Kind of. I mean, sure, I think it would be really neat to be able to replace the L1 cache myself, but this would probably necessitate a less efficient CPU. Is it fair to fault Apple for using an Intel CPU which itself has no mechanism for internal repair/replaceability? I think it is if we were able to apply the same standard to every product (this is where the VAT idea came from); then the repairability becomes the aggregate of all the parts used, and it's the manufacturer's job to pick repairable parts, but how far do you go?
Does it make sense to lower the repairability score because I can't replace the L1 cache? If the answer is "no", then what is the difference between an L1 cache replacement and a CPU replacement on a fully integrated motherboard (like Apple's new computers)? The answer in my mind is one of convention. Perhaps the solution to this is to enshrine the Von Neumann architecture as a legal standard. But this solution seems like it could limit innovation.
What about smaller components? If we follow this to the logical conclusion we cannot use integrated circuits at all, as each individual transistor must be replaceable.
The rules needs to be written in a way that allow things, like integrated circuits, to exist but at the same time prevent intentional preventing of repairability.
> The answer in my mind is one of convention. Perhaps the solution to this is to enshrine the Von Neumann architecture as a legal standard. But this solution seems like it could limit innovation.
Right - I think this is the issue. People cite cars as an example, but car architecture evolves much more slowly than computer architecture.
Arguably at the present, as we come up against Moore’s law limitations, architectural changes should be expected to increase.
Government mandated computer architecture sounds like a great way to completely cede technological advance.
To specifically answer your question, I don't think anyone's advocating for users to be provided with parts and schematics to repair a faulty L1 cache on a CPU, or for the CPU to be designed to support that kind of repair.
Most "repairs" come in the form of replacing a component. Many products can be divided into components and those components can recursively be broken down into smaller components. At some point, though, a component cannot reasonably be further divided for the purpose of repair and peterlk appears to imply that a CPU is an example of such a component. Indeed, repairing a faulty L1 cache would probably require so much investment that even the manufacturer itself would not be able to justify such a repair.
> My point is that at some point, things start to get blurry, and it is in the manufacturers best interest to continue blurring these lines.
peterlk goes on to point out that manufacturers often use this fact to their advantage by convincing people that it's not practical to componetize bigger, more expensive parts for the sake of repair, such as the main board assembly on a laptop. In reality, component-level board repair is a practical solution for many repair scenarios, provided you can source the necessary components and schematics.
We lack objective criteria to determine what parts can reasonably be made "repairable" via componetization and manufacturers abuse that to confuse people. I myself thought component-level board repair was a dead end until I stumbled upon Rossmann's channel. I think both you and peterlk are getting to the same point: we need that criteria if we're going to reach the full potential of the right-to-repair movement.
Do they? Given that you later claim we don’t have the necessary criteria to determine repairability, this seems like an unsupportable claim.
I hear just as much from people obsessed with repair that confuses people in the other direction.
> I myself thought component-level board repair was a dead end until I stumbled upon Rossmann's channel. I think both you and peterlk are getting to the same point:
> we need that criteria if we're going to reach the full potential of the right-to-repair movement.
We do and we also need to understand the costs at scale.
Repairability may not be better overall if it comes at the expense of durability, or more use of materials.
It’s not at all obvious, for example, that we have the numbers to support say, making iPhones more repairable vs simply designing them to neeed fewer more expensive repairs and then recycle them when failed.
> Do they? Given that you later claim we don’t have the necessary criteria to determine repairability, this seems like an unsupportable claim.
They absolutely do. We don't have objective legal criteria, but Louis Rossmann's channel and business is a testament to how component-level board repair is not only viable, but very profitable. Apple (or most other computer manufacturers, for that matter) will never offer you something like that.
I think that criteria is necessary for legal action. Otherwise, right-to-repair could be dead on arrival (eg France's implementation) or go way too far (eg requiring a CPU's L1 cache to be replaceable). However, even without that criteria, it's plain to see that manufacturers are minimizing repair options as much as possible to extract profits.
> I hear just as much from people obsessed with repair that confuses people in the other direction.
Okay? There isn't any issue for which I can confidently claim to agree with everyone in the same camp as me. Right to repair is no different. I can only defend my own thoughts and statements.
> Repairability may not be better overall if it comes at the expense of durability, or more use of materials.
> It’s not at all obvious, for example, that we have the numbers to support say, making iPhones more repairable vs simply designing them to neeed fewer more expensive repairs and then recycle them when failed.
I agree completely. I do not want us to enact legislation which forces manufacturers to sacrifice quality and progress just for the sake of repair.
That said, there is clearly a lot of room for repair to become more accessible without changing designs or adding hardly any cost to the manufacturer. That is what I'd like to see addressed with right-to-repair.
Are you suggesting the right to repair should include this?