I worked on an embedded system with an Allwinner chipset. We tried to reduce power consumption, not to save battery since it was a cabled system, but to reduce heat. It turns out nobody, not Allwinner who provided the BSP, not the board designer, nor the final customer who developed the application software cared to optimize the OS much. The CPU had four cores, but only one was occupied most of the time. But all cores were at 100% frequency. Configurable voltages were also all on the upper edge. I enabled frequency scaling, switched to the correct scheduler, and now the board was much cooler and ran with more than twice the performance.
I'm always surprized how many low-hanging fruit there are in these kind of systems.
It would become very easy if they realized that a lot of users who buy a Linux phone would probably be happy to trade thinness for battery life.
I’d even go as far as saying that even with Apple’s efficiency and all, I’d still prefer it if they had bigger batteries.
And those are a lot thicker. Maybe Librem should not try to compete with Android and iOS at all in terms of design and just go for “privacy by default” as their campaign. (If they ever sort out delivery problems)
It sounds like Librem and Pine, if they haven’t already, should do the same and create a battery life optimization team, responsible for coordinating that effort across hardware, software, internal teams, and external volunteers.
I’ve read somewhere recently that in order to keep the phone mostly free of proprietary firmware, Purism had no choice but pick lots of discrete components. That would be in contrast to most other smartphone designs with more closely integrated chipsets, they wrote.
That discrete-ness, according to the author, is likely to be an upper bound for battery lifetime on the Librem 5. After all, all those chips have to be powered at least part of the time, and that allegedly consumes more energy than a single package would.
In other words: the Librem 5 may never gain a decent battery lifetime. As drivers mature, battery usage may improve a little. But not much. Buyers may want to keep their hopes realistic.
This is only a half the story. The Purism still runs on proprietary firmware. The discrete components in question are an actual full auxiliary CPU core, its only purpose being to load the hard-coded blob to properly boot the primary SoC. This was done because under FSF's definition, if the end user cannot update the firmware blob, then the firmware is considered one with the hardware, and the hardware as a whole is "free enough".
So they took away the user's ability to update the firmware, fused it in a ROM with every possible bug and inefficiency frozen in place for the rest of eternity, wasted silicon and engineering time to do so, only to grant themselves an arbitrary, honorary badge.
This isn't even radicalism anymore, this is hypocrisy. As a power-user who values freedom, and a long-time Free Software sympathiser, I am personally offended, and won't give my money to either party until they reverse course on their user-hostility.
That's simply not true. Users can upgrade those firmwares if they want (and absolutely no weird tricks like disassembling or soldering are necessary for that). PureOS doesn't distribute any non-free updates, but if you want, you absolutely can reflash these blobs.
> The RYF has a “secondary processor” exclusion that can be granted on a case by case basis. We will leverage this exclusion to load and train the DDR PHY on the i.MX 8. We will use a secondary processor to keep binary blobs out of u-boot and the kernel.
Librem 5 does not lock the firmwares in any kind of ROM. Nothing is taken away from the user.
To be able to conserve battery apps works differently than programs, apps can be suspended. That is usually the problem with normal programs, they are not developed with battery conservation in mind.
I wonder how Librem have solved this, perhaps in their scheduler, or intends to solve this in the future.
https://developer.puri.sm/Librem5/Apps/Guides/Design/Constra...
As for apps being suspended, most apps are suspended when there's nothing to do. If a graphical application is minimised so it doesn't have to redraw the screen then it should either be doing nothing or occasionally polling a server if that's it's design.
Web browsers are an interesting corner case as web sites often have JavaScript that wants to run all the time and there's some trade-off between doing what the web site wants and saving CPU/energy. But that's probably not going to be an OS issue for PureOS but an Epiphany browser issue.
https://developer.apple.com/documentation/uikit/app_and_envi...
https://developer.android.com/guide/components/activities/ac...
I'm not an app expert nor an expert on GNOME development either, but I got a bit sceptical when I read their app example code, python with GNOME, neither is famous for being snappy.
For the moment the Librem 5 seems to be using apps designed to work on PC desktops but at lower resolutions.
Sadly I cannot confirm this for my 6+ different laptops I owned. And I tried all the crazy settings, grub, tlp, .. and even compiling the kernel myself.
It is hard to beat loads of dedicated engineers, who are paid regular and well and have access to all the proprietary device information and even manufacturing.
edit: maybe there was one time, when a linux stock install performed maybe equally than a stock windows install. But that was only because of the windows bloat, which linux does not ship. But I can easily remove most of the bloat, but I cannot just write a better gpu driver. But if windows continues its bloat path, it will be soon inferior despite way better drivers.
