Warning: Some posts on this platform may contain adult material intended for mature audiences only. Viewer discretion is advised. By clicking ‘Continue’, you confirm that you are 18 years or older and consent to viewing explicit content.
I appreciate the thought, but I think you’re giving the concept too much credit, and also misunderstanding exactly what Apple did or why it was bad.
“Graceful degradation” is simply the existence of a wider range of failure modes. The flashlight is nice because there are more conditions where you can do something with it, but the life cycle of such a product is obviously not limited by the replaceable batteries.
Apple’s hidden power management hacks were also, in fact, an example of “graceful degradation”. As a lithium-ion battery degrades, high-amperage loads (i.e., the the processor when executing an intensive workload) will cause an increasingly large voltage drop. If the voltage supplied to the processor drops too low, the latches inside the processor will destabilize and begin to produce incorrect results (a 1 that should have been a 0, or vice versa). This is immediately catastrophic for obvious reasons.
Given this, you have two choices: either the device shuts down when the voltage drop becomes too large (at, e.g., 40% charge, depending on the specific properties of the battery), or you reduce the maximum current draw of the processor by reducing its clock frequency.
Apple chose the latter, which probably makes sense in the grand scheme of things. However, this was still pretty bad for two reasons: they didn’t inform the user that they were doing it, and first-party battery replacements were prohibitively expensive until recently. Because of this, most users would assume that their phone was slowing down because it was old, not because their battery could no longer supply adequate power to sustain the maximum clock frequency. Worse yet, even if they did somehow figure this out, it was rarely worthwhile to shell out the $130+ Apple was charging to replace the battery (which basically just involves removing two screws and a ribbon cable).
The other problem is they didn’t give a choice to users. If I recognise I’m only going to keep my phone for another 6 months then I might prefer to just run the risk of a failure while maintaining high CPU function when the battery has sufficient charge.
And of course it wasn’t really a safety issue it wasn’t dangerous for the device to fail it was minorly irritating you just start the device again. So they basically made a unilateral decision on everyone’s behalf without asking anybody or telling them what they had done.
Apple got sued for lack of communication essentially.
Given this, you have two choices: either the device shuts down when the voltage drop becomes too large (at, e.g., 40% charge, depending on the specific properties of the battery), or you reduce the maximum current draw of the processor by reducing its clock frequency.
Yep, Apple took one route and Google took the other. There wasn’t a great solution short of replacing users batteries which no company is going to do without being forced.
I appreciate the thought, but I think you’re giving the concept too much credit, and also misunderstanding exactly what Apple did or why it was bad.
“Graceful degradation” is simply the existence of a wider range of failure modes. The flashlight is nice because there are more conditions where you can do something with it, but the life cycle of such a product is obviously not limited by the replaceable batteries.
Apple’s hidden power management hacks were also, in fact, an example of “graceful degradation”. As a lithium-ion battery degrades, high-amperage loads (i.e., the the processor when executing an intensive workload) will cause an increasingly large voltage drop. If the voltage supplied to the processor drops too low, the latches inside the processor will destabilize and begin to produce incorrect results (a 1 that should have been a 0, or vice versa). This is immediately catastrophic for obvious reasons.
Given this, you have two choices: either the device shuts down when the voltage drop becomes too large (at, e.g., 40% charge, depending on the specific properties of the battery), or you reduce the maximum current draw of the processor by reducing its clock frequency.
Apple chose the latter, which probably makes sense in the grand scheme of things. However, this was still pretty bad for two reasons: they didn’t inform the user that they were doing it, and first-party battery replacements were prohibitively expensive until recently. Because of this, most users would assume that their phone was slowing down because it was old, not because their battery could no longer supply adequate power to sustain the maximum clock frequency. Worse yet, even if they did somehow figure this out, it was rarely worthwhile to shell out the $130+ Apple was charging to replace the battery (which basically just involves removing two screws and a ribbon cable).
The other problem is they didn’t give a choice to users. If I recognise I’m only going to keep my phone for another 6 months then I might prefer to just run the risk of a failure while maintaining high CPU function when the battery has sufficient charge.
And of course it wasn’t really a safety issue it wasn’t dangerous for the device to fail it was minorly irritating you just start the device again. So they basically made a unilateral decision on everyone’s behalf without asking anybody or telling them what they had done.
Apple got sued for lack of communication essentially.
Yep, Apple took one route and Google took the other. There wasn’t a great solution short of replacing users batteries which no company is going to do without being forced.