12 ms·

Cars do get weaker over time as the parts wear. And I would certainly expect the microcontroller to adjust to provide the best performance possible under the ne

Cars do get weaker over time as the parts wear. And I would certainly expect the microcontroller to adjust to provide the best performance possible under the new limitations. I assume most well-made, modern cars already do that.

dozzie 9y ago
> Cars do get weaker over time as the parts wear. And so does the battery life get shorter as the battery wears. We know that. This in no way allows to anticipate that CPU gets slower.
lloeki 9y ago
Well you'd loose maybe some couple percent tops over a dozen years but basically nothing more on a properly maintained car (regular oil change, spark plugs, bearings, whatever). Here we're talking about halving the device performance after a single year because the thing is designed on such a tight rope that it can't handle the slightest voltage drop which happens to come way too soon because the battery wears out so fast. From an engineering perspective it looks like the latest devices are designed so that they provide blistering performance out of the box, satisfactory performance 6 months in and something ranging to barely adequate to crapshoot 1 year in. That is, picture a graph f(time) with a horizontal line that defines "This device is OK" WRT your favorite metric (performance, battery life, whatever) on the vertical axis, and it seems to me that the device should be designed to stay above this line long enough, but when objectively looking at it, it appears that the line is crossed way too soon, around the 6 month mark. That is when I started to notice my phone wasn't quite holding up all day, that there were performance glitches here and there that started to annoyingly grow on me, but I was dismissing all of this because hey I shelled out xxx€ for this device, and it can't break down that so fast, because Apple, so it must be all in my head or something. I cannot imagine Apple not having such graphs when designing a device prior to release, nor after release with all the telemetry they gather. That is, to me, where the fraud can lie: a design where the device produces delightful performance for a known short while, followed by a growing frustration to be released again through delight when a new device is bought. The question is, is this premeditated?

dozzie 9y ago

> Cars do get weaker over time as the parts wear. And so does the battery life get shorter as the battery wears. We know that. This in no way allows to anticipate that CPU gets slower.

lloeki 9y ago

Well you'd loose maybe some couple percent tops over a dozen years but basically nothing more on a properly maintained car (regular oil change, spark plugs, bearings, whatever). Here we're talking about halving the device performance after a single year because the thing is designed on such a tight rope that it can't handle the slightest voltage drop which happens to come way too soon because the battery wears out so fast. From an engineering perspective it looks like the latest devices are designed so that they provide blistering performance out of the box, satisfactory performance 6 months in and something ranging to barely adequate to crapshoot 1 year in. That is, picture a graph f(time) with a horizontal line that defines "This device is OK" WRT your favorite metric (performance, battery life, whatever) on the vertical axis, and it seems to me that the device should be designed to stay above this line long enough, but when objectively looking at it, it appears that the line is crossed way too soon, around the 6 month mark. That is when I started to notice my phone wasn't quite holding up all day, that there were performance glitches here and there that started to annoyingly grow on me, but I was dismissing all of this because hey I shelled out xxx€ for this device, and it can't break down that so fast, because Apple, so it must be all in my head or something. I cannot imagine Apple not having such graphs when designing a device prior to release, nor after release with all the telemetry they gather. That is, to me, where the fraud can lie: a design where the device produces delightful performance for a known short while, followed by a growing frustration to be released again through delight when a new device is bought. The question is, is this premeditated?