This [0] from this morning was interesting. Apparently AppleInsider has been monitoring a number of Genius Bar and authorized service centers for warranty repairs of 2014 thru 2017 laptops. It seems 2016 MBPs had the most keyboard issues, and the 2017 version brought the claims back down to (and below) 2014/15 levels.
So, it might be reasonable to say they fixed it last year, in terms of prevalence of issues, although absurd cost to repair issue remains; combined with PR, probably explains this year's mod.
Anytime there's a new chassis launched you have to expect greater failure rates on the early parts, and the 2016 model was the first one with a new chassis, so I'm not sure the data they have tells the entire story.
It's likely there were manufacturing issues that were still being worked out throughout the 2016 cycle and that there were very minor tweaks made during that year's run to improve reliability. If that were the case, the issues would be weighted toward 2016 models manufactured early in the year, and we'd only be able to know if we compared return data to build date.
In my personal experience with a 2016 MBP, one of the issues with the keyboard was that when the system was under load for a significant amount of time, the resulting increase in temperature around the left side of the keyboard would cause keys to stick. The keys would unstick after temps decreased. This was an issue discussed by several people in forums. A manufacturing change to something like the application of thermal paste on the CPU/GPU might change that, without any change to the keyboard design itself.
I’d be interested to see a breakdown of reasons the keyboards failed for each model year.
And what percentage of the pre butterfly design were because of sand/dust etc. then contrast that to the butterfly design percentage. They could have merely improved reliability in another way(s) with the butterfly design.
That would certainly be interesting -- from your other comments comments in this discussion it seemed like you already knew the answer! Guess I assumed too much.
That's literally the point of the article - the design was changed in a significant way to incorporate a fix to reduce dust, which is spelled out in the patent.
I fail to see how what they did is change “in a significant way”. It seems more of a bandaid to me. It almost signals an unwillingness to fundamentally change the keyboard design for the better
Well one person's band-aid is another's significant fix, and only the data will tell.
EDIT: Below is wrong, but keeping it for posterity:
Note that the desktop Magic Keyboards that employ the butterfly mechanism do not seem to have the same key sticking issues as those on MacBooks, so it's likely they felt the fundamental design of the mechanism was fine but that it did need an additional silicone barrier for the lower-travel MacBook variant.
That’s the question raised here: Apple has said that they’ve made the keyboard “quieter”. The jury’s out on whether that was codewords for fixing the issue.