Google also invested a lot of time into optimizing battery consumption (which hackers and people like the guy from Commonsware derogatory call "War against background processing"). If you look up through history of Android releases, there isn't a single release where there would't be a pretty major change in how Android puts device to sleep and how it wakes it up again.
That stuff is really hard since a single bad service can drain your battery in matter of hours and needs seriously tight coordination between all software makers on your device to avoid problematic edge cases.
One of the things I want to do on my Librem5 is monitor my servers, so that will involve polling things every few minutes. PowerTop says that I can save power by changing the polling for USB, but that changes Wifi ping times from ~1ms to ~350ms. Eventually I'll probably try experimenting with that to get an option with a 10ms ping time that still saves some power.
I get two days of intensive use from my old Huawei P30 Pro.
Same with Linux/Windows versus Macs. It is only recently that Linux/Windows laptops have begun to approach Macs in terms of battery life, and their battery efficiency is still far behind, especially with Apple Silicon being a thing.
I absolutely love Calyx, and wouldn't move from it. However, the battery life even compared to my old iPhone 6S is pretty bad. I'm not even a heavy phone user, I don't use social media or have many apps sending me notifications (only Signal) and I feel like my phone can't even last two days of light usage. My iPhone I could use similary and I could usually get 3 days of usage.
(To be fair, I ran LineageOS without Google Play Services installed, which makes a huge difference, so it's not exactly apples to apples.)
Specifically: While iPhones are noticeably more power efficient than Android phones the latter have been sufficient for my usecases especially given that there are typically options with larger batteries.
I mean...without any metrics (from you or the parent) the statement is pretty hollow and doesn't mean anything.
In absence of that, even an anecdotal comparison seems more relevant than a statement that only considers one of the two items being compared.
Update: Because I was curious whether my subjective experience was backed by real numbers, I looked up the top few Android and iPhones with the greatest battery life as per the first website I found [1] and calculated their efficiency based on their battery capacity. Various iPhone 13 models used 3.1 to 3.6 mAh per minute whereas the Android phones used 4.0 mAh/min (Moto G9 Power), 4.2 mAh/min (Samsung Galaxy A03s, Realme 9 Pro), 4.3 mAh/min (Nokia G21).
[1] https://www.techrankup.com/en/smartphones-battery-life-ranki...
If Apples battery efficiency is so good they should release a phone with a 5000mah battery /as well as/ said efficiency and market their battery life that beats all their competitors by multiple days, I'm sure a lot of people would buy that in a heartbeat.
The downsides of larger batteries are increased weight, volume, and cost. Being more efficient gets the same battery life for less of those three, allowing them to either be reduced or the budget to go to other components.
As for Xiaomi being comically large, I found the mid-range Samsung and Motorola models to have larger bezels and to generally be larger for similar specs when I bought my Xiaomi Redmi Note 7 a couple years ago. The reality is that the majority of phones from Motorola, Xiaomi, LG and Nokia are designed and manufactured by 3 Chinese ODMs (Wingtech, Huaquin and Longcheer), and even 20% of Samsung's phones come from these 3 ODMs. See: https://amosbbatto.wordpress.com/2021/12/10/comparing-l5-and...
Android has great battery life, spyware doesn't ;-).
I remember being able to run Ubuntu in the background on an unrooted Android phone while browsing the Internet. You can’t do that with iPhone.
That said, I rather have battery predictability over features, but I always thought that if Android dropped background apps, they would have the same battery usage as an iPhone.
It would be nice to be able to reliably run background apps on Android.
Also as an aside Android doesn't appear to reliably kill background processes, it kills them if it thinks that something else needs the resources. Running the Facebook app is one way of triggering Android to kill a bunch of background apps.
Unfortunately trusting developers to use those allowances wisely did not pan out.
The point was that it's reasonable to expect the battery life of plasma OS and friends to be comparable to Android, similar to how Linux on a laptop is comparable to Windows.
The Purism CSO has been running a Librem 5 as his primary desktop PC for over a year.
When Plasma was first released I was probably running hardware slower than a Librem 5.
Platform Model Life Size Efficiency
iOS Apple iPhone 13 Pro Max 21.7 4352 4.98
iOS Apple iPhone 13 16.8 3227 5.21
iOS Apple iPhone 13 Pro 16.6 3095 5.37
Android ASUS ROG Phone 5 16.6 6000 2.77
Android ASUS ROG Phone III 16.5 6000 2.75
Android ASUS ROG Phone II 16.2 6000 2.7
Android Samsung Galaxy S21 Ultra (S888) 15.9 5000 3.18
iOS Apple iPhone 11 Pro Max 15.6 3969 3.93
It's almost as if they deliberately added this seemingly bad option so people would choose the most expensive option iPhone 13 Pro Max. Because it would seem like the best deal out of three. I mean the Pro Max is just $100 more expensive than Pro, right?
