Hacker News new | past | comments | ask | show | jobs | submit login
MacBook Battery Time Remaining (kaspars.net)
65 points by konstruktors on Dec 25, 2016 | hide | past | favorite | 59 comments

The time remaining indicator was really useful. It was more than accurate enough. It allowed me to see at a glance if I had enough charge to finish the movie, if it would last for the rest of the train ride, etc. If I saw that the battery wouldn't last as long as I had expected, closing Slack would fix it.

The problem with the time remaining indicator was that people started reporting the displayed time as actual run time. I read a couple of blog posts where people claimed that the battery life time of the Macbooks was reduced, based on the estimates they saw in the battery menu. They didn't actually time the run time of their Macbooks, presumably because that would take 5-10 hours, or longer if you want to repeat the measurement. So Apple is claiming that these people are wrong, the indicator was inaccurate, and they removed it.

Of course, the problem is that most likely battery life actually is lower. The battery of the new MacBook Pro is significantly smaller than the predecessor's, but the components don't use significantly less power. With the increasing popularity of Electron based apps, average power consumption is probably even increasing.

For the user, this is a huge loss. We've learned to live with the battery time estimate, and we know that the time remaining isn't absolute; just like we know that the remaining range displayed on the dashboard of most recent cars depends heavily on the driving style.

Fortunately, there are 3rd party tools that display the estimate Apple removed -- I downloaded Coconut Battery. But I wish we wouldn't have to rely on 3rd party tools for such basic functionality. Macs used to come with "batteries included".

> With the increasing popularity of Electron based apps, average power consumption is probably even increasing.

I can only hope OS vendors (both mobile and desktop) will one day be driven to actually incentivise frugal use of resources. We are currently suffering a massive tragedy of the commons for RAM, cpu and even disk I/O (see Spotify, recently).

I would be absolutely thrilled if, just for example, App Store margins were calculated based on median resource usage. Or if apps got a clear UI annotation indicating resource use. Or something... anything. Like a carbon tax, but for software :)

Currently it's the Wild West out there, the cpu tuna is being massively overfished and we don't put the blame with the right parties so why would anyone stop. We the consumers end up paying for all this.

Edit: and if google could do the same with search ranking I would convert to any religion you asked me to.

This makes sense for iPad or the 12" MacBook but not for MacBook pros. I invest a lot of money in a MacBook Pro for work and it makes sense the apps I'm using needs more RAM, CPU and graphics card usage.

If the thinness of current MacBook pros prevent them on displaying remaining battery time, then making the MacBook pros thinner was not a good decision. You can make the regular MacBooks thinner but not the MacBook pros -- unless Apple finds another battery material that outputs significantly more power per cubic cm

> I invest a lot of money in a MacBook Pro for work and it makes sense the apps I'm using needs more RAM, CPU and graphics card usage.

100% agreed. The crucial word there is "I". You choose. You get to decide what to think of wasteful apps. Text editor in Electron? Well you're a programmer, so ok. No hard feelings. Image editor using up 50% of your cpu? I'll look for a different one. Might be just the other way around for a graphic designer.

The important part is: let the market figure out what is the balance of acceptable resource usage. Right now, there is no accountability. Spotify was trashing the disk for about a year and finally got off their asses to fix it when it became a serious pr issue. That is not a sustainable system.

Edit: to drive the point home about your mbp: I have one too, and I actually don't think it's acceptable for apps to use a lot of cpu / gpu. I'm on battery a lot and I constantly monitor the tasklist to shut apps that are using too much. Already in n=2, you and I have different preferences. And that's great! Let us collectively figure out a balance like a normal market does :).

You are not getting OPs point I think. He's talking about efficiency not overall resource usage. When people code in electron and embed an entire friggin browser into their application just so that they don't have to port their webapp or so that they can apply their webdev skills, you the user gain nothing.

These apps don't do more, they do less but require more resources. They trade user convenience for developer convenience.

> Of course, the problem is that most likely battery life actually is lower. The battery of the new MacBook Pro is significantly smaller than the predecessor's, but the components don't use significantly less power

Exactly. Apple has always designed their devices to a battery life target. But as "wireless web" and "movie playback" both depend on use cases that get more and more optimized as Apple improve their software, they get more and more disconnected from the reality of pro users. How about a "editing a Swift Playground" battery life estimation? Or Photoshop? What Pro user spends all their day browsing the web and watching movies? Those make sense for the regular MacBook, not the Pro.

