Should I leave the MBA plugged in when it’s at my desk, regardless of battery level?

Al Dente piggybacks off all that work. It just sets charge target key in the SMC instead of letting the system’s machine learning battery health management pick the number. The rest of the app is just UI for the user to pick the number. The system continues to handle the voltage management. Pretty clever, and probably not much smaller of a team than the people who own that SMC key and battery health checkbox at Apple.

On the product design side, this is just meeting the needs of people for whom one-size-fits-all machine learning isn’t working. Both approaches are based on the same li-ion science.

1 Like

By definition machine learning is not one-size-fits-all.

How do this developer or its customers know more than Apple does about setting the charge target?

2 Likes

I have to say, I generally agree with this sentiment.

That said, Apple neglected this area for such a long time that it became something third-party devs at least initially did do much better than Apple.

And it wouldn’t be the first time that Apple’s default implementation of something…erm…was less-than-optimal for some use cases. :slight_smile:

So for me it’s not a matter of knowledge, but a matter of whether or not they’ve made a good implementation of the knowledge they have.

You can see in this thread that it can’t figure out everyone’s usage patterns. Just as it doesn’t understand how to smartly charge my phone when my sleep schedule becomes inconsistent. Since macOS doesn’t give the user configuration options, the ML model has to solve every usage pattern without input and it isn’t capable of that. Apple, probably correctly, chose to ship one method that worked best for the large majority of its users, and this niche product fills in the gaps.

No one has said this. The developers are providing an alternate method to achieve what Apple’s method is trying to do. It’s better or worse depending on how well the system method understands the user. And yes, it does give users power to make worse decisions than the system would.

Generally, may I say that hating on a responsive indie developer for giving users options they want is out of character for this forum.

3 Likes

I think “hating on” is a bit of hyperbole.

But I do think there is a notable difference between software hacking with no risk of harm to the machine, vs hardware hacking where you truly can damage the machine.

1 Like

I explicitly did not ‘hate’ on the developers (whatever that means) – I recognised that they may be very talented and well-intentioned, and good luck to them.

But, there may be a risk of damage to the machine, and I have no way of quantifying that risk. If Apple gets it wrong they have to replace my machine (within defined limits). What happens if the Indie developer gets it wrong? Will Apple still replace the machine?

For me, the possible (probably minor) benefits don’t outweigh the risks. If you or the developers can provide a guarantee that there will be both significant improvements and no increased risk of damage, then I will be genuinely happy to reconsider.

1 Like

Fair enough! I guess I’d point again to what I wrote about the SMC key. I don’t really see how you could conclude that is dangerous. But I don’t even use Al Dente, so I’ve probably posted too much about it. :slight_smile:

Fair enough! I guess I’d point again to what I wrote about the SMC key. I don’t really see how you could conclude that is dangerous.

Because I don’t have the expertise to know whether it is dangerous or not, of course. How many users genuinely have the technical expertise to know whether changing an SMC value can be risky?

Which brings us back to the issue of trust: I trust Apple’s engineers enough to have given them over £2k of this machine (and I’ve never had battery problems by following their advice with four other Macbooks over the years). I don’t decry people who do understand the risks and make another decision. :slight_smile:

Heh, yes. That gets to the part that makes me sad. Would love to figure out how to write in a way that is more helpful to other users here.

1 Like

FWIW, I agree with the rest of your logic - but I don’t actually think this is true…at least not for any “defined limits” that would be meaningful to any end user. I think the issue would have to be massively widespread and/or you’d have to be able to prove that Apple’s software broke something in order to get any relief from Apple.

Consider the butterfly keyboard. A specific design issue from Apple went almost completely unaddressed (in a “will you accept responsibility and fix my laptop” sense) for three years, and even then I don’t think the replacement parts actually address the issue - they just replace the existing, defectively-designed part with a new, defectively-designed part.

Not arguing that devs are smarter than Apple. Just questioning, given past performance, whether Apple would actually 'fess up and take care of such a machine outside of AppleCare (and if the AppleCare deductible would apply).

1 Like

Apple would replace a machines damaged by a swollen battery during AppleCare, and under UK consumer legislation, you have additional rights for up to six years. Of course, for that you do need to show that the product isn’t fit for purpose, and the longer you get into the six years, the more proof you need, but still, I don’t think it would be particularly hard to show that a swollen/unworkable battery wasn’t fit for purpose within, say, 3 years.

My MacBook 12" butterfly keyboard was replaced free of charge by Apple after nearly two years (not under AppleCare), without complaint, when the spacebar started to fail. In fact, I didn’t even have to mention consumer legislation, they just offered it, and this was well before the formal replacement policy.

I’m not worried about whether the battery will last longer under either approach (after all, I’ll never know one way or the other): the concern is of serious damage. But if Apple have got the design so badly wrong I end up with an unusable machine, I have some avenues to pursue, which may be invalidated by tinkering with the hardware. That’s all.

I had the same experience. Had the battery replaced, then traded the machine in. Still nervous about clamshell…

1 Like