not just tell the public why they can't release the full schematics? What else would prevent someone from talking about why they couldn't talk about something? Refusing to even explain themselves just makes it all seem more suspicious.

Sometimes NDAs prevent even revealing the scope of the NDA, because knowing the scopes of things is then potentially exposing information a party doesn't want known by others. Sure, maybe Framework could just go and state those things, but they're potentially burning their relationships with key partners that might be necessary to make the product at the moment.

I'm not arguing this is a good thing, just trying to share some perspective.

I think the difference is in the (perception of) honesty or hypocrisy. If Apple jealously guards details about their hardware and impedes repairs, it's bad but it's expected. If Framework guards details about their hardware and impedes repairs, it's in direct opposition to their claimed mission and the reason people were interested, and it feels like a betrayal. (Obviously this varies; not everyone cares about the same things or to the same degree. But I suspect that's this is where a lot of the dissent comes from.)

> . We should celebrate attempts to change the status quo

yes but... no. if you go half way to do something you will not reap much of the expected benefits. in certain use cases hard thresholds are important to consider. in this case half available schematics is not good enough to do repairs even if its better than no schematics at all.

But half way to something here is demonstrably better. The computer opens with ordinary tools and the parts are modular and available! It was also stated above that this repairer did not ask for the schematics.

> and they’ll complain just as loudly about you as they do the default (far worse) option.

This is optimistic. They'll lampoon you and call you hypocrites for claiming to be helpful but not doing the zillions of things they have in mind, while giving the unfriendly incumbent a pass.

I'm not sure who they is in this context but the person who runs the channel linked in the topic and who is normally in the videos has consistently praised Framework for their efforts, including the release of these public schematics, regardless of being partially complete, and has reminded viewers how it compares to other brands that don't release anything of the kind for helping repair shops.

In a recent video they were considering buying a second Framework laptop of their own just to have it displayed in videos as essentially free advertising. If some of these characterizations are about them it seems misplaced tbh.

> I'm not sure who they is in this context

Same "they" as the parent-- "for some subset of users you will never be good enough, and they’ll complain just as loudly about you as they do the default (far worse) option." Not implying that it's the channel's stance.

On any criterion: ease of repair, privacy, security, good citizenship, etc: there's some subset of critics who treat the best behaved companies and individuals worse than the worst. These are the people who e.g. give Apple a much harder time on privacy than Meta.

> (...) no matter how much better you are than the default, for some subset of users you will never be good enough (...)

What if it is really not good enough?

Should everyone just accept someone's claims that they addressed a problem when they might be actually making it worse by pretending it's solved when it isn't?

I recall the days of old when no laptop supported linux, and the impact there was when a model showed up with claims it indeed supported linux. I bought one based on those claims (don't recall the make-model, might have been an Acer) and I still recall the disappointment I felt when I discovered critical stuff like wifi did not worked at all. There were a few comments like "It boots, right?" followed by "for some subset of users you will never be good enough" when critical problems were pointed out, but the truth of the matter was that a purchase was made based on a feature that was advertised but was in fact not offered.

Similarly I could definitely see Intel having some trade secrets as you get closer the CPU.

If they're "secrets", they're probably one of the worst-kept in the industry.

You can already find plenty of other, probably leaked, schematics of laptops using the same if not very similar platform. They are all based on reference designs from Intel, after all. It's inevitable that yours will leak too, hopefully sooner rather than later.

Hence I am strongly siding with the "marketing BS" people here. Claiming to be so strongly in support of RtR, and then turning around and keeping the barely-existing artificial moat up, is practically not much better than all the other laptops that already have schematics available.

do you think Intel's lawyers would accept that argument? getting blacklisted (or sued) by Intel would be catastrophic for Framework

Hopefully Intel's lawyers realise that suing its customers who are buying its products would be a bad idea. If anything, getting Intel to open up more documentation --- like it used to have --- would increase sales and community reputation. Especially with its diminishing performance lead, and the rise of performant but even more horribly proprietary ARM SoCs, opening up seems to make even more sense. Remember that x86 and the PC architecture dominated the computing industry because it was open.

Hopefully may or may not be sufficient for them to bet the farm on.

Knowingly breaching a contract you've already invested many dollars and hours building and understanding, against the advice of your legal council, seems risky.

Additionally - while the legacy bemoth Intel is getting onboard with open source schematics (and figuring out its strategy to open source the VHDL for its 12th gen core to compete against RISC-V - perhaps now that Gelsinger who is more technical than the previous CEO - there is some hope!) -- the question is whether we'd rather see a product like Framework make inroads despite the roadblocks and build up marketshare to further the case toward right to repair.

Exactly.

It's really easy to make speculative quasi-legal decisions when it's not your company.

Agree - I think Intel would have a big say in what it wants released under no NDA. It's important to keep in mind - NDA's aren't always meant to prevent the average person from accessing it ever. In fact it'd be most likely primarily used to enable an Intel (or Framework) legal team to have some leverage to go for damages against the most extreme cases of abuse [1].

Also - w.r.t. to right to repair - I've always gotten from their marketing more of a right to repair at the iFixit level (swap SSD, memory, LCD panels, keyboards, etc.) versus de-popping the Intel CPU.

Module level repair novel you say?

Parts Removal and Replacement Videos - PCs and Smart Devices - Lenovo Support GB

https://support.lenovo.com/gb/en/solutions/ht505031-parts-re...

Interesting. That said - its one thing to have a knowledge base likely targeted toward corporate IT teams with direct access to order-able parts via Lenovo value added resellers and another to put it on the landing page for a consumer facing laptop.

It'd be a huge win in fact for right to repair if Lenovo did do this in response to Framework.

Thinkpads have always been very repairable at the module level since they were still called IBM Thinkpads. You can also find schematics for them on other sites.

Reply on news.ycombinator.com