> Or Photoshop

Consistent 20-30% processor usage no matter what the processor when not in focus and no files open.

It's natural that Apple designs a battery target for one usage. It's totally ok to tell '10hrs watching a movie', with actual figures being lower when doing other activities. PC competitors never ever hit the target, even for movie-watching, so we were perfectly happy to have 10 actual hours for one usecase.

Of course I don't get what customers gain from .07 nineteenth of an inch less of thickness. Next version Apple will provide a thinner macbook, with no battery, and the power supply will come from an NFC table...

There are several non apple products for years with way better battery life. My 6yo T420 easily makes 12-14 hours on Wifi. More if i disable stuff.

I just wish they'd show figures for a use case that matters to me. If you keep 10 hours of "wireless web" mostly by optimizing your web browser, that's great for people who spend their time browsing Facebook, but if that means I lose 2 hours of my usage, I wish they had some number that showed that as well. Of course, I get that if they show a "heavy productivity - 3 hours" then pundits will latch onto that and compare it to other makers' web browsing numbers... It's a crap situation.

You have a point that many PC manufacturers have fantasy numbers - I don't want Apple to venture into that realm. ("10 hours... as long as literally the only process you have open is Safari")

Phrasing it another way... I hope Apple doesn't optimize their batteries for their own benchmark, to the detriment of real-world use cases.

> I don't get what customers gain from .07 nineteenth of an inch less of thickness

It is difficult to make thin, light and functional laptops. This results in fewer doing it, with those that do being able to command a margin for it.

It also amortises your product R&D. Minitiarisation continues, delivering not only more computing power but more in smaller form factors.

By focussing on thin and light, they can keep tweaking with decisions they had to make when putting a computer in a phone and then a watch and will need to face putting it into a contact lens, et cetera.

> PC competitors never ever hit the target

Start looking at proper business models maybe.

I've read several reviews by critics, as well as some users here on HN, that report that the new MacBook Pro battery is actually lasting quite a bit longer than its predecessor. I always found this surprising considering the current ongoing controversy.

I myself bought a brand-new 2015 model 15 inch, and I'm getting all day battery close to 12 hours. I've never had a laptop deliver such battery life before. And that's doing C++ development all day!

Sorry did you mean 2016 (if you did and correct it I'll delete this comment, so don't reply to it) which your context implies

2015 model, bought brand new a couple weeks ago. Apple made the smart move of still selling last year's model as part of the current product lineup.

So you would get 9 hours on this year's model (likely enough for you.)

This year's has give or take a 75 watt hour battery while last year's has give or take 100 watt hours, so to arrive at the above figure I quoted three quarters of what you reported.

Perhaps. Some reports by varied users really do claim improved battery life on the new 2016 model, so perhaps, especially with this last macOS update a few days ago, I'd actually get more than 12 hours now. Who knows, I ain't buying this first iteration of the 2016 model. But when I gave it a try at the store, I must admit I liked the feel a lot. The battery and graphics issues, and my unwillingness to deal with dongles (for now), and the extra 400 euros, led me to the proven 2015 design.

By the way 3/4 the battery wattage does not mean 3/4 the battery time. The new 2016 supposedly uses lower power consumption, especially with the display itself (which is where much of a battery goes). Which is why many people are getting longer battery life from a smaller battery (if you believe many of the reports out there, including by tech journals).

Have they fixed the issue that batteries dont stop charging once they are full? Otherwise those 9 hours wont be 9 houra for long ether.

Was this an issue in the recent 2016 model? All major laptops by all manufacturers have prevented overcharging in batteries for years. It has always been safe to leave a Macbook plugged in long after the battery was full.

This was ether introduced in 2014 or 2015 models. My 2015 MBP did suffer from it. Mostly used in a office environment it lost at least 10-20% of battery volume within a year.

Yes i also never heard of any other brand still having that issue.

On my phone on a bus right now. But there is a thread somewhere in a mac support forum that suggests to put tape over (i think) the 2 middle connectors on the charger which somehow fixes the issue.

Edit:// i dont use apple anymore so take that with a grain of salt. I mostly remember because the all mac company i worked for then was in rage. Thought about switching to thinkpads. And already replaced a few of those faulty things in mid 2016. I did not really care because it knew i just get a new one. I did however confirm it myself, it did not wait even 1% to charge again. Next to my battery dying faster than anything i've seen before.

Do you compile your code locally?

Yes. My workflow is emacs + clang.

