Mastodawn

I swear people don’t understand planned obsolescence. PLANNED is the key

If I stop supporting a 6-year-old version of something because I have data suggesting hardly anyone uses it anymore, that’s just garden variety obsolescence

If I plan on day one to stop support after 6 years, THAT is planned obsolescence. Same if I just make design decisions such that most of them will break before then.

If I deliberately break old things, that’s not planned obsolescence, that’s FORCED obsolescence.

calcifer 🧯Aug 20, 2024

Forced obsolescence is mostly illegal and ought to be completely illegal.

Planned obsolescence is grayer. If I note that 95% of people replace their fridge within 10 years, so I decide to save money by only building fridges that last around 10 years… that’s not necessarily a problem (yes, we can get into how disposable culture is a problem, but I could write a whole paper on the complexities of blame there).

It’s definitely a problem when it’s more like hey I noticed people will not be upset or buy a different brand in the case a fridge only lasts 8 years, so let’s only make ours last 8 years so we can sell more.

FoolishOwl Aug 20, 2024

@calcifer I've thought that Moore's Law was really about the CEO of Intel making a public declaration about the rate of planned obsolescence throughout the computer industry. Since Intel had an effective monopoly on key components, it was in a position to impose that rule.

calcifer 🧯Aug 20, 2024

@foolishowl close. It was a genuine observation, but then Intel set it as an engineering target.

But that isn’t an example of planned obsolescence. The chips don’t stop working or stop being supported; Intel didn’t do anything to artificially make it harder to use old chips or anything like that. Simply releasing new things in the hopes people will upgrade is not planned obsolescence

FoolishOwl Aug 20, 2024

@calcifer More in line with your original point, it's obvious that with mobile phones, they're deliberately designed to shorten their lifespans. They used to have replaceable batteries, removable microSDs, and operating systems that could be rooted and modified or replaced. The batteries are particularly egregious: batteries inevitably wear out.

calcifer 🧯Aug 20, 2024

@foolishowl yes, non-user-serviceable wear parts is often a good example of planned obsolescence (sometimes there are other reasons for that choice, but 90% of the time PO is the controlling factor)

calcifer 🧯Aug 20, 2024

@foolishowl the non-rootable OS isn’t really though; the controlling reason to avoid that is security (mostly for the benefit of the cellular providers, but also for corporate adopters). Stopping updates at a planned time is PO, as is carriers blocking old hardware.

FoolishOwl Aug 20, 2024

@calcifer Points taken.

(I'm still sore that I selected my current phone based on its manufacturer's reputation for enthusiastic support for rooting, only to find that they'd reversed policies before this model came out.)

✧✦Catherine✦✧Aug 20, 2024

@calcifer @foolishowl I don't agree at all. I run my own firmware on the phone I'm typing this on, and it has all the same signing & verified boot mechanism as Google's. technically it could still give me superuser access, though I chose not to enable that personally

David Chisnall (*Now with 50% more sarcasm!*)Aug 21, 2024

@calcifer @foolishowl It’s hard to support installing user keys in a secure way, but not impossible. A lot of Android devices require you to first connect a computer and enable debug mode, which requires the ability to unlock the phone. Then the transition to an unlocked bootloader also wipes the device, which ensures that unlocking the bootloader does not compromise any data installed on the device before it is unlocked. This process could easily be augmented with a new process for installing a new key for checking the loader.

For corporate customers, this might even be better because it would allow their IT departments to ensure everyone ran a blessed version of the firmware, rather than the vendor’s version with pre-installed crapware.

calcifer 🧯Aug 21, 2024

@david_chisnall @foolishowl the user keys aren’t the issue; support cost is. Providers and phone makers have a relationship, and providers have a vested interest in making sure that it’s difficult for their customers to make modifications to any part of the phone that controls the interface with the cellular radios/phy/network. Moving those things to software and getting sign off from providers involves a commitment to protect that.

It’s very hard to support that commitment while leaving the hardware open to OS/firmware replacement

calcifer 🧯Aug 21, 2024

@david_chisnall @foolishowl in the case of IT departments, they already have the control they want: a blessed version, signed by a vendor, and remote controls that let them lock that. That’s a huge risk transfer for IT; no one wants to take on the support costs of having to manage every phone update, nor the risk of having BYOD with user-installed OSes of dubious origin.

I’m pro having paths to replace an OS on any device, but we can’t ignore the very real problems the current state exists to solve.

Denys F.Aug 20, 2024

@calcifer excellent point, thank you for explanation!

Sophie Schmieg Aug 20, 2024

@calcifer and there is a big difference in how that is communicated as well. For example, if I support a draft standard while the standardization effort is ongoing, and from the very beginning say that I will deprecate and replace with the actual standard once it is ready, then there is IMHO nothing objectionable happening, even though I planned the obsolescence from the start.

Arthur David Olson Aug 20, 2024

Me getting duller with old age is bland obsolescence.

SpaceLifeForm Aug 20, 2024

The windows are clear.

Andrew Aug 23, 2024

@calcifer there’s also the separate thing of “having a design lifetime in mind” which is just how hardware engineering works