But remove this decoy option, and most people would just buy the iPhone 13.
> On the iPhone 13 series, there’s a few complex behaviours to consider: First off, the iPhone 13 Pro and its LTPO panel noticeable decreases the minimum baseline power consumption of the phone by around a massive 100mW… While 100mW doesn’t sound much, when using the phones at lower screen brightness, this can represent a large percentage of the overall device power consumption, and vastly increase battery life for the new iPhone 13 Pro models.
> Comparing the 13 to the 13 Pro, the phones have quite different curves – while the 13 Pro uses less power to display full white up until 140 nits, the regular 13 becomes more efficient afterwards. We’re also seeing different curve shapes, meaning the phones are driven differently in regards to their PWM and emitter voltages.
The Pro phones are more battery efficient than the regular ones, and probably use the extra space gained with a slightly smaller battery for some other featureful hardware (cameras?).
The trade offs made seem to be reasonable for a Pro that’s between the regular and the Pro Max. It doesn’t seem to me like just a decoy.
The $100 is irrelevant to the target market of the Pro line, it's basically pocket change. You buy the phone that's most comfortable for your hand because that matters way more than the $100.
A nice by-effect will be that whatever tricks come out of the bag will almost equally be viable to boost the battery life on laptops.
It can't be possibly be attractive enough to compete with the billions of dollars that Apple can spend on this. The company has nearly $200 billion in cash sitting around. FOSS just can't compete with what Apple can invest on hardware development.
This could be the greatest phone ever made, but if it doesn't last a full day, it's worse than every other phone I've owned, or will own.
[1] 100% doesn't mean that the battery is fully charged, it means that the detected level is what the charger is prepared to go to. To conserve battery life it is not uncommon to have a 100% indicator when the battery is only 90% charged, and a 10% indicator when the battery is only 20% charged.
Although Motorola smartphones have been pretty solid for a while now, so not discounting that there has probably been a lot of R&D effort invested on their end too.
My Nokia was getting 8 days before I started putting all the apps on it. The more apps I added, the worse it got.
Nowadays I generally restrict apps from running anything in the background/sync'ing etc etc if I can.
I am happy to wait a few seconds for email to sync when I open the app rather than have my phone beep at me for every new email...
It's actually the opposite - most modern (especially flagship) phones don't really last more than a day (and here day isn't even 24 hours, but more like 16 hours). It was a massive deal when Apple finally increased the battery life of their iPhones with iPhone 13 series so it can outlast a day of normal use.
There's some difference in mid-range market though - since people buying cheaper phones tend to value battery life, you can get mid-range Android phones with downright massive batteries. Especially in Asia.
Deleted that part of my comment.
They don't run regular Android though, with Google Play, just the open source version with some Chinese marketplace. I have followed one (now dead) e-ink phone on indiegogo (forgot the name, sorry), and they wrote that Google does not let (black-and-white) e-ink smartphones pass Android certification, because they cannot correctly display the colors in their apps. So they will never run Google Play, and there will never be just e-ink phone with built-in regular Android and Google Play, sadly.
There is a hacky way to get Google Play on the hisense phones, but... ugh. That's too much hackiness for me.
Not months worth of battery saving like an eInk screen maybe but still enough to be insignificant compared to general "screen off" battery consumption.
It is baffling to me that Apple still hasn't implemented this feature even on their top end "Pro" iPhones... but this feature by itself isn't enough for me to switch to Android full time.
Since I am usually either at home or at work, it isn't far away from a charger so battery life isn't a huge deal for me.
Both of these figures are way better than they were a year ago, back then I managed to get 8 to 9 hours with WiFi off.
There’s work happening to make suspending the phone an option (that’s how the less efficient PinePhone, that’s equipped with a smaller battery manages to deliver better battery life).
In the case if my fairly new laptop battery they said it "adjusts" and gets better after a while. I'm not sure if it's true and how much better it actually gets.
https://forums.digitalspy.com/discussion/2128704/do-new-batt...
Battery life is an area that may be difficult for smaller phone makers to compete on. I think Apple especially puts a ton of engineering effort and coordination into making iOS and their apps work efficiently with their hardware, reducing complexity, runtime cycles, and power consumption as much as possible, on top of already highly-efficient ARM hardware.
Over years of doing that (kaizen), the result is optimized hardware/software fusion with industry-leading battery life. But it seems like it takes a non-trivial amount of additional engineering time and effort to accomplish this, that will be difficult to match by smaller mobile tech startups.
I hope the open source community around Librem and Pine will be able to replicate that effort, but I'm not sure this kind of consistent incremental upgrade work is attractive enough to volunteer FOSS developers. And being maximally effective at it most certainly requires the parent company to coordinate the effort across hardware, software, internal teams, and external volunteers.