The usage of Electron, while convenient, is continually irritating to me. Power consumption matters and "hur dur lets repurpose our web app for desktop" is annoyingly bad for power.

I get why companies do it, I just wish they wouldn't. Alternatively, perhaps an investment in electron as a platform in the power consumption space is what really needs to happen.

> The battery of the new MacBook Pro is significantly smaller than the predecessor's

I think I heard that Apple originally intended to have a different battery design than the one they shipped, and had to revert at the last minute. Perhaps it would have better battery life had that not happened.

There's Time Remaining indicator in the Activity Monitor, you don't need 3rd party tools.

Excuse me?!

That's how it worked for years and everyone knew that, and still it was useful to estimate the current workload of your machine.

Why all of the sudden something that worked for years isn't viable anymore coincidentally when people started complaining about new MBP's shitty battery life? Nobody's buying it.

Also the news today reports "Apple Working with Consumer Reports on Macbook Pro Battery Issue" [1]

Reading between the lines:

"Apple Pressures Consumer Reports to Change its Story"

You don't need to "work with" Consumer Reports to fix the issue. Just fix it.

Why "work with"? Does Consumer Reports have engineers that will help Apple understand the problem? Is Consumer Reports an isolated incident that just needs some tech support to help them understand the computer? Smacks of whitewashing to me.

[1] https://www.cnet.com/news/apple-working-with-consumer-report...

Imagine for a moment that you thought you'd made a good device. Now imagine that a respected organization did their own testing and said they couldn't recommend it.

Would you want to work with them to understand why you came to opposing conclusions?

There's no real chance of bullying, as CR buys the products they test, and has decades of experience with huge companies not loving their reviews.

Consumer Reports already sent Apple their logs on the test machines before the review. CR had already made a good faith effort to find the source of the opposing conclusions. However:

> In his tweet, Schiller linked to a story from iMore that says Consumer Reports was just going for a pre-Christmas headline and should have done more testing.

Apple is trying to bully/influence the public reception to CR's review. Gruber has spent more time dissing the CR review (and distracting from Gurman's expose on the Mac troubles which he hasn't linked to on DF)

Don't know how much you know about Consumer Reports, but I don't think they can't be pressured.

What if Safari attracted a heavy WebGL ad in their web browsing test? What if Chrome had an ad blocker in place? What if they used an automation system to drive the test which wired on the discrete gpu?

I see a lot of questions an engineer could answer to help Consumer Reports better understand why their test showed such large variances and then they can decide if it is a testing artifact or indicative of a real variance.

And yet none of these issues seemed to plague CR in the past, when they've reviewed every single MBP release of the last several years (in addition to hundreds of other laptops)...

Getting tired of people pointing out that "it was never accurate, it doesn't know if you're going to open a game in 20 minutes" as if it's some sort of massive revelation and justifies this move just to make up for their design problems with their latest laptop because the shaped battery didn't work out in time to ship.

As I see it, as the difference between high power and low power states increases, the accuracy of a "time left" predictor decreases.

Back in the day, "active" was one power draw, with a somewhat lower one for idle. And people intuitively understood that idle was not part of a timed test.

Now idle is significantly less power. Discrete GPU is vastly more. Ill behaved app waking from idle only to sleep again is a huge hit. Most of these behaviors are invisible to the user.

You can't just say "at last minute's usage rate, each percent of battery is 6 minutes" and multiply it out. The wildly varying input gives a wildly varying output.

The old algorithm with the new inputs was confusing people.

It's not like they are stopping anyone from watching movies or surfing the web and timing how long a computer runs. The truth is still out there.

Having used and trusted the "time remaining" calculation in Macs for years, it's very hard not to be suspicious of the timing here. A series of less-than-stellar reports of battery life in the new MacBook Pros, and then all of a sudden the indicator is inaccurate? Nope, not buying it.

I know why apple hasn't, but I'd kill for a $1500 macbook stuffed into a macbook pro case, and all of that spare space taken up by battery.

Give me a core-m, 8GB RAM, a high res display, and a massive battery.

All these years I've wanted a lightweight 15" (or 17") MacBook/Air but Apple insists on segmenting the larger laptops at more expensive prices. Apple won't even offer a 15" MBP without the dGPUs now.

Apple should simply offer two lines of laptop:

11"/13"/15" MacBooks optimized for thinness

13"/15"/17" MacBook Pros optimized for power

It would weight like a similarly-sized coffee-table book, you'd need a cart to drag it around.

But yeah, overall Apple should give us more options. It was fine before with dedicated lineups - Air for battery-life and portability, Pro for ultimate power. The current lineup makes both segments pay more while giving them stuff they don't need.

I would also settle for a battery pack which charges the MacBook at full speed over USB. I know the latest one charges from battery packs, but they can usually provide 15-20W, so it's a bit from the charger's 65 or 85W.

Is this really news to anyone? Of course it's hard to estimate, the same goes for MPG/miles until empty in a petrol or diesel car.

That doesn't make the measurement of current estimate useless. It's still a useful tool. Users understand if they close windows in their browsers, or shut down Battlefield 958 that will prolong their battery life and that if they turn up screen brightness that will cost them battery life.

Apple chose the lazy way out of this problem. The estimate was not any more inaccurate than it ever was in the past (at least I haven't seen any evidence that it was), and user's aren't any dumber. Apple just didn't like the press and made a stupid change to try to correct that issue. Lame.

Love that Apple keeps adding "features" like this. Not even SlimBatteryMonitor still works thanks to this update.

What a time to be alive, watching Apple choke on their own arrogance.

Their one shot at a big comeback is the iPhone 10th anniversary edition. If it is as much of a smash hit as Robert Scoble keeps heralding then Apple will be golden for another decade.

This hilarious Scoble piece?


I wonder what he has been smoking.

I love this sentence:

"You pop it into a headset which has eye sensors on it, which enables the next iPhone to have a higher apparent frame rate and polygon count than a PC with a Nvidia 1080 card in it."

That's quite an insult to pot. Robert Scoble is just an idiot.

Definitely not pot. Probably bath salts.

You can smoke things other than pot...

> "You pop it into a headset which has eye sensors on it, which enables the next iPhone to have a higher apparent frame rate and polygon count than a PC with a Nvidia 1080 card in it."

Interesting. Ive said that the touch bar was the first step in a new paradigm. Maybe the next step is connect your iPhone as the display. That would explain why Apple stopped investing in displays.

I'm using Battery Monitor (http://www.rockysandstudio.com/index.html#freeapps) and it works great.

Of course, it would've been better if I didn't have to install a third-party app and I had one less icon in my menu bar, but whatever.

pmset -g batt

Bonus points: easy to add in tmux-powerline running fullscreen in iTerm2

Maybe it'll be a special collector's piece, like the Twentieth Anniversary Macintosh.

Bring on the machine learning battery gauges?

if it runs on the GPU it will have to take its own power draw into account

> if it runs on the GPU

It will struggle to function at all seen as no Apple machine ships with a decent CUDA card.

Have they never heard of averages?

Just show time remaining based on last 5 minutes, 1 hour, 5 hours, and all data.

Or does apple not like that kind of "complexity" and wants just a single number?

Average makes bad estimations for non linear phenomenon. And nothing can be done about it.

Math 101.

And what could be better? Well, nothing we don't have the math for that problem in the first place. If we had, we could predict weather accurately with a time window bigger than 24hours...

24 hours is the accuracy weather forecasts are having since decades even with Moore laws multiplying the power of processors...

Even by throwing more money and engineers at them some problems cannot be solved.

The only way to average non linear phenomenon is to scale the problem to an extra large population or make prediction on ultra short time.

Actually I guess that if users were seeing there battery lifetime sucked dramatically every time they use WiFi, or GPU it would create an adverse effect on computer users of making them anxious (human beings are weird animals) and some of them would use less their computers.

Imagine if SUV were showing an estimation of mileage dropping sharply when you make a big acceleration? People would stop accelerating like mad men, hence it would kill the SUV/sport car market. You don't want to kill a really good and rational economy, do you?

Sellers don't want you to be economical in your use of anything, else it kills the market as sure as a pair of crocs lasts for years.

We live in an era where consumers and producers want to use moar resources, moar energy and feel good about it, not anxious.

Does the planet have limited resources, is it irresponsible?

No! Let's throw more money and engineers at a problem and we surely can solve it.

I guess immortality and super powers are just around the corner. lol.

>> Imagine if SUV were showing an estimation of mileage dropping sharply when you make a big acceleration?

That's how it works with any vehicle that has a real-time MPG readout

Oh, easy - we're just holding our macboocks wrong, that's messing with battery estimates.

Exactly. It's like a car manufacturer saying how many miles to the gallon you're going to get, without knowing that you're a terrible driver who never gets out of 2nd gear ;)